Biennial Review of Health Care Management: Meso Perspectives, Volume 8

BIENNIAL REVIEW OF HEALTH CARE MANAGEMENT: MESO PERSPECTIVES ADVANCES IN HEALTH CARE MANAGEMENT Series Editors: John ...

Author: Grant Savage

32 downloads 1656 Views 1MB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!

Report copyright / DMCA form

DOWNLOAD PDF

BIENNIAL REVIEW OF HEALTH CARE MANAGEMENT: MESO PERSPECTIVES

ADVANCES IN HEALTH CARE MANAGEMENT Series Editors: John D. Blair, Myron D. Fottler and Grant T. Savage Recent Volumes: Volume 1:

Volume 2:

Volume 3:

Advances in Healthcare Management, Edited by John D. Blair, Myron D. Fottler, and Grant T. Savage Advances in Healthcare Management, Edited by Myron D. Fottler, Grant T. Savage, and John D. Blair

Advances in Healthcare Management, Edited by Grant T. Savage, John D. Blair, and Myron D. Fottler Volume 4: Bioterrorism, Preparedness, Attack and Response, Edited by John D. Blair, Myron D. Fottler, and Albert C. Zapanta Volume 5: International Healthcare Management, Edited by Grant T. Savage, Jon A. Chilingerian, and Michael Powell Volume 6: Strategic Thinking and Entrepreneurial Action in the Health Care Industry, Edited by John D. Blair, Myron D. Fottler, Eric W. Ford, and G. Tyge Payne Volume 7: Patient Safety and Health Care Management, Edited by Grant T. Savage and Eric W. Ford

ADVANCES IN HEALTH CARE MANAGEMENT VOLUME 8

BIENNIAL REVIEW OF HEALTH CARE MANAGEMENT: MESO PERSPECTIVES EDITED BY

GRANT T. SAVAGE University of Missouri School of Medicine, USA

MYRON D. FOTTLER University of Central Florida, USA

United Kingdom – North America – Japan India – Malaysia – China

Emerald Group Publishing Limited Howard House, Wagon Lane, Bingley BD16 1WA, UK First edition 2009 Copyright r 2009 Emerald Group Publishing Limited Reprints and permission service Contact: [email protected] No part of this book may be reproduced, stored in a retrieval system, transmitted in any form or by any means electronic, mechanical, photocopying, recording or otherwise without either the prior written permission of the publisher or a licence permitting restricted copying issued in the UK by The Copyright Licensing Agency and in the USA by The Copyright Clearance Center. No responsibility is accepted for the accuracy of information contained in the text, illustrations or advertisements. The opinions expressed in these chapters are not necessarily those of the Editor or the publisher. British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library ISBN: 978-1-84855-672-0 ISSN: 1474-8231 (Series)

Awarded in recognition of Emerald’s production department’s adherence to quality systems and processes when preparing scholarly journals for print

CONTENTS LIST OF CONTRIBUTORS

vii

LIST OF REVIEWERS

ix

INTRODUCTION: THE 2009 BIENNIAL REVIEW OF HEALTH CARE MANAGEMENT Grant T. Savage and Myron D. Fottler

xi

SECTION I: ORGANIZATIONAL BEHAVIOR PERSPECTIVES THE EFFECT OF EMOTIONAL EXHAUSTION AND DEPERSONALIZATION ON PHYSICIAN–PATIENT COMMUNICATION: A THEORETICAL MODEL, IMPLICATIONS, AND DIRECTIONS FOR FUTURE RESEARCH Eric S. Williams, Ericka R. Lawrence, Kim Sydow Campbell and Steven Spiehler USING SELF-CONCEPT THEORY TO IDENTIFY AND DEVELOP VOLUNTEER LEADER POTENTIAL IN HEALTHCARE Francine Schlosser, Deborah M. Zinni and Andrew Templer LEADERSHIP STRATEGIES FOR BIOTECHNOLOGY ORGANIZATIONS: A LITERATURE REVIEW Lynn Johnson Langer

v

3

21

49

vi

CONTENTS

ATTRIBUTION THEORY AND HEALTHCARE CULTURE: TRANSLATIONAL MANAGEMENT SCIENCE CONTRIBUTES A FRAMEWORK TO IDENTIFY THE ETIOLOGY OF PUNITIVE CLINICAL ENVIRONMENTS Patrick A. Palmieri and Lori T. Peterson

81

SECTION II: ORGANIZATIONAL DEVELOPMENT AND STRATEGIC MANAGEMENT PERSPECTIVES MEASURING UP: ARE NURSE STAFFING MEASURES ADEQUATE FOR HEALTH SERVICES RESEARCH? Lynn Unruh, C. Allison Russo, H. Joanna Jiang and Carol Stocks MATRIX MENTORSHIP IN ACADEMIC MEDICINE: SUSTAINABILITY OF COMPETITIVE ADVANTAGE Jay A. Fishman THE IMPACT OF HOSPITAL OWNERSHIP CONVERSIONS: REVIEW OF THE LITERATURE AND RESULTS FROM A COMPARATIVE FIELD STUDY Lawton R. Burns, Rajiv J. Shah, Frank A. Sloan and Adam C. Powell

115

155

171

LIST OF CONTRIBUTORS Lawton R. Burns

Departments of Management and Health Care Management, The Wharton School, University of Pennsylvania, Philadelphia, PA, USA

Kim Sydow Campbell

Department of Management and Marketing, Culverhouse College of Commerce and Business Administration, University of Alabama, Tuscaloosa, AL, USA

Jay A. Fishman

MGH Transplant Center and Infectious Disease Division, Massachusetts General Hospital and Harvard Medical School, Boston, MA, USA

H. Joanna Jiang

Center for Delivery, Organization, and Markets, Agency for Healthcare Research and Quality, Rockville, MD, USA

Lynn Johnson Langer

Advanced Biotechnology Studies, Johns Hopkins University, Rockville, MD, USA

Ericka R. Lawrence

Department of Management and Marketing, Culverhouse College of Commerce and Business Administration, University of Alabama, Tuscaloosa, AL, USA

Patrick A. Palmieri

Center for American Education, School of Administrative Sciences, Universidad San Ignacio de Loyola (USIL) Lima, Peru

Lori T. Peterson

Department of Management and Labor Relations, Nance College of Business, Cleveland State University, Cleveland, OH, USA vii

viii

LIST OF CONTRIBUTORS

Adam C. Powell

Department of Health Care Management, The Wharton School, University of Pennsylvania, Philadelphia, PA, USA

C. Allison Russo

Thomson Reuters, Washington, DC, USA

Francine Schlosser

Odette School of Business, University of Windsor, Windsor, Ontario, Canada

Rajiv J. Shah

Director of Agricultural Development, Bill and Melinda Gates Foundation, Seattle, WA, USA

Frank A. Sloan

Center for Health Policy, Law & Management, Duke University, Durham, NC, USA

Steven Spiehler

Department of Management and Marketing, Culverhouse College of Commerce and Business Administration, University of Alabama, Tuscaloosa, AL, USA

Carol Stocks

Center for Delivery, Organization, and Markets, Agency for Healthcare Research and Quality, Rockville, MD, USA

Andrew Templer

Odette School of Business, University of Windsor, Windsor, Ontario, Canada

Lynn Unruh

Department of Health Professions, College of Health Services Administration, Health and Public Affairs, University of Central Florida, Orlando, FL, USA

Eric S. Williams

Department of Management and Marketing, Culverhouse College of Commerce and Business Administration, University of Alabama, Tuscaloosa, AL, USA

Deborah M. Zinni

Faculty of Business, Brock University, St. Catharines, Ontario, Canada

LIST OF REVIEWERS Ruth A. Anderson Duke University, USA

Louis D. Marino The University of Alabama, USA

Margarete Arndt Clark University, USA

Ann S. McAlearney The Ohio State University, USA

Barbara Bigelow Clark University, USA

Kathleen Montgomery University of California – Riverside, USA

Eric W. Ford Texas Tech University, USA

Stephen J. O’Connor University of Alabama – Birmingham, USA

Leonard H. Friedman George Washington University, USA

Cheryl Rathert University of Missouri, USA

Jonathon R. B. Halbesleben University of Wisconsin – Eau Claire, USA

Leo van der Reis University of Missouri, USA Jacqueline Zinn Temple University, USA

Naresh Khatri University of Missouri, USA

ix

This page intentionally left blank

INTRODUCTION: THE 2009 BIENNIAL REVIEW OF HEALTH CARE MANAGEMENT Advances in Health Care Management is a peer-reviewed research annual that publishes state-of-the-art reviews (odd years) and research on special topics (even years) in the ﬁeld of health care management. As conceived by the founding editors, John D. Blair, Myron D. Fottler, and Grant T. Savage, and as originally commissioned by JAI Press, Advances in Health Care Management provides a forum for leading research on health care management. Volumes 1–3 offer reviews of the ﬁeld, research on selected topics, and best papers from the Health Care Management Division of the Academy of Management. In contrast, volumes 4–7 focus on a range of special topics, from bioterrorism to international health care management to entrepreneurship to patient safety. Volume 8 is the inaugural volume for the biennial review of health care management research. This volume provides state-of-the art reviews of health care management, linking concerns about health care workforce management with health care organization management issues. This meso perspective, linking micro- and macro-organizational processes, should appeal to health care management researchers and doctoral students. Review articles in this volume orient new and established scholars about current themes within health care management as well as emerging themes and divergent views. The authors evaluate these future directions and offer their perspective on the direction or directions that would help build theory and improve the practice of health care management. Speciﬁcally, the volume focuses both on health care workforce management issues, including nurses, physicians, and volunteers, and on health care organization management issues, ranging from performance metrics to leadership development to strategy within health care organizations. Four chapters draw upon organizational behavior perspectives, and three chapters focus on organizational development and strategic management perspectives. xi

xii

INTRODUCTION

We discuss each of these chapters and their contributions in the following sections.

ORGANIZATIONAL BEHAVIOR PERSPECTIVES The Effect of Emotional Exhaustion and Depersonalization on Physician–Patient Communication: A Theoretical Model, Implications, and Directions for Future Research Eric Williams and his colleagues review the literature on both physician burnout and physician–patient communication. A major contribution in this chapter is a model based on these two literatures, which outlines the impact that physician burnout can have on the physician–patient interaction and, therefore, patient outcomes. When physicians become emotionally exhausted, they begin to depersonalize to cope and focus on biomedical issues rather than communicating with the patient. When the patient is approached with this communication style from their physicians, they become less satisﬁed, trusting, and compliant. Less compliance results in worsened clinical outcomes, especially for patients with chronic disease. The authors discuss both the implications of this model and future directions for research.

Using Self-Concept Theory to Identify and Develop Volunteer Leader Potential in Healthcare Francine Schlosser and her colleagues review and apply self-concept theory to health care volunteers in the Canadian health care industry. They propose that volunteer leaders are differentiated from other leaders and volunteers by how they view their roles in the organization and their ability to make a difference in these roles. This interpretation reﬂects self-concept theory because each person’s self-concept inﬂuences self-perception, reactions to experiences, and shapes motivation. A case study of one Canadian hospital is used to illustrate the impact of a volunteer leader’s self-concept on his or her behavior based on a series of focus groups and interviews with key stakeholders. The management of student volunteers was identiﬁed as a particularly problematic area of volunteer management. Their study concludes with a discussion of the implications for managing volunteer

Introduction

xiii

leaders and volunteers, which include better selection processes, leadership coaching, training and development, and reward programs. Leadership Strategies for Biotechnology Organizations: A Literature Review Lynn Johnson Langer reviews the literature on leadership strategies in the biotechnology industry. Most biotechnology ﬁrms are started by basic science researchers, and the core workforce in these ﬁrms typically includes other research scientists. She identiﬁes several leadership themes that are speciﬁc to biotechnology, including the continuous learning and adaptability of these research science leaders, which inﬂuences their ability to lead effectively in a dynamic environment. Other precepts she draws from the literature include the need for biotechnology leaders to create learning organizations, to express their vision throughout their organization, to be strategic decision-makers, and to focus on creating a culture and working environment that aligns with their vision. She ends by noting the need for additional research that examines the leadership needs of biotechnology organizations as they progress from small start-up ﬁrms to large, multinational corporations. Attribution Theory and Healthcare Culture: Translational Management Science Contributes a Framework to Identify the Etiology of Punitive Clinical Environments Patrick Palmieri and Lori Peterson consider adverse events in health care through the lens of attribution theory. To date, this cognitive theory, which is well-known in the organization behavior and management literature, has not been translated to the health care work environment. Palmieri and Peterson demonstrate that attribution theory provides the cognitive rationale to explain the etiology of punitive hospital cultures, in which clinicians are routinely blamed for adverse patient events. They discuss the historical evolution of attribution theory in relation to human behavior in clinical practice and then discuss work environments existing in today’s health care organizations. Their Health Care Attribution Error model illustrates how the concepts from attribution theory are related to the emerging Just Culture perspective. In summary, this chapter provides managers and clinicians with theoretically supported insights to reduce inappropriate and inaccurate attributions to improve the health care work environment.

xiv

INTRODUCTION

ORGANIZATIONAL DEVELOPMENT AND STRATEGIC MANAGEMENT PERSPECTIVES Measuring Up: Are Nurse Stafﬁng Measures Adequate for Health Services Research? Lynn Unruh and her colleagues examine nurse stafﬁng measures currently being used in health services research. More speciﬁcally, their study assesses whether the measures, sampling framework, and data sources meet the needs for research in the areas of stafﬁng assessment; patient, nurse, and ﬁnancial outcomes; and predictors of nurse stafﬁng. They perform a systematic review of articles from 1990 to 2007 that use hospital nurse stafﬁng measures or address the validity, reliability, and availability of these measures. Their review identiﬁes more than 100 articles that use nurse stafﬁng measures for original research. Data sources range from small-scale surveys to national databases such as that of the American Hospital Association. The latter is the most frequently used data source, but its nurse stafﬁng measures have several limitations. The data limitations noted in this review and previous research suggest a need for improvement in nursing stafﬁng data content, scope, and availability. The authors conclude with implications and recommendations for enhancing nurse stafﬁng data, research, and administrative practice. Matrix Mentorship in Academic Medicine: Sustainability of Competitive Advantage Jay Fishman reviews the sparse literature on academic medical center strategic competitiveness and the emerging literature on physician mentorship. He outlines the strategic challenges currently facing academic medical centers to continue competing in the academic world while also competing with other health care organizations in terms of efﬁciency, quality, and safety. Unfortunately, academic medical centers are ill-prepared for the leadership challenges posed by this environment. He argues that they need to create physician change agents willing to assume leadership roles to guide the academic medical center’s evolution. His review of traditional, one-on-one relationships between mentors and trainees demonstrates that they do not provide the breadth of guidance needed for this evolving environment. Dr. Fishman suggests that a structured system of ‘‘matrix mentorship’’ would provide an essential set of physician leader skills and values to enhance

Introduction

xv

competitive advantage. He cautions, however, that a formal program in matrix mentorship must include multiple mentors to guide all aspects of the professional and psychosocial development of future physician leaders as well as a large and sustainable institutional commitment of resources. The Impact of Hospital Ownership Conversions: Review of the Literature and Results from a Comparative Field Study Lawton Burns and his colleagues examine the nature of various types of hospital ownership conversions and their impacts on a variety of hospital performance measures such as access, quality, and cost. The ﬁrst part of their study is a literature review that examines the rate at which conversions have occurred over time, the relative frequency in conversions between speciﬁc ownership categories (i.e., non-proﬁt to for-proﬁt) and the effects of conversion on hospital operations and performance. They conclude that the impact of ownership conversion on different outcome measures is mixed with some evidence of improved hospital efﬁciency. They suggest that the impact of conversion may be mitigated by changes in a hospital’s strategic content and processes. The second part of their study is a comparative ﬁeld study of 16 hospitals that experienced ownership conversions from nonproﬁt to for-proﬁt, public to for-proﬁt, public to non-proﬁt, and for-proﬁt to non-proﬁt. They ﬁnd that hospitals change ownership for ﬁnancial reasons, experience revenue increases, experience increases in capital investment, and pursue labor force reductions post-conversion. They conclude that the speciﬁc changes in ownership may be less important than the change from free-standing status to multihospital system membership in explaining post-conversion strategies and outcomes. In closing, the chapters in this inaugural biennial review volume offer a rich resource for health care management and health services researchers. The various authors contribute to the discipline by systematically reviewing extant literature and by synthesizing various streams of literature and offering new theoretical models and perspectives. As a result, health management and health service researchers will ﬁnd much to learn and to further explore from reading this volume of Advances in Health Care Management. Grant T. Savage Myron D. Fottler Editors

This page intentionally left blank

SECTION I ORGANIZATIONAL BEHAVIOR PERSPECTIVES

This page intentionally left blank

THE EFFECT OF EMOTIONAL EXHAUSTION AND DEPERSONALIZATION ON PHYSICIAN–PATIENT COMMUNICATION: A THEORETICAL MODEL, IMPLICATIONS, AND DIRECTIONS FOR FUTURE RESEARCH Eric S. Williams, Ericka R. Lawrence, Kim Sydow Campbell and Steven Spiehler ABSTRACT The physician–patient relationship is the cornerstone of care quality. Unfortunately, it may be adversely affected by physician burnout, which is becoming more prevalent according to the literature. We present a model, based on the burnout and physician–patient communication literatures, which delineates the impact of physician burnout on the physician–patient interaction and ultimately on patient outcomes. In short, when physicians Biennial Review of Health Care Management: Meso Perspectives Advances in Health Care Management, Volume 8, 3–20 Copyright r 2009 by Emerald Group Publishing Limited All rights of reproduction in any form reserved ISSN: 1474-8231/doi:10.1108/S1474-8231(2009)0000008005

3

4

ERIC S. WILLIAMS ET AL.

use depersonalization to cope with emotional exhaustion, their communication style becomes more biomedically oriented. Faced with this communication style when interacting with their physician, patients are less satisﬁed, trusting, and adherent. The implications of this model and directions for future research are presented.

Burnout is deﬁned as a psychological response to work stress, which consists of three principle elements: emotional exhaustion, depersonalization, and lack of personal achievement (Maslach & Jackson, 1986). Emotional exhaustion is the feeling or sensation of not caring about things or people previously considered important. Simply put, burned out people are emotionally spent. Depersonalization is an emotional detachment from others. These individuals withdraw from their environment. Lack of professional achievement is a strong sense that achievements mean little. It is essentially a substantial discounting of ones actions or achievements. Historically, burnout has been investigated in service occupations (Fruedenberger, 1974). Investigators have been interested in the causes of burnout as well as its impacts on service quality and performance (Halbesleben & Rathert, 2008; Maslach & Leiter, 1997). Within the health care industry, one important question concerns the impact of physician burnout on the physician–patient relationship. This is not an idle question given the increasingly stressful medical workplace brought on by changes in contemporary medical care that include disparities in access and quality, inequities in compensation, and increased work demands with decreased control over multiple aspects of daily work life. Researchers have documented high levels of burnout in both academic physicians (Ramirez et al., 1995) and private practice physicians (Deckard, Hicks, & Hamory, 1992; Whippen & Canellos, 1991). Additional research suggests physician burnout may result in increased absenteeism, increased turnover intentions, anxiety, depression, and lower job performance (Halbesleben & Bowler, 2007; Maslach, Schaufeli, & Leiter, 2001). A subtle and perhaps more pernicious aspect of physician burnout is its impact on patient care quality. The available research suggests burnt-out physicians may not be able to deliver safe, high-quality patient care. Shirom, Nirel, and Vinokur (2006) reported that increased levels of emotional exhaustion were a negative predictor of quality of patient care, whereas Keijsers, Schaufeli, LeBlanc, Zwerts, and Miranda (1995) found linkages between all three burnout dimensions with both objective and subjective measures of unit performance. One study, performed in an internal medicine

Emotional Exhaustion, Depersonalization and Physician–Patient Communication

5

residency program (Shanafelt, Bradley, Wipf, & Back, 2002), reported that three-quarters of the residents exhibited symptoms of burnout and that these burnt-out physicians reported more suboptimal patient care practices than their colleagues. Of the three components of burnout, only depersonalization was associated with suboptimal patient care. A similar study found that burnout was related to physician-reported likelihood of making an error and suboptimal patient care (Williams, Manwell, Konrad, & Linzer, 2007). Dugan et al. (1996) examined the effect of job stress among nurses on turnover, absenteeism, injuries, and patient incidents. Signiﬁcant zero-order correlations were found between a perceived stress index and patient incidents overall, medication errors, and patient falls, but not IV errors. However, a second stress measure composed of reported stress symptoms found more modest, non-signiﬁcant correlations. A report prepared for the Agency for Healthcare Quality and Research (Hickam et al., 2003) on the linkage between working conditions (including job stress, dissatisfaction, and burnout) and patient safety cautioned that these results are suggestive, but not conclusive given the limited literature.

GAP IN THE LITERATURE Both the burnout and physician–patient communication literatures are voluminous and well established. The burnout literature tends to focus on antecedents, with limited attention to its outcomes (Halbesleben & Rathert, 2008). The physician–patient communication literature has the opposite shortcoming. It examines the impact of communication behaviors with outcomes such as patient satisfaction, recall, compliance, but has had little to say on antecedents of physician communication behaviors (Hall, Roter, & Katz, 1988). Although research has linked physician burnout with patient outcomes (Halbesleben & Rathert, 2008), it has not looked at the speciﬁc mechanisms by which burnout affects patient outcomes through physician– patient communication. The gap that exists at the intersection of these two literatures suggests one such mechanism. That is, neither literature examines the effect of burnout, or more correctly, depersonalization, on the quality of the physician–patient relationship: speciﬁcally, the verbal and non-verbal behaviors of burnt-out physicians and the impact of those behaviors on patient outcomes. This gap has not gone unnoticed (Halbesleben, 2006; Williams, Savage, & Linzer, 2006). Hall, Horgan, Stein, and Roter (2002) note that ‘‘motives and

6

ERIC S. WILLIAMS ET AL.

emotional states that inﬂuence how physicians and patients behave’’ need development in the literature. This chapter presents a theoretical framework that explains how physician burnout may lead to suboptimal communication behaviors, which, in turn, result in poor patient outcomes. In developing this framework, diverse literature on physician burnout and physician–patient communication will be synthesized. The implications of this model and directions for future research will be presented.

A THEORETICAL FRAMEWORK Fig. 1 illustrates our theoretical framework, which suggests that emotionally exhausted doctors may use depersonalization as a coping mechanism during interactions with patients. Depersonalization in verbal (a strong focus on biomedical content) and non-verbal communication behaviors (limited eye contact, body orientation away from patient) undermines the physician– patient relationship. Patients react negatively to such behaviors and adverse outcomes ensue. The theoretical underpinnings of this model comes from Hobfoll’s (1989) Conservation of Resources (COR) model, which examines the psychological process underlying stress using the lens of valued resources. Resources are deﬁned ‘‘as those objects, personal characteristics, conditions, or energies that are valued by the individual or that serve as a means for attainment of these objects’’ (Hobfoll, 1989, p. 516). Stress results from one of three processes: (1) loss of resources, (2) threat to resources, or (3) inadequate return on resource investment. Hobfoll and Freedy (1993) extended the COR model to burnout through the idea that burnout results from inadequate return for work resources invested across time. That is, work resources are continually invested, but are depleted (or, at least, placed at risk) as they do not generate sufﬁcient return. This resource depletion is particularly insidious as resource loss has greater psychological import than resource gain (Tversky & Kahneman, 1974). Hobfoll and Freedy (1993) Emotional Exhaustion

Depersonalization • Cynicism • Withdraw

Fig. 1.

Medical Encounter • Biomedical Content

Conceptual Model

Patient • Satisfaction • Trust • Adherence • Health Status

Emotional Exhaustion, Depersonalization and Physician–Patient Communication

7

suggest then that ‘‘workers are more sensitive to workplace phenomena that translate into losses for them’’ (p. 118). In a longitudinal study, Bakker, Schaufeli, Sixma, Bosveld, and Dierendonck (2000) investigated the relationship of burnout and patient demands. They found that physician perceptions of lack of reciprocity on the part of patients were related to a physician’s emotional exhaustion. The remainder of this section will further elucidate the theoretical model through the consideration of the three sets of links: emotional exhaustion – depersonalization, depersonalization – physician communication behaviors; physician communication behaviors – patient outcomes.

Emotional Exhaustion – Depersonalization Hobfoll’s (2001) forth corollary of COR theory suggests ‘‘those who lack resources are likely to adopt a defensive posture to conserve their resources’’ (p. 356). That is, as physicians become increasingly emotionally exhausted and command fewer resources, they cope by being increasingly careful about how they invest their resources at work. This idea corresponds with the earlier process models of burnout (Cherniss, 1980; Golembiewski & Munzenrider, 1988; Leiter & Maslach, 1988). These models theorize that as emotional exhaustion develops as a result of substantial and sustained job demands (Karasek, Baker, Marxer, Ahlbom, & Theorell, 1981; Landsbergis, 1988; Schaufeli & Bakker, 2003), some doctors respond by depersonalizing their interactions with clients, exhibiting cynicism and withdrawal (thereby preserving scarce resources). The empirical literature supports the link between emotional exhaustion and depersonalization (Lee & Ashforth, 1993; Leiter, 1991). For example, in 178 matched pairs of patients and physicians, Halbesleben and Rathert (2008) found that physician emotional exhaustion was positively related to physician depersonalization. When examining patient encounters, they found that patients of burnt-out physicians perceived depersonalization in their physician and responded with less satisfaction and longer recovery times.

Depersonalization – Physician Communication Behaviors The impact of depersonalization and withdrawal on physician communication behaviors lies at the core of our model. Moore (2000) deﬁned depersonalization as ‘‘a negative, callous, or excessively detached response

8

ERIC S. WILLIAMS ET AL.

to other people who are usually the recipients of one’s service or care.’’ Maslach and Leiter (1997) describe depersonalization as cynicism, ‘‘a cold distant attitude toward work and the person on their job’’ (p. 18). They go on to suggest that such individuals decrease their work involvement and, in extreme cases, retreat from their ideals. In this context, COR theory focuses on resource conservation. This typically translates into a strong focus on biomedical content with little, if any, psychosocial content by the physician in a medical encounter. At least since Goffman (1967) we have recognized that interaction is a resource-intensive activity because of the simultaneous need to protect our own face and support our partner’s face. Simply put, engaging in biomedical communication consumes fewer resources than engaging in both biomedical and psychosocial communication. Support for this observation comes out of Campbell’s (2008) adaptation of the competing values framework for manager–subordinate communication (Quinn, Hildebrandt, Rogers, & Thompson, 1991) (Fig. 2) to physician–patient communication. In this framework, two axes create four quadrants of ‘‘values’’: the horizontal axis represents a continuum beginning on the left with ‘‘conventional structure’’ and ending on the right with ‘‘dynamic content;’’ and the vertical axis represents a continuum beginning at the bottom with ‘‘instrumental logic’’ and ending at the top with ‘‘relational awareness.’’ Each quadrant reﬂects a different value orientation

Fig. 2.

Competing Values Model of Physician–Patient Communication (Quinn et al., 1991)

Emotional Exhaustion, Depersonalization and Physician–Patient Communication

9

on the part of a physician toward interaction with a patient. The top-left quadrant suggests that when physicians communicate relational messages, they function as ‘‘motivators’’ and establish credibility, generate trust, and build rapport with patients. Despite their conventional structure, such messages require emotional resources to achieve relational awareness; such resources may be in short supply for the burnt-out physician. In contrast, in the bottom-left quadrant, physicians function as ‘‘analysts’’ when they communicate informational messages. The conventional structure and instrumental logic of such messages establish clear understanding with patients but do not require emotional resources. Hence, burnt-out physicians may rely more heavily on such messages in patient interaction. The bottom-right quadrant suggests that when physicians communicate instructional messages, they function as ‘‘task-masters,’’ who direct the actions of patients through dynamic content and instrumental logic. Finally, in the top-right quadrant, physicians function as ‘‘information-gatherers’’ and stimulate patient contribution and change. As with the other top quadrant, the prerequisite relational awareness of such psychosocial messages implies that burnt-out physicians might conserve emotional resources by minimizing such messages in patient interaction. Prior physician communication research makes clear that physicians do sometimes communicate with limited relational awareness of their patients and that such behavior is less than optimal. For instance, Roter et al. (1997) empirically derived a model of physician communication styles based on actual primary care encounters: narrow biomedical, expanded biomedical, biopsychosocial, psychosocial, and consumerist. The narrow biomedical style (32% of visits) focused strictly on biomedical issues characterized by closed-ended questions (i.e., communicating primarily informational messages within a single quadrant of the competing values framework). Buller and Buller (1987) label a focus on biomedical content and control over the medical encounter through information withholding as a ‘‘controlling style.’’ In this style, physicians used closed-ended questions, provided limited information, and used medical jargon. Similarly, Emanuel and Emanuel (1992) identify a ‘‘paternalistic’’ style in which the physician’s role is to deliver the most appropriate treatment as determined by the physician, and the role of the patient is passive and limited to adhering to treatment. Roter et al.’s (1997) expanded biomedical style (33% of visits) also focused on biomedical content, but included more psychosocial content and a substantial amount of physician-directed questioning (i.e., communicating messages within several quadrants of the competing values framework). They also found that, while all physicians sometimes used more than one

10

ERIC S. WILLIAMS ET AL.

style, a majority used one pattern for most of their visits, with most physicians preferring one of the two biomedical patterns. Even if a burnt-out physician wants to maintain a communication including psychological content (i.e., transformational and relational messages in the competing values framework), they may lack the emotional resources to communicate such messages convincingly when suffering from burnout. The emotional labor literature (Hochschild, 1983) suggests that internal emotions exert a substantial inﬂuence on displayed emotions and, presumably, communication behaviors (Pugh, 2001). Brief (1998) suggests two models. The ﬁrst contends that when people try to conceal or fake a particular emotion (which is especially difﬁcult if the emotion to be hidden is strong), true emotions will leak out (Ambady & Rosenthal, 1992; Ekman, 1985). Brief (1998) also suggest that emotional contagion (Hatﬁeld, Cacioppo, & Rapson, 1992, 1994) may be at work. In this model, participants in a service encounter tend to, without conscious thought, mimic one another to arrive at a similar emotional state. Thus, an emotionally exhausted physician may try to hide his or her distress, but unconsciously betray inner feelings through a biomedical communication style. Regardless of which model may be at work, physicians may successfully disguise their burnout only to a limited degree. Eventually, emotional exhaustion and depersonalization, especially in severe cases, may win out and result in communication at odds with optimal physician–patient communication. We argue, following COR, that emotionally exhausted physicians reduce their resource investment in patient interaction in part by depersonalizing their communication behaviors: they focus on messages requiring relatively little relational awareness within the competing values framework. Such depersonalization behaviors include limiting the amount of communication with patients, especially transformational and relational interaction, and focusing instead on biomedical content (i.e., informational and instructional interact).

Biomedical Communication and Reciprocity in Patient Outcomes The insidious aspect of the burnout process is that physicians do not suddenly decide to reduce their resource investments; such changes occur over time. Changes in the level of relational awareness or engagement by physicians during medical encounters are a violation of the norm of ‘‘reciprocity’’ embedded in the joint expectations of a medical encounter.

Emotional Exhaustion, Depersonalization and Physician–Patient Communication 11

Social exchange theory (Blau, 1964) suggests that ‘‘balance’’ in the physician–patient relationship is maintained through provision of professional services by the physician and by the patient’s gratitude, trust, compliance with medical advice, and symptom alleviation. Two studies suggest that patients clearly understand this balance and are unhappy when the balance is upset. Hall et al. (2002) examined the mutuality of liking between physicians and patients. They found that ‘‘how much each liked the other was related to how much each was liked,’’ suggesting reciprocity in terms of liking. Moreover, they found that these impressions were generally accurate, suggesting that neither physician nor patient was fully able to disguise their feelings for the other. Even more interesting was a very strong correlation between patient liking for the physician and patient description of the physician’s behavior during the visit. The investigators suggest that patients may be ‘‘grounding their liking on how appropriate or inappropriate they consider their physicians’ behavior to be.’’ Similarly, Beach, Roter, Wang, Duggan and Cooper (2006) investigated the accuracy with which patients perceived the level of respect physicians displayed toward them and the level of positive communication behaviors. They found that patients were able to accurately perceive when they were respected by their physicians. Furthermore, when physicians felt respect toward a patient, they provided more information and expressed more positive affect with those patients. Thus, reciprocity goes both ways. If a physician is able to fully engage with a patient both biomedical and psychosocially, the patient will tend to respond favorably. If a physician is distant (withdrawn or depersonalized), focusing on biomedical information, patients will tend to respond negatively (Roter et al., 1997). The literature seems to support the contention that patients of depersonalizing physicians will be less satisﬁed. For instance, Stewart (1984) examined audiotapes from 140 medical encounters in 24 physician ofﬁces. Analyses revealed that a higher frequency of patient-centered (i.e., transformational and relational), as opposed to physician-controlled (i.e., informational and instructional), behavior was strongly related to treatment compliance and satisfaction. Bertakis, Roter, and Putnam (1991) investigated 550 patient visits with 127 physicians in 11 geographically disbursed sites. They found that encounters focused on biomedical content were negatively related to satisfaction, whereas those focused on psychosocial content were positively related. They also found that physician-controlled conversation was associated with lower patient satisfaction. Using the same dataset, Roter et al. (1997) found that patients were least satisﬁed with biomedical communication styles. An even more intriguing ﬁnding was that

12

ERIC S. WILLIAMS ET AL.

physicians’ own satisfaction was lowest when they interacted in a narrow biomedical style. Although a comprehensive review of the research on individual communication behaviors and patient outcomes is beyond the scope of this work, a brief look at several meta-analyses is revelatory. Hall et al. (1988) found patient satisfaction was strongly associated with information giving and partnership building, but not with question asking. They also found that patient compliance was less strongly related to such behaviors than satisfaction, but was associated with giving more information, asking fewer overall questions, asking more questions about compliance, more positive talk, and less negative talk. Finally, they found patient recall and understanding were predicted by more information giving, less question asking, more partnership building, and more positive talk. In Stewart’s (Stewart, 1995) review of 21 studies reported positive associations between quality of communication and patient health outcomes. Question asking was related to reduced patient distress and anxiety as well as increased symptom relief. Patients who were encouraged to ask more questions were successful at obtaining information were provided with information programs and were given clear information, tended to respond with less anxiety and distress and perceived fewer physical and role limitations. Beck, Daughtridge and Sloane (2002) reviewed 22 studies assessing physician verbal and non-verbal behaviors with patient outcomes and identiﬁed 14 behaviors associated with negative health outcomes: negative social– emotional interactions, formal behavior, passive rejection, high rates of biomedical questioning, interruptions, one-way information ﬂow, directive behavior, irritation, nervousness, anxiety, tension, and dominance. Taken as a whole, these results suggest a grim picture of how the communication behaviors of emotionally exhausted, depersonalizing physicians negatively inﬂuence patient outcomes.

IMPLICATIONS AND DIRECTIONS FOR FUTURE RESEARCH COR theory suggests that building up personal resources or interrupting resource loss spirals are the key to successful interventions (Hobfoll & Freedy, 1993). In keeping with this, we present three implications and directions for future research focusing on physician–patient communication. Parenthetically, it should be noted that effective interventions to manage

Emotional Exhaustion, Depersonalization and Physician–Patient Communication 13

physician (and health provider) burnout would make these implications less salient. However, the high rates of burnout and poor communication behaviors documented in this work argue for the importance of both decreasing physician burnout and improving physician-patient communication. The ﬁrst implication of the biomedical communication style is rather obvious: when physicians experience burnout and engage in depersonalization in an effort to conserve emotional resources, their communication style during the medical encounter is more focused on the informational and instructional quadrants of the competing values framework instead of the transformational or relational quadrants. The physician’s inclination to engage in communication styles that are not transformational may present problems in the physician–patient relationship not only because patients prefer more patient-centric communication styles (Beck et al., 2002; Roter et al., 1997), but these styles may actually produce better clinical outcomes (Hsiao & Boult, 2008) through better compliance with medical treatment (Golin, DiMatteo, & Gelberg, 1996). Hsiao and Boult (2008) reviewed the literature on various quality outcomes with actual patient outcomes, identifying ﬁve studies looking at the impact of communication and patient outcomes. All ﬁve studies supported the idea that transformational communication (e.g., active listening, encouragement of discussion of complains, and concerns) was associated with better health outcomes. Kinmouth, Woodcock, Grifﬁn, Spiegal, & Campbell (1998) found that type 2 diabetes patients that received patient-centered communication (e.g., active listening, encouragement of discussion of complications, and concerns) reported greater well-being than patients who received routine care. In a longitudinal study, Heszen-Klemmens and Lapinska (1984) found that positive physician behaviors predicted patient adherence and both subjective and objective measures of patients’ health status. Having a physician with a patientcentric communication style is even more important for patients with chronic diseases. For example, Kaplan, Greenﬁeld, and Ware (1989) report on four clinical trials examining the effect of communication behaviors on patient outcomes. They found that behaviors associated with a patientcentered style were associated with positive health outcomes. Future research should examine speciﬁc behaviors that are exhibited by physicians in medical encounters when they are burned out and engaging in depersonalization. Ideally, the design for the study should follow the methodology developed by Roter and colleagues. Such a design should include several elements: (1) measurement of physician emotional exhaustion and depersonalization, (2) use of an analysis technique speciﬁcally

14

ERIC S. WILLIAMS ET AL.

designed for this purpose (e.g., Roter Interactive Analysis System), and (3) audio and videotaping of the encounters, so that raters can analyze both verbal and non-verbal aspects of medical encounters. Future research may also want to link such a study with patient outcomes such as patient satisfaction, trust, or to clinical outcomes similar to that of Halbesleben and Rathert (2008). Ratanawongsa et al. (2008) used several of these elements. They, unfortunately, found that there were no signiﬁcant differences in physician communication based on physician burnout. However, patients of highburnout physicians gave twice as many negative rapport-building statements as compared to patients of low-burnout physicians. Physician burnout was not signiﬁcantly related to physician or patient affect, patient-centeredness, verbal dominance, or length of the encounter. Physician burnout was also not signiﬁcantly related to patients’ ratings of their satisfaction, conﬁdence, or trust. However, patients were found to engage in more rapport-building behaviors. While valuable, this effort was limited by a focus on burnout in general rather than a speciﬁc focus on depersonalization (and emotional exhaustion), the long latency between burnout measurement and patient encounters (15 months on average), and small sample sizes (40 physicians with 235 patients). Future research would be well advised to follow this work, but use larger samples, ensure much less latency between burnout measurement and patient encounter, measure both emotional exhaustion and depersonalization, and both videotape and audiotape the encounter. The second implication of our model lies in the literature dedicated to training physicians in more patient-centric styles. Although the medical interview is taught with increasing emphasis on patient-centered communication models (Makoul, 1999), many physicians continue to use the biomedical communication style as their default (Roter et al., 1997) despite its association with negative patient outcomes. Research suggests that communication interventions focused especially on patient-centered communication (Rao, Anderson, Inui, & Frankel, 2007) affect the quality of medical encounters and patient outcomes. Roter et al. (1995) designed a clinical trial to assess a communication skills training program focused on reducing patients’ emotional distress (e.g., depression, anxiety). The two 4-hour training sessions focused on the types of psychosocial issues found in primary practice and the communication skills (e.g., active listing) associated with successful reduction of patients’ emotional distress. The ﬁnal part of the training employed practice with simulated patients followed by feedback. Physicians in the experimental group diagnosed more

Emotional Exhaustion, Depersonalization and Physician–Patient Communication 15

psychological problems in their patients, and their patients experienced less emotional distress at two-week, one-month, and six-month assessments. Trummer, Mueller, Nowak, Stidl, and Pelican (2006) designed a clinical trial to train physicians and nurses on a surgery unit on communication techniques that empower patients to be more effective co-producers of their own recuperation after surgery. Patients of physicians and nurses in the experimental group experienced lower length of stay, lower incidence of post-surgery tachyarrhythmia, faster transfers out of the ICU, and higher patient ratings of communication quality. Tamblyn et al. (2007) obtained clinical skills examination scores from Canadian physicians who were completing requirements for licensure. The investigators found that low patient–physician communication scores in the clinical skills examination predicted future complaints to medical regulatory authorities. As mentioned, previous research has suggested that physician training improves physician communication in the medical encounter and results in positive outcomes for both patients and physicians. However, physician training may induce some pressures on physicians. Therefore, future research should elucidate both positive and negative consequences of physician communication skills training. Haskard et al. (2008) found that trained physicians exhibited improved information-giving and improved quality of care; however, they also experienced a signiﬁcant decrease in satisfaction with the interpersonal aspects of their professional lives. This decrease may be related to perceptions of the additional demands associated with training in conjunction with other current organizational and occupational demands (i.e., increased administrative duties). Researchers may also want to look at the differential effects of training due to physician gender, age, and practice experience. For instance, it may be hypothesized that the negative relationship between physician training and life satisfaction may be stronger for women than for men. Another line of future research might examine the short- and long-term effects of physician training. In the short term, the time demands of training and the emotional demands of personal change may cause an increased amount of stress and reduce a physician’s resource investment in the patient interaction. However, physician communication may improve in the long term as a result of a skill-consolidation effect (Haskard et al., 2008). The third implication of our model is the potential beneﬁt of training patients in communication to overcome ineffective communication behavior from burnt-out physicians. This concept is an extension of the argument that patient involvement and participation in care, as demonstrated by question asking, information exchange, and shared decision-making, is

16

ERIC S. WILLIAMS ET AL.

signiﬁcantly correlated with patient outcomes (Haywood, Marshall, & Fitzpatrick, 2006). This consideration is even more important given the expanded availability of medical information and the rise of medical consumerism. McGee and Cegala (1998) consider information exchange to be central to the patient–physician medical interview and a key component by which physicians make accurate diagnoses and effective treatment recommendations. They developed a study in which patients were trained in information-verifying skills for use during medical interviews. Findings revealed that trained patients asked signiﬁcantly more direct and indirect questions throughout the medical consultation and during the postexamination stage of the consultation than untrained patients; trained patients asked more information-verifying questions about medical topics than untrained patients; and trained patients had more accurate recall of information about treatment recommendations than untrained patients. Cegala, McClure, Marinelli, and Post (2000) found similar results when they examined the effectiveness of a training booklet designed to enhance patient communication skills in information exchange. Trained patients engaged in more information seeking, provided more detailed information about their medical conditions, employed more summarizing to verify information, and demonstrated a more patient-controlled style of communication. Overall, this body of research suggests that trained patients are more communicatively active during medical consultations. Haskard et al. (2008) extended the training paradigm to design a randomized study of a communication skills training program for both physicians and patients. Patient satisfaction and perceptions of choice, decision-making, information, and lifestyle counseling were measured one month post-intervention. Physician satisfaction and stress and global ratings of the communication process were assessed one month and six months post-intervention. Overall, physician training improved their informationgiving and lifestyle health-behavior counseling, and increased patients’ quality of care ratings and their willingness to recommend the physician. It also increased physician satisfaction with physical exams. The impact of training on physician satisfaction and stress were less clear; there was a relative increase in stress and decrease in satisfaction when only one, either physician or patient, was trained. Additional research examining training techniques is warranted. For example, current research supports Anderson and Sharpe’s (1991) view that modeling and practice are important components of effective communication skills training (Delvaux et al., 2005) while Cegala et al.’s (2000) work booklet may represent a viable alternative or complementary training model

Emotional Exhaustion, Depersonalization and Physician–Patient Communication 17

and tool. Formal training providing patients with skills on informationverifying and information recall strategies has also been successful (McGee & Cegala, 1998). Future research should examine the optimal method for delivery of communication skills instruction, with particular considerations for low literacy, cost-effectiveness, and ease of dissemination.

CONCLUSION Physician burnout is a real phenomenon with real consequences for patients. Understanding which physician communication behaviors impact patient satisfaction, trust, adherence, and ultimately patient health is essential for developing effective communication skills training for both physicians and their patients. Increased research focusing on the effect of physician burnout on communication during the medical encounter is needed, as are new models of communication skills training for both physicians and patients.

REFERENCES Ambady, N., & Rosenthal, R. (1992). Thin slices of expressive behavior as predictors of interpersonal consequences: A meta-analysis. Psychological Bulletin, 111, 256–274. Anderson, L. A., & Sharpe, P. A. (1991). Improving patient and provider communication: A synthesis and review of communication interventions. Patient Education Counseling, 17, 99–134. Bakker, A., Schaufeli, W., Sixma, H., Bosveld, W., & Dierendonck, D. (2000). Patient demands, lack of reciprocity, and burnout: A ﬁve-year longitudinal study among general practitioners. Journal of Organizational Behavior, 21, 425–441. Beach, M. C., Roter, D. L., Wang, N. Y., Duggan, P. S., & Cooper, L. A. (2006). Are physicians’ attitudes of respect accurately perceived by patients and associated with more positive communication behaviors? Patient Education and Counselling, 62(3), 347–354. Beck, R. S., Daughtridge, R., & Sloane, P. D. (2002). Physician-patient communication in the primary care ofﬁce: A systematic review. The Journal of the American Board of Family Practice, 15(1), 25–38. Bertakis, K., Roter, D., & Putnam, S. (1991). The relationship of physician medical interview style to patient satisfaction. The Journal of Family Practice, 32(2), 175–181. Blau, P. M. (1964). Exchange and power in social life. New York: Wiley. Brief, A. P. (1998). Attitudes in and around organizations. Thousand Oaks, CA: Sage. Buller, M. K., & Buller, D. B. (1987). Physicians’ communication style and patient satisfaction. Journal of health and Social Behavior, 28(4), 375–388. Campbell, K. S. (2008). Physicians and patients: How professionals build relationships through rapport management. In: G. F. Hayhoe & H. M. Grady (Eds), Connecting people with technology: Issues in professional communication (pp. 145–154). Amityville, NY: Baywood.

18

ERIC S. WILLIAMS ET AL.

Cegala, D., McClure, L., Marinelli, T., & Post, D. (2000). The effects of communication skills training on patients’ participation during medical interviews. Patient Education and Counseling, 41, 209–222. Cherniss, C. (1980). Professional burnout in human services organizations. New York: Praeger. Deckard, G. J., Hicks, L. L., & Hamory, B. H. (1992). The occurrence and distribution of burnout among infectious disease physicians. Journal of Infectious Diseases, 16, 224–228. Delvaux, N., Merckaert, I., Marchal, S., Livert, Y., Conradt, S., & Boniever, J. (2005). Physicians’ communication with a cancer patient and a relative: A randomized study assessing the efﬁcacy of consolidation workshops. Cancer, 103, 2397–2411. Dugan, J., Lauer, E., Bouquot, Z., Dutro, B. K., Smith, M., & Widmeyer, G. (1996). Stressful nurses: The effect on patient outcomes. Journal of Nursing Care Quality, 10(3), 46–58. Ekman, P. (1985). Telling lies. New York: Norton. Emanuel, E. J., & Emanuel, L. (1992). Four models of the physician-patient relationship. Journal of the American Medical Association, 267, 2221–2226. Fruedenberger, H. J. (1974). Staff burnout. Journal of Social Issues, 30, 159–164. Goffman, E. (1967). Interaction ritual: Essays on face-to-face behavior. New York: Partheon. Golembiewski, R., & Munzenrider, R. (1988). Phases of burnout: Developments in concepts and applications. New York: Praeger Publishers. Golin, C. E., DiMatteo, M. R., & Gelberg, L. (1996). The role of patient participation in the doctor visit: Implications for adherence to diabetic care. Diabetes Care, 19(10), 1153–1164. Halbesleben, J. R. (2006). Patient reciprocity and physician burnout: What do patients bring to the patient-physician relationship? Health Services Management Research, 19(4), 215–222. Halbesleben, J. R., & Bowler, W. M. (2007). Emotional exhaustion and job performance: The mediating role of motivation. Journal of Applied Psychology, 92(1), 93–106. Halbesleben, J. R., & Rathert, C. (2008). Linking physician burnout and patient outcomes: Exploring the dyadic relationship between physicians and patients. Health Care Management Review, 33(1), 29–39. Hall, J., Horgan, T., Stein, T., & Roter, D. L. (2002). Liking in the physician-patient relationship. Patient Education and Counseling, 48, 69–77. Hall, J., Roter, D., & Katz, N. (1988). Meta-analysis of correlates of provider behavior in medical encounters. Medical Care, 26(7), 657–675. Haskard, K. B., Williams, S. L., DiMatteo, M. R., Rosenthal, R., White, M. K., & Goldstein, M. G. (2008). Physician and patient communication training in primary care: Effects on participation and satisfaction. Health Psychology, 27(5), 513–522. Hatﬁeld, E., Cacioppo, J. T., & Rapson, R. L. (1992). Primitive emotional contagion. In: M. S. Clark (Ed.), Review of personality and social psychology: Volume 14, Emotion and social behavior. Newbury Park, CA: Sage. Hatﬁeld, E., Cacioppo, J. T., & Rapson, R. L. (1994). Emotional contagion. Cambridge, UK: Cambridge University Press. Haywood, K., Marshall, S., & Fitzpatrick, R. (2006). Patient participation in the consultation process: A structured review of intervention strategies. Patient Education and Counselling, 63(1–2), 12–23. Heszen-Klemmens, I., & Lapinska, E. (1984). Doctor-patient interaction, patients’ health behavior and effects of treatment. Social Science & Medicine, 1984(19), 1. Hickam, D., Severance, S., Feldstein, A., Ray, L., Gorman, P., Schuldheis, S., et al. (2003). The effect of health care working conditions on patient safety (No. 03-E031). Washington, DC: Agency for Healthcare Quality and Research.

Emotional Exhaustion, Depersonalization and Physician–Patient Communication 19

Hobfoll, S. E. (1989). Conservation of resources. A new attempt at conceptualizing stress. The American Psychologist, 44(3), 513–524. Hobfoll, S. E. (2001). The inﬂuence of culture, community, and the self in the stress process: Advancing conservation of resources theory. Applied Psychology an International Review, 50(3), 337–421. Hobfoll, S. E., & Freedy, J. (1993). Conservation of resources: A general stress theory applied to burnout. In: W. Schaufeli, C. Maslach & T. Marek (Eds), Professional burnout: Recent development in theory and research (pp. 115–133). Washington, DC: Taylor and Francis. Hochschild, A. R. (1983). The managed heart. Los Angeles, CA: University of California Press. Hsiao, C., & Boult, C. (2008). Effects of quality on outcomes in primary care: A review of the literature. American Journal of Medical Quality, 23(4), 302–310. Kaplan, S. H., Greenﬁeld, S., & Ware, J. E. (1989). Assessing the effects of physician-patient interactions on the outcomes of chronic disease. Medical Care, 27(3), S110–S127. Karasek, R., Baker, D., Marxer, F., Ahlbom, A., & Theorell, T. (1981). Job decision latitude, job demands and cardiovascular disease: A prospective study of Swedish men. American Journal of Public Health, 71(1), 694–705. Keijsers, G. J., Schaufeli, W. B., LeBlanc, P. M., Zwerts, C., & Miranda, D. R. (1995). Performance and burnout in intensive care units. Work and Stress, 9(4), 513–527. Kinmouth, A. L., Woodcock, A., Grifﬁn, S., Spiegal, N., & Campbell, M. J. (1998). Randomized controlled trial of patient centered care of diabetes in general practice: Impact on current wellbeing and future disease risk. British Medical Journal, 317, 1202–1208. Landsbergis, P. (1988). Occupational stress among health care workers: A test of the job demands-control model. Journal of Organizational Behavior, 9, 217–239. Lee, R., & Ashforth, B. (1993). A longitudinal study of burnout among supervisors and managers: Comparisons between the Leiter and Maslach (1988) and Golembiewski et al. (1986) models. Organizational Behavior & Human Decision Processes, 54, 369–398. Leiter, M. (1991). Coping patterns as predictors of burnout: The function of control and escapist coping patterns. Journal of Organizational Behavior, 12(2), 123–144. Leiter, M., & Maslach, C. (1988). The impact of interpersonal environment on burnout and organizational commitment. Journal of Organizational Behavior, 9, 297–308. Makoul, G. (1999). Report III. Contemporary issues in medicine: Communication in medicine. Washington, DC: Association of American Medical Colleges. Maslach, C., & Jackson, S. E. (1986). Maslach Burnout Inventory Manual. Palo Alto, CA: Consulting Psychologist Press. Maslach, C., & Leiter, M. (1997). The truth about burnout: How organizations cause personal stress and what to do about it. San Francisco: Jossey-Bass Publishers. Maslach, C., Schaufeli, W. B., & Leiter, M. P. (2001). Job burnout. Annual Review of Psychology, 52, 397–422. McGee, D., & Cegala, D. (1998). Patient communication skills training for improved communication competence in the primary care medical consultation. Journal of Applied Communication Research, 26, 412–430. Moore, J. E. (2000). Why is this happening? A causal attribution approach to work exhaustion consequences. Academy of Management Review, 25(2), 335–349. Pugh, S. D. (2001). Service with a smile: Emotional contagion in the service encounter. Academy of Management Journal, 44(5), 1018–1027.

20

ERIC S. WILLIAMS ET AL.

Quinn, R. E., Hildebrandt, H. W., Rogers, P. S., & Thompson, M. P. (1991). A competing values framework for analyzing presentation communication in management contexts. Journal of Business Communication, 28, 213–232. Ramirez, A. J., Graham, J., Richard, M. A., Cull, A., Gregory, W. M., Leaning, M. S., et al. (1995). Burnout and psychiatric disorder among cancer clinicians. British Journal of Cancer, 71, 1263–1269. Rao, J. K., Anderson, L. A., Inui, T., & Frankel, R. (2007). Communication interventions make a difference in conversations between physicians and patients. Medical Care, 45(4), 340–349. Ratanawongsa, N., Roter, D., Beach, M. C., Laird, S. L., Larson, S. M., Carson, K. A., et al. (2008). Physician burnout and patient-physician communication during primary care encounters.. Journal of General Internal Medicine, 23(10), 1581–1588. Roter, D., Hall, J., Kern, D., Barker, R., Cole, K., & Roca, R. (1995). Improving physicians’ interviewing skills and reducing patients’ emotional distress.. Archives of Internal Medicine, 155, 1877–1884. Roter, D., Stewart, M., Putnam, S., Lipkin, M., Stiles, W., & Inui, T. (1997). Communication patterns of primary care physicians. Journal of the American Medical Association, 277(4), 350–356. Schaufeli, W. B., & Bakker, A. B. (2003). Job demands, job resources, and their relationship with burnout and engagement: A multi-sample study. Journal of Organizational Behavior, 25, 293–315. Shanafelt, T., Bradley, K., Wipf, J., & Back, A. (2002). Burnout and self-reported patient care in an internal medicine residency program. Annals of Internal Medicine, 136, 358–367. Shirom, A., Nirel, N., & Vinokur, A. D. (2006). Overload, autonomy, and burnout as predictors of physicians’ quality of care. Journal of Occupational Health Psychology, 11(4), 328–342. Stewart, M. (1984). What is a successful doctor-patient interview: A study of interactions and outcomes. Social Science & Medicine, 19(2), 167–175. Stewart, M. (1995). Effective physician-patient communication and health outcomes: A review. Canadian Medical Association Journal, 152(9), 1423–1433.. Tamblyn, R., Abramhamowicz, M., Dauphinee, D., Wenghofer, E., Jacques, A., Klass, D., et al. (2007). Physician scores on a national clinical skills examination as predictors of complaints to medical regulatory authorities. Journal of the American Medical Association, 298(9), 993–1001. Trummer, U. F., Mueller, U. O., Nowak, P., Stidl, T., & Pelican, J. (2006). Does physicianpatient communication that aims at empowering patients improve clinical outcome: A case study. Patient Education and Counseling, 61, 299–306. Tversky, A., & Kahneman, D. (1974). Judgement under uncertainty: Heuristics and biases. Science, 185, 1124–1131. Whippen, D. A., & Canellos, G. P. (1991). Burnout syndrome in the practice of oncology: Results of a random survey of 1,000 oncologists. Journal of Clinical Oncology, 9(10), 1916–1920. Williams, E. S., Manwell, L., Konrad, T. R., & Linzer, M. (2007). The relationship of organizational culture, stress, satisfaction, and burnout with physician-reported error and suboptimal patient care: Results from the MEMO study. Health Care Management Review, 32(3), 203. Williams, E. S., Savage, G., & Linzer, M. (2006). A proposed physician-patient cycle model. Stress and Health, 20(2), 131–137.

USING SELF-CONCEPT THEORY TO IDENTIFY AND DEVELOP VOLUNTEER LEADER POTENTIAL IN HEALTHCARE$ Francine Schlosser, Deborah M. Zinni and Andrew Templer ABSTRACT Resource constraints in the Canadian publicly funded healthcare system have created a need for more volunteer leaders to effectively manage other volunteers. Self-concept theory has been conceptualized and applied within a volunteer context, and the views of healthcare stakeholders, such as volunteers, volunteer leaders, and supervisors, triangulated to form an understanding of the attitudes and behaviors of volunteer leaders. We propose that leaders are differentiated from others by how they view their roles in the organization and their ability to make a difference in these roles. This interpretation can be informed by self-concept theory because each individual’s notion of self-concept inﬂuences how employees see themselves, how they react to experiences, and how they allow these $

A previous version of this paper was presented at Academy of Management, Healthcare Interactive Papers, August 2008.

Biennial Review of Health Care Management: Meso Perspectives Advances in Health Care Management, Volume 8, 21–47 Copyright r 2009 by Emerald Group Publishing Limited All rights of reproduction in any form reserved ISSN: 1474-8231/doi:10.1108/S1474-8231(2009)0000008006

21

22

FRANCINE SCHLOSSER ET AL.

experiences to shape their motivation. A small case study proﬁles a volunteer leader self-concept that includes a proactive, learning-oriented attitude, capitalizing on signiﬁcant prior work experience to fulﬁll a sense of obligation to the institution and its patients, and demands a high level of respect from paid employees.

INTRODUCTION Most non-proﬁt organizations rely on volunteer support to reach their goals. The value of hospital volunteering has an estimated payoff of almost seven times the cost to manage volunteer labor (Handy & Srinivasan, 2004, p. 23). The importance of volunteering to charities extends beyond free labor, to building conﬁdence in that institution (Bowman, 2004). This image is important to the success of any healthcare institution, in fact more than 60 percent of 778 participants in a recent study attributed altered views of nursing facilities to volunteering (Keith, 2005). Despite such recognized beneﬁts, it has become more difﬁcult to retain and recruit volunteers. Volunteers have more demands placed upon their time, as they work longer hours at home and at work. Additionally, organizations face stricter requirements in the selection of appropriate volunteers, such as legal responsibilities regarding criminal background checks as well as privacy of information. These considerations limit the pool and expertise of volunteers. Budget cutbacks in government-funded healthcare have also resulted in fewer paid personnel who can coordinate volunteer work. A recent study of 28 hospitals using more than 2 million volunteer hours annually (Handy & Srinivasan, 2005) concluded that organizational demand for volunteer labor is a decreasing function of their costs. Thus, the willingness of hospitals to use volunteer labor diminishes as the cost to manage the volunteer force increases. These issues have increased the demand for volunteer leaders who can manage a team of volunteers. This demand is critical in healthcare, but is also felt across the non-proﬁt sector. The success of non-proﬁt organizations hinges upon the effective and coordinated behaviors of volunteer leaders. Leaders can be characterized by their strength of personality or character (Bowden, 1926) and in their activities inducing group change (Cooley, 1902; Mumford, 1906/1907). Leaders are also able to inﬂuence others toward a common goal (Stogdill, 1950). We propose that leaders are differentiated from others by how they

Identifying Volunteer Leader Potential in Healthcare

23

view their roles in the organization and their ability to make a difference in these roles. This interpretation can be informed by self-concept theory because each individual’s notion of self-concept inﬂuences how employees see themselves, how they react to experiences, and how they allow these experiences to shape their motivation. Organizations that rely heavily on volunteer leaders must better understand this relationship between leader self-concept and the volunteer environment. Self-concept contributes to how individuals are able to recognize their inherent strengths and weaknesses so that organizations are better able to meet the needs of their respective volunteers by reﬂecting this understanding in training and assignment of duties. Other researchers have conceptualized in general terms how the selfconcept of leaders and followers inﬂuences each other (Lord & Brown, 2001; Lord, Brown, & Freiberg, 1999). These general conceptualizations will have more value to organizational decision-makers if they are examined within speciﬁc organizational situations. We suggest that the burgeoning importance of volunteer leader development in the resource-constrained voluntary sector provides some impetus to consider how self-views might affect the leadership potential of ageing volunteers. Volunteers’ views of their roles are shaped by both individual- and relational-level processes. Accordingly, this chapter ﬁrst conceptualizes the attributes and behaviors that distinguish effective volunteer leaders and then applies this model using a case study methodology. In addition, we explore how a volunteer leader self-concept develops, thus informing the actions that institutions can take to encourage the emergence of new volunteer leaders.

Volunteering Vocation and Commitment As a broad deﬁnition, volunteerism is deﬁned as ‘‘any service to the community given without payment through a group or organization’’ (Warburton & Terry, 2000, p. 249). At a more general level, the deﬁnition of a volunteer is ‘‘an individual who is unpaid and gives freely of his or her time’’ (Brudney & Kellough, 2000; Hartenian, 2007; Wilson & Pimm, 1996). Wilson (2000) noted the importance of commitment to the volunteer role and the organization and concluded that volunteering must be viewed within the concept of self-identity, that is, how volunteers see themselves as people who help others. He described the three main reasons why volunteers withdraw from volunteering: (1) their efforts are unrecognized, (2) there is a poor ﬁt between their interests and volunteer roles, and (3) their efforts to

24

FRANCINE SCHLOSSER ET AL.

help clients/patients are limited by lack of autonomy. These reasons point to a need for fulﬁllment as a potential reason for volunteerism. Volunteers will feel fulﬁlled if their efforts are recognized and aligned with individual goals and if they are given autonomy. Volunteers’ identiﬁcation with the organization and their internalization of its goals are related to what volunteers are willing to contribute (Farmer & Fedor, 2001; Katz & Kahn, 1966/1978; Pearce, 1993). Previous research has explained that volunteers are driven by both altruism and egoism motives (Bowman, 2004). Keith (2003) concluded that differences in volunteer motivation and skills were explained by age, education, and prior volunteer experience. Esteem needs, such as acknowledgment, recognition, and self-actualization, motivate volunteers. The way that individuals perceive their personal interests and contributions may provide insights as to how to construct a volunteer self-concept. A volunteer’s self-view is framed by the fulﬁllment of their own goals and obligation to the organization’s goals. The following section describes self-concept according to previous theorists and is used to construct the conceptualization of volunteer and volunteer leader self-concept.

Constructing a Volunteer Leader Self-Concept In their review of past literature, Markus and Wurf (1986, p. 299) concluded that the self-concept could be deﬁned as ‘‘a set or collection of images, schemas, conceptions, prototypes, theories, goals, or tasks’’ held by an individual. Markus and Wurf (1986) highlighted the self-concept as a mediator and regulator of behavior and contended that ‘‘It interprets and organizes self-relevant actions and experiences; it has motivational consequences, providing the incentives, standards, plans, rules and scripts for behavior; and it adjusts in response to challenges from the social environment.’’ However, the non-speciﬁc nature of self-concept makes it difﬁcult to test empirically and creates a need to split the self-concept into different levels for practical ﬁeld analysis. Lord et al. (1999) suggest that ﬁeld researchers should distinguish stable and highly accessible schemas from other more obscure and variable schemas. The more accessible schemas determine which level of self-identity is salient at any given time. To clarify, volunteers have multiple deﬁnitions of self, divided into two levels of social inclusion (Brewer & Gardner, 1996; Lord et al., 1999; Markus & Wurf, 1986). The ﬁrst level of self focuses inward on the individual, with social inclusion limited to comparisons with others. The

Identifying Volunteer Leader Potential in Healthcare

25

second level expands social inclusion by focusing outward on how the individual relates to others.

The Individual Self At the individual level, a person deﬁnes the self as a separate unique entity. Individuals operating at this level develop feelings of self-worth by comparing their own traits to others. The individual self-identity is predicated on the idea of a person’s uniqueness and separation from others. Markus and Wurf (1986, p. 315) have suggested that, at the individual level, ‘‘the self-concept mediates intrapersonal processes, which include selfrelevant information processing, affect regulation and motivation processes.’’ First, self-relevant information involves a social comparison to others or different situations implying a judgment of personal capability (Bandura, 1997). In a study of aging female volunteer leaders who sought out planning and developing workshops to help them organize and manage their groups, Burden (2000) noted the signiﬁcance of a developmental perspective when theorizing volunteering. These women felt more capable after experiencing the workshop. Thus, discussion of the volunteer leader role involves comparing their accountabilities and developmental needs relative to others. Second, affect regulation describes feelings of personal control, through emotional self-control. Volunteers in a healthcare context must manage signiﬁcant stressors related to seriously ill and dying patients, grieving and concerned families, and stressed-out employees (Dein & Abbas, 2005). Active volunteers demonstrate more empathic concern (Bekkers, 2005) and altruism (Mowen & Sujan, 2005). A recent study of volunteers in an AIDS organization concluded that volunteers who showed empathic concern and perspective taking believed that their volunteer experiences had more value (Stolinski, Ryan, Hausmann, & Wernli, 2004). Consequently, how leaders view their abilities to deal with this stress relative to others will shape their willingness to continue to volunteer as a post-employment leisure activity. Finally, intrinsic motivation is related to the concept of role satisfaction because individuals will view themselves in comparison to other possible selves or other people. Previous researchers have concluded that volunteer involvement (total time volunteered) is predicted by satisfaction (Davis, Hall, & Meyer, 2003) and that motivation is strongly correlated to satisfaction (Reeser, Berg, Rhea, & Willick, 2005). Older volunteers experience greater increases in life satisfaction over time as a result of their

26

FRANCINE SCHLOSSER ET AL.

volunteer hours than their younger counterparts (Van Willigen, 2000) as well as improved physical and mental health (Lum & Lightfoot, 2005). Additionally, older volunteers who do not experience signiﬁcant roles in other life relationships (such as partner, employment, and marital) develop their psychological well-being through volunteering (Greenﬁeld & Marks, 2004). In a longitudinal study of 144 volunteers, Omoto, Snyder, and Martino (2000) found that younger volunteers tend to be motivated by and achieve outcomes related to interpersonal relationships. This is in contrast with older volunteers who are more likely to be motivated by service and obligation to the community. Consequently, when desiring to retain and develop older volunteers as leaders, organizations must understand volunteer attitudes and behaviors related to motivation and satisfaction.

The Interpersonal Self At a second level of self-concept, researchers have discussed the importance of the interpersonal (relational) self to the understanding of human relationships. For example, Brewer and Gardner (1996) deﬁned the interpersonal self within the context of relationships to others. At the relational level, the self develops through interaction or membership. For example, Lord et al. (1999) suggested that the relational self-concept will be inﬂuenced by a leader’s feedback. This relational self could also include the ‘‘reﬂected self’’ or the way that the individual perceives others to feel about him/her (Lord et al., 1999, p. 4). Similarly, in symbolic interactionist literature, Cooley (1902) developed the concept of the ‘‘looking glass self,’’ which reﬂects a picture of the self relative to others, speciﬁcally how individuals sense that others view their actions or roles. To summarize, both individual and relational levels of self-concept develop over time. The individual self-view derives through comparison with an alternate self, such as a past or future conception of one’s self or even a referent person. For example, volunteers may view their unpaid contribution in terms of the education and skills they developed in their previous paid careers. For example, a retired automotive executive may wish to utilize previous corporate leadership experiences in an unpaid volunteer assignment in a healthcare facility. In contrast, the relational self is deﬁned in terms of how others view the individual. Accordingly, volunteer leaders deﬁne their contribution through their interaction with other hospital stakeholders such as patients, paid employees, other volunteers, and institutions such as the union and the hospital management.

Identifying Volunteer Leader Potential in Healthcare

27

Distinguishing Leaders from Others Business researchers have deﬁned leadership as personality traits, actions, context, and relationships. For example, earlier research pertaining to leadership might be considered as a nexus of group change, activity and process (Cooley, 1902; Mumford, 1906/1907), as a strength of personality or character (Bingham, 1927; Bowden, 1926), and as a way to inﬂuence others toward a common goal (Stogdill, 1950). Bass (1990) relates this inﬂuence to the use of persuasion, power, and the creation of structure. We can build upon these ideas to conceptualize how volunteer leaders may be distinguished from other volunteers at an individual level and relational level. Fig. 1 depicts elements of individual and relational selfconcept relevant to leaders and other volunteers. Effective volunteers may view themselves in relation to their volunteering activities as both obligated and fulﬁlled. However, we propose that volunteer leaders act as a nexus of change and activity by being proactive in situations where other volunteers may choose a passive reaction. We suggest that leaders are aware of the personal sources of power, and this awareness will shape their conceptualization of self at the individual level. The relational self-concept is developed through the use of this power to inﬂuence other volunteers toward the institution’s goal. We explain this by applying a case study methodology.

Leader Individual Self-Concept

Other Volunteer Individual SelfConcept

Proactive Powerful

Reactive/Passive

Volunteer Self Concept Fulfilled Obligated

Leader Relational Self Concept Proactive Influential

Fig. 1.

Other Volunteer Relational SelfConcept Reactive/Passive

Volunteer Self-Concept.

28

FRANCINE SCHLOSSER ET AL.

METHODS A qualitative study was undertaken, utilizing a case study methodology, to better explore the different attitudes and behaviors of volunteer leaders and followers. Employing a case study methodology enables a researcher to closely examine the data within a speciﬁc context. In most cases, a case study method selects a small geographical area or a very limited number of individuals or organizations as the subjects of study. Case studies, in their true essence, explore and investigate contemporary real-life phenomenon through detailed contextual analysis of a limited number of events or conditions and their relationships. Yin (1984) deﬁnes the case study research method ‘‘as an empirical inquiry that investigates a contemporary phenomenon within its real-life context; when the boundaries between phenomenon and context are not clearly evident; and in which multiple sources of evidence are used.’’ The case study method has been used effectively in a number of ﬁelds including sociology and within government, management, and education. Therefore, a case study allowed us to select an organization that had many volunteers and that had already begun to develop a team structure with volunteer leaders.

Study Context A hospital context was chosen for this study because volunteer followers and leaders play a crucial role in delivering care to a broad section of society. Hospital volunteerism has been the focus of some previous research but has focused on motivation for volunteering (e.g., Liao-Troth, 2005) or recently the economic implications of volunteers in the resource-constrained health sector (Handy & Srinivasan, 2004, 2005). Data were collected from volunteer followers and leaders and employees in a large Canadian hospital. This hospital had played an important community role for 125 years. Founded in 1888 by an Order of Catholic nuns, within 20 years, the hospital had established a School of Nursing that later became part of local post-secondary institutions. In 1994, the hospital merged with two other religious-based hospitals and consolidated to two locations within the city. Although the hospital provides universal healthcare, the religious nature of this institution is worth noting because it may amplify the vocational expectations of employees, volunteers, and patients. The hospital has continued to experience funding cutbacks and ﬁnancial difﬁculties. At the time of the study, 500 individuals actively volunteered at the hospital,

Identifying Volunteer Leader Potential in Healthcare

29

coordinated by one paid staff member (the ‘‘volunteer coordinator’’) whose primary responsibility was to recruit, coordinate, and retain volunteers. Another important point is that this particular hospital is highly unionized, which has implications for those wishing to volunteer. For example, unionized workers may perceive that a volunteer is performing the work that should be given to a paid unionized worker. Sample The case samples of ‘‘leaders’’ and ‘‘other’’ volunteers were selected in consultation with the paid volunteer coordinator. As the sole person in the organization responsible for recruitment and development of volunteers, the perceptions of this coordinator reﬂected the expectations of the organization. The volunteer coordinator differentiated leaders from others by identifying volunteers who had decreased his/her personal management load by taking on additional responsibilities. With the help of the coordinator, we were also able to identify volunteers who exempliﬁed long-term, reliable volunteer commitment. Accordingly, we were able to assess participant attitudes and behaviors that shape these desired outcomes. Therefore, this sample is a purposeful sample. The value of this type of sample is that it can provide unique cases that are especially informative (Neuman, 2004). All participants identiﬁed by the paid volunteer coordinator were contacted and asked if they would like to participate. All those contacted agreed to meet with the researchers. Table 1 summarizes the demographics of the study participants. Most of the study participants were over 50 years of age and were collecting some form of retirement or disability pension. This aligns with previous research on volunteer leaders. For example, Edwards, Mooney, and Heald (2001) concluded that older community volunteers were more likely than student volunteers to assume responsibility for planning and coordinating services. Thus, the age of the participants in this study is representative of the general volunteer population. Data Collection Data were collected over two months in a series of focus groups and interviews. First, a two-hour-long focus group was conducted with volunteer leaders, followed by a second with ‘‘other’’ volunteers. Additionally, paid employees who supervised volunteers were interviewed for

30

FRANCINE SCHLOSSER ET AL.

Table 1. Qualitative Method

Healthcare Organization Participant Proﬁles. Participants

Demographics

Focus group

Six volunteers who were considered effective by volunteer co-ordinator

Focus group

Five team leader volunteers

Interview Group interview

Training coordinator (1) Gift shop manager and (2) temporary volunteer co-ordinator Volunteer co-ordinator Director of Mission, Pastoral and Volunteer Services Five supervisors of volunteers on various ﬂoors

Interview Interview

Interviews

Total

22 participants

Five females One male Three retired pensioners Two disability pensions Prior positions include teaching, administrative, auto hourly, and both disability pensioners were former patients Four females One male Five retired pensioners Three prior employees One 40-year volunteer One retired supervisor from auto industry Male Two females

Female Female

Relationship

Volunteers

Volunteers

Volunteer Employees

Employee Employee

Four females Employees One male Two administrators, one renal dialysis, eye surgery clinic, Priest Chaplain 17 Females 5 Males

9 Employees 13 Volunteers

30 minutes in person or by telephone. All interviews were tape-recorded with the permission of the interviewees. The questionnaires are available for perusal from the lead researcher. Data Analysis The tape-recorded interviews were content analyzed with the assistance of the qualitative software program NVivo. NVivo is a powerful program for

Identifying Volunteer Leader Potential in Healthcare

31

coding and interpreting textual data. The initial coding was performed by a research assistant and was audited by the author. The narratives were coded using standard qualitative analysis techniques (Creswell, 1998). The minor discrepancies that existed between the coders were resolved by the coders together examining the data. The cases were initially coded at the sentence level with each substantive sentence assigned to one or more of various themes.

DEVELOPMENT OF THEMES AND CONSTRUCTION OF A VOLUNTEER LEADER SELF-CONCEPT The data were analyzed to identify attitudes and behaviors that distinguished volunteer leaders from other volunteers. Additionally, we investigated the relationship between paid supervisors and volunteers and how each assessed volunteer effectiveness and satisfaction. Review of verbal and non-verbal communication patterns in the focus group discussions revealed common themes related to both the individual self and the relational self. This process allowed us to understand volunteer leadership using a framework of self-concept. Leaders and Other Volunteers This section highlights individual and relational self-concept themes found in the data, using quotes to proﬁle the differences in attitude and behavior between volunteer leaders and other volunteers. Table 2 proﬁles differences that are related to each aspect of the individual and interpersonal self and identiﬁes whether connections exist to proactive and passive behaviors and to the use of power and inﬂuence. Based on the content analysis and previous literature, we sorted the data into the individual-level themes of self-relevant information processing, affect regulation, and motivation, noted by Markus and Wurf (1986). This approach highlighted differences in how the volunteer leader and follower focus groups expressed thoughts and experiences speciﬁcally in the areas of training and development, coping strategies, and their previous work experiences. Differences were noted in individual themes related to duties and accountabilities relative to other volunteers and paid employees. Other volunteers noted their role was to willingly cooperate and be part of a team.

32

FRANCINE SCHLOSSER ET AL.

Table 2.

Comparison of Volunteer Leaders and Other Volunteers.

Theme

Individual level Self-relevant information processing (training and development) Affect regulation (coping strategies)

Motivation (previous experience) Relational level Self reﬂected through union Self reﬂected through paid employees Self reﬂected through patients Self reﬂected through other volunteers Self reﬂected through institution

Volunteer Leader

Other Volunteers

Proactive

Passive/reactive

Actively seek out training to meet needs of hospital

Fit current skills into needs of hospital

Personal development as a coping strategy for emotionally challenging roles, seeking out leadership roles or training Previous jobs shaped personal power in expanding volunteering role

Important of ﬁt between person and emotional demands of role

Previous jobs provided skills that allowed better ﬁt

Active conﬂict resolution Passive conﬂict response Complaining: Unappreciated Satisﬁcing: Appreciated by by employees employees Feedback and interaction (gratitude) Communication of important Participation in social job-related information gatherings No inﬂuence tactics No inﬂuence tactics Previous experiences with the institution shaped a view of volunteering as a way to give back

In contrast, volunteer leaders were willing to take on managing tasks (such as scheduling and training) and displayed leadership attributes (such as willingness to stand up for the rights of volunteers). Volunteer leaders viewed their duties differently from other volunteers, noting the importance of training and scheduling tasks and heightened unit responsibilities. A volunteer leader noted as follows: We need at least 4 people a day, and if someone can’t make it, then I have to call someone else or I will come in myself.

This same sentiment was echoed by another volunteer leader who noted, ‘‘If somebody does not show up or I cannot ﬁnd someone else then I am it. Sometimes it is three days a week. We have some new volunteers but they are not all very responsible.’’

Identifying Volunteer Leader Potential in Healthcare

33

As previously discussed, leadership has been deﬁned as a nexus of group change, activity and process, strength of character, and a way to inﬂuence others toward a common goal by being proactive. Although agreeing on a lesser accountability than paid work, volunteer leaders in the study more often deﬁned their role based on what needed to get done, whereas other volunteers were more likely to view it as secondary to paid employees. By demonstrating the willingness to get the job done, leaders were inﬂuencing others toward a common goal of exemplary ‘‘patient care.’’ For example, a leader noted as follows: If a nurse asks you to do something you might do it, but if the nurse asks a porter, then they may say that it is not my job, but I as a volunteer would never say it is not my job.

In addition to setting an example of completing whatever task is presented, these leaders are also demonstrating extra-role behavior. In addition to patient care, volunteer leaders sometimes found themselves in other peripheral activities of the hospital such as in the lost and found, records, or even the coffee and gift shops. Although they are not core, peripheral activities are often staffed with volunteers and are important to the effective functioning of the organization. As one volunteer noted of a volunteer leader: She is also in charge of the lost and found y It is a big job. If it sits around for a long time and it is not clean she will take it home and wash it and clean it and mend it and then she will take the clothes over to our church that is in a poor area and then they sort the clothes into different groups and gives them to needy families.

The differences in the way that leaders and other volunteers deﬁned their volunteer roles were also reﬂected in how they described situations in the hospital that involved effectiveness and compassion. Both described effectiveness using terms such as compassion, self-less, social conscious, thoughtful, and providing a hug or a hand to hold. Additionally, they described their own effectiveness and compassion relative to the effectiveness of paid employees. The focus group of volunteer leaders voiced high expectations of paid employees. For example, a volunteer leader described a time when she was waiting with family members for 11 hours, during a patient’s serious, day-long surgery. As the volunteer’s shift was about to end, the volunteer leader asked a nurse to call up to the operating room to ﬁnd out how much longer the surgery would last. She noted as follows: I get chills telling you that they [the nurses] wouldn’t do it for me y but I had to leave and I could hardly go back and tell the patient’s family that I’m leaving now. They [the

34

FRANCINE SCHLOSSER ET AL. family] said, ‘‘what about my mother?’’ Why wouldn’t they [the nurses] have feelings for that family?

In contrast, other volunteers appeared to have less expectation that the effectiveness of paid staff would be linked to compassion, Like if you’re in a room with patients, and you sense that they need to talk or have tears in their eyes, you sit down and ask them what’s wrong, you have that time whereas a nurse may not.

Differences in the way that volunteer leaders and other volunteers regulated their emotions were linked to personal development and decisionmaking. A volunteer leader who performed a task that was perceived to be a nurse’s responsibility remedied her lack of experience by taking a course for helping people cope with dying. Other volunteer leaders also held leadership executive roles in the volunteer association. In contrast, non-leader volunteers stressed the importance of ‘‘ﬁt’’ with their volunteering role, rather than personal development. As one volunteer indicated, ‘‘I think if you match the person’s attributes to the job, that’s the best, they’ll be good at it. It will be a trial and error process, depending on the person’s personality.’’ Matching an individual’s skills to a particular job is certainly advantageous to the hospital; however, volunteer effectiveness in this instance would be about an individual going where they are needed, as opposed to where they may feel they are better suited – an attribute that volunteer leaders hold as important to effective volunteerism. In yet another example, a volunteer indicates that they want to be told what to do, rather than make decisions, ‘‘Here it’s kind of nice because someone can just tell you what they’d like you to do and you just do it and you’re not in the planning.’’ Previous experience was viewed as important to being an effective volunteer by leaders, other volunteers, and paid employees. However, in leaders, this referent self played a greater role in shaping their sense of personal power, leading to a lengthy discussion in the volunteer leader focus group. During this discussion, a leader noted as follows: Going back to ‘‘We are just a volunteer’’, I think staff has to realize that before we came to volunteer we had a position in the community. B – was an engineer, a highly respected position. I get the impression [from some employees] that we are just somebody off the street volunteering. I think I sort of shook a few minds because I was helping a patient up and one of the nurses said it was not my job and then C – who is the manager down there who was one of my students said that ‘‘D – has been a nurse here for so many years, she knows what she is doing.’’

Identifying Volunteer Leader Potential in Healthcare

35

Other volunteers ﬁt their prior skill sets to the tasks given. They were not as comfortable in going beyond their assigned volunteer duties even if they were well qualiﬁed to do so. For example, Your past skills count for a lot y Because I’m a skilled tradesperson, they’re [paid employees] very sticky about what your job is and who does what y Going out of your way is okay, but not if you’re offending people.

The Volunteer Leader Relational Self The data proﬁled in Table 2 also included themes that informed our understanding of the relational self-concept of volunteers and volunteer leaders. In this table, we note how volunteer leaders and other volunteers differ in how they see themselves reﬂected by other hospital stakeholders. Relationships with unionized personnel (e.g., porters), paid employees (e.g., nurses), patients, volunteer peers, and the institution shaped the volunteer leader and follower self-perception. Throughout our focus group meetings and discussions with various participants, union sensitivities were consistently discussed. Both unionized and non-unionized employees are wary of the activities of the volunteers, concerned that their jobs may be in jeopardy. The thinking is that if a volunteer does the porter job, then there would be no need to employ the porter. The same sentiment does not hold true for jobs requiring more skill, such as for nurses. As one participant noted, ‘‘The higher the skill set, those people are not insecure they appreciate us. The lower the skill set they are insecure.’’ What this suggests is that leaders must be politically astute in how to work around porters or jobs where workers may feel insecure toward the volunteers. In essence, they will do what is necessary to get the job done but not by offending anyone. For example, in their relationship with the union, volunteer leaders were more likely to feel that they should take a leadership role to resolve conﬂict. Quoting a leader, When a doctor gave me something to do they went to management. And another thing I should have done y they said to me, do you want to meet with the union leader and I said, look, I’m just a volunteer and the volunteer services lady went to the meeting, but I think I was wrong. I think I should have went to the meeting. Now when I look back to what had happened, because the person sent was also in the union, I should have gone y. I would now [speak for and defend her team] but I thought that being a volunteer y I didn’t think I should have to go into that kind of thing y I am not here to argue. I wanted to let someone else handle it.

36

FRANCINE SCHLOSSER ET AL.

Furthermore, several other volunteers noted conﬂicting situations, showing how sensitive they needed to be. As one participant notes as follows: There’s one thing you have to remember about volunteering, and this we can’t take the place of a paid worker y and y not infringe on union conditions.

Yet another participant states as follows: I make sure I say it y I am M – the volunteer. I went through a lot of difﬁculties with upsetting the porters. It is a shame that I have to watch what I do with the porters around.

Therefore, even though leaders and volunteers were sensitive to certain unionized employees’ insecurities, they felt that the needs of the patient must be fully considered. Therefore, in instances related to employees such as porters, volunteer leaders would make sure these employees were not around before engaging in activities directly associated with their duties, not as a way to undermine the employees’ responsibilities but for the sake of patient care. For example, one leader told us as follows: When the porters leave, then I tell the girls that they can go over and pick up the patients, but they don’t do it while they [the porters] are there because they are threatened and it is really a shame. Because there are a lot of other things they could be doing, but we still do it. You have to play your cards right.

Regarding their relationship with paid employees, participants in the volunteer leader focus group discussed their feelings that paid employees did not respect volunteers, for example as follows: When I started volunteering 28 years ago, the staff did not appreciate having volunteers around they thought it was taking work away from them and some of them in the elevators would ask why would you want to work for free? But now they do appreciate it.

In contrast, other volunteers discussed their beliefs that paid employees valued and respected their volunteering efforts, ‘‘I heard the nurses tell the patients that they don’t know what they would do without the volunteers.’’ Their relationships with patients most strongly deﬁned the volunteering motivations and vocational identity of volunteer leaders and other volunteers. A volunteer leader noted as follows: It helped me to be in the recovery room and the O.R y that I am not the only one going through different things. I could help people and I could relate to what they were going through. It really is nice to volunteer in the recovery room because the families really do appreciate it.

Identifying Volunteer Leader Potential in Healthcare

37

A volunteer (non-leader) noted as follows: I have a child care degree, a bit of university and a college degree. I started out as a mental patient myself, and now I feel I have the empathy for these people.

Both volunteer leaders and other volunteers described patient satisfaction as the most important measure of their effectiveness as volunteers. Feedback and interaction with patients developed a positive volunteer self-view, apparent in the following quotes: Volunteer Leader: When a patient says thank you it makes your day. We don’t need someone from the ﬁrst ﬂoor to give you a pin or a pat on your back. It is self satisfaction. Volunteer (non-leader): By the thanks you get. Just when they say ‘‘you’re an angel.’’ Because it’s their comments that show how effective you are.

The self-concepts of volunteer leaders and non-leaders were also shaped through relationships with other volunteers, forming a culture amongst the volunteers. For example, a feeling of belonging related to volunteer uniforms was evident in this leader’s quote: We have new shirts and I have had many comments from people who are not appreciative of them switching the color from pink to blue. They say ‘‘we have known you as the pink ladies for decades and now all of a sudden we are the blue ladies. We walked in looking for the pink ladies.’’

Leaders discussed the importance of meeting to communicate important issues such as security during the SARS outbreak, whereas other volunteers were more likely to describe relationships with other volunteers in terms of gatherings and lunches. The leaders were assuming their leadership duties, taking care of everyone to ensure continued patient care, where volunteers focused on the social aspect associated with their volunteer role. The last theme regarding relational self involved their relationship with the institution. Both volunteer leader and non-leader participant relationships with this hospital originated through previous experiences as patients, employees, or community members who were then strengthened by current volunteering experiences. For example, different leader participants noted that Participant 1: older family members were served very well here so I thought it would be a good place to give back. Participant 2: I was more familiar with this institution. As a child if we were ever hurt, this is the hospital that we would go to, or when my parents had surgeries this is where they were.

38

FRANCINE SCHLOSSER ET AL. Participant 3: You walk in and everyone sincerely smiles and says hi.

However, through discussion and otherwise, all the volunteers and leaders were aware of the value they contributed to the bottom line of the hospital, noting that there were huge cost savings with their volunteering. This denotes that the volunteers understood their role in the context of the overall organizational effectiveness and took responsibility for their role in patient care and comfort.

Leader Volunteer Inﬂuence Behaviors The Volunteer Leader focus group had a lengthy discussion about their informal role modeling efforts to inﬂuence people outside the hospital to become volunteers. Examples include the following: Participant 1 – Well one nurse one day in the elevator said she was ready to retire and she said I think I am going to come back and volunteer because of you. She said to me ‘‘I am paid so much an hour and you ladies are not paid anything and there you are smiling and there we are with the long face y’’ Participant 2 – No people leave and you don’t hear from them. I go to the retirement luncheon twice a year. I have tried to get people to come back and they are just not willing to come back and I think that is a great sadness.

However, the focus groups did not highlight typical relationship themes between leaders and other volunteer team members. There were no quotes that exempliﬁed vertical tactics (downward from the volunteer leaders nor upward tactics from the other volunteers) nor horizontal inﬂuence tactics. For example, leaders highlighted the frustrations of working with high school–aged volunteers (40 volunteer hours were mandated in the local high school system). These student volunteers were not perceived to have the same level of fulﬁllment and obligation to their volunteering role. References were repeatedly made to the lack of commitment from these students as well as a lack of maturity. However, neither the volunteer leaders nor the other longer term volunteers in the focus groups described any attempt to inﬂuence the student volunteers toward more effective behaviors and goals while admitting that younger people could make a difference in brightening up the lives of some of the patients with their warm smiles. Furthermore, the comments showed a line drawn in the sand between the older and the younger volunteers. As mentioned earlier, there is a bifurcation in the volunteer composition – they are either younger students or aging volunteers, with few in-between. This bifurcation of volunteers is

Identifying Volunteer Leader Potential in Healthcare

39

quite disturbing as it reﬂected potential issues in developing a sustainable chain of strong and effective volunteers. A few leaders described requests by the other volunteers for the leaders to use their leadership role and power to intervene in group conﬂicts. In spite of this, the leaders did not ﬁnd an opportunity to do so, because the volunteer scheduling made it difﬁcult to meet as a team. Also, although the hospital paid staff had identiﬁed the participants in the ﬁrst focus group as effective volunteer leaders, and the second focus group as effective volunteers, almost all of the second group was unaware of the presence of team leaders. Further investigation into this phenomenon indicated that volunteer leaders were relegated to a few areas of the institution and that their role was based on long-term tenure and their particular knowledge in certain areas. Their main duty entailed the scheduling of volunteers, thus taking the onus away from the understaffed volunteer coordinator’s ofﬁce, who were in need of assistance. Essentially, these leaders had no say in the type of volunteers who worked in their respective departments, nor did they play in role in the performance management of the volunteers. They did however liaise with the departments on any concerns that may arise between volunteers, workers, and patients; however, they do not have the power to resolve organizational issues, forwarding them to the volunteer coordinator. This further indicates the lack of inﬂuence exerted by those in volunteer team leader roles. This is particularly concerning because the lack of formal structure in the volunteer hierarchy can lead to confusion and undermine the effective functioning of volunteer leaders.

DISCUSSION This case study examined the differences in attitudes and behaviors linked to the self-concepts of ‘‘effective’’ volunteer leaders and other volunteers. To summarize (based on focus group consensus), they differed at an individual level of self-concept by (a) their duties as volunteer leader or other volunteer, demonstrated by leaders having heightened responsibility for covering volunteer needs; (b) their active or passive views of volunteer activities relative to paid responsibilities; (c) their conﬂict coping techniques, speciﬁcally with leaders choosing more learning oriented techniques; (d) their sense of personal power and efﬁcacy related to prior work experiences. At the relational level, leaders and followers differed in (a) how their self-efﬁcacy was reﬂected in employee appreciation or lack of appreciation and (b) how their volunteer or volunteer leader role shaped a

40

FRANCINE SCHLOSSER ET AL.

sense of self reﬂected in how other volunteers interacted with them (socially versus task oriented). Volunteer leaders and followers were similar in their identiﬁcation with other volunteers and the value they placed on serving patient needs. They were also likely to have a positive pre-volunteering experience with the hospital. Attitudes expressed in both focus groups reﬂected a knowledge of institutional norms and familiarity with the job content, gained through long-term experience with the hospital. This long-term experience encompassed both pre-volunteer and volunteer encounters. Followers and leaders placed importance on working within the system to increase patient care. Much discussion centered on how volunteers were required to balance patient and employee needs for volunteer help with other structured roles in the organization. For example, limits were placed on their roles by either a lack of professional expertise (nursing was done by registered nurses) or collective agreements. Similar to Keith (2003), the behaviors and attitudes of most effective leader volunteers interviewed in the current study reﬂected skills learned as employees and speciﬁcally knowledge of the organizational culture and routines of the hospital. One participant had not worked at the hospital but had worked in a unionized shop, and this helped him to understand the union–volunteer relationship. Women who had been stay-at-home moms, could not as easily understand this relationship, opting to ensure the care of the patients was never compromised. Additionally, more than half of the volunteers and volunteer leaders were currently or had previously volunteered with other organizations, and this demonstrated their commitment to volunteering, and potential competition for their volunteer involvement. Study participants, from both leader and follower volunteer perspectives, experienced a strong relational attachment to the hospital. This attachment was formed from prior experiences with the hospital, for example, as employees, patients, or relatives of patients. This indicates that they experienced a strong relational social exchange before joining the institution on a volunteer basis. However, there was also substantial discussion about the management problems at the hospital and how other retired employees refused to consider post-employment volunteering. This suggests that prevolunteering experiences with the institution and speciﬁcally the type of relationship that emerges may inﬂuence the development of volunteer organizational commitment and the decision to volunteer. Leaders and other volunteers unanimously perceived volunteering at this organization to require a deep commitment to the patients. For example, they discounted the effectiveness of students who volunteered to meet

Identifying Volunteer Leader Potential in Healthcare

41

community service hours (required by high schools and some government programs). Paid employees who supervised volunteers noted that they assessed volunteer effectiveness by asking other supervisors and workers for assessment or by quantifying the value of their unpaid work. In contrast, volunteers and leaders measured their own effectiveness by patient satisfaction. Consequently, a social exchange also developed between volunteers and their clients – patients and relatives. This case study ﬂagged problems with the volunteer leaders. Without the use of direct inﬂuence on other volunteers, they were unable to direct all volunteers on a common path toward the same goal and commitment. The hospital’s volunteer retention and attraction issues and the aging volunteer base may stem from the lack of inﬂuence tactics used by volunteer leaders with new student volunteers and with the other longer term volunteers.

Using Self-Concept Theory to Identify and Develop Volunteer Leader Potential This case holds implications for hospitals and other large formalized institutions in the identiﬁcation and management of volunteer leaders. Volunteer leaders have heightened expectations of effectiveness regarding both paid and unpaid workers. Thus, there are indications that volunteer leaders are sensitive to being a part of a team with high standards. Leaders believed that compassion was a part of everyone’s role, and not just an extra duty delegated to volunteers because paid staff did not have time. Volunteers were more likely than paid employees in this study to describe effectiveness and compassion as being central to patient care as well as patient care being the ultimate measure of effectiveness. This study implies that institutions who encourage paid employees to delegate the role of compassion and hand-holder to volunteers may risk alienating both the experienced leader volunteer and the patients themselves who expect ‘‘care’’ to be exempliﬁed in paid healthcare professionals. However, in the face of diminishing funds, and shortage of skilled healthcare employees, volunteers are increasingly assisting with work previously accomplished by paid employees. For example, nurses have less time to spend on compassionate tasks associated with healthcare delivery. Therefore, whatever role the volunteer takes on should be based on the peripheral aspects of other’s job(s) and not on the main tasks. This goal can only be achieved by instituting clearly deﬁned job descriptions for both the volunteer leader and their respective volunteers. However, consideration has to be made to keep

42

FRANCINE SCHLOSSER ET AL.

these job descriptions as generic as possible to ensure that there is ﬂexibility for the volunteer leaders to make appropriate judgment calls as the task demands and to keep the task of continued upgrading of job descriptions to a minimum to keep costs in-line with budgetary constraints. The study also indicates that institutions should consider the previous job experiences of volunteers when choosing leaders. While it was clear that the leaders we spoke to were ideally suited to their respective areas, there is an increased need to utilize more leaders in the hospital. As such, an internal posting system should be developed to encourage the growth in this area. It was evident that a number of volunteers would not want such a role, but there were others who had come from high-level positions in their previous working lives who might value this leadership role and who probably bring a lot of value-added because of their prior experience. However, an increase in the amount of volunteers indicates that better recruitment and selection of leaders is warranted. In a workplace setting, human resource managers commonly use personality testing (Barrick & Mount, 1993; Le, Oh, Shaffer, & Schmidt, 2007) and structured job interviews (Olmstead, 2007) to select paid workers. The same concepts can be utilized to ensure person-job ﬁt for unpaid workers. One area of concern was the misuse of student volunteers. Since it is difﬁcult to attract a large pool of volunteers, students offer an available pool of recruits. Unfortunately, as indicated in our interviews, they are not seen as effective as others demonstrating a lack of commitment, no loyalty, and a ‘‘get me out of here as soon as possible’’ attitude that does not meet with the attitudes of effective volunteers. Special assignments should be identiﬁed that better ﬁt the short-term and inexperienced nature of student volunteering. With volunteer leaders having appropriate authority, they can discipline students who do not wish to take on the required roles and behaviors needed. Without successful endorsement on completion of their volunteer assignment, volunteer leaders can choose to either pass or fail the student. A failure would result in the student not achieving their volunteerism hours and having to repeat some aspect of it. Alternatively, if the student is not fulﬁlling the needs of the requirement, they could be dropped from the volunteerism program. Therefore, a program that awards points of merit, fulﬁlling the required hours, and fulﬁlling identiﬁed action items could be a way to ensure students are more effective volunteers. Assigning student volunteer leaders may also be effective to encourage students to take on leadership roles, which are useful for them when they apply to other jobs and may encourage them to be better volunteers.

Identifying Volunteer Leader Potential in Healthcare

43

More notably, and concerning, is that as the healthcare industry continues to experience shortages in paid workers, unpaid workers now become critical to healthcare delivery (American Society of Training and Development, 2007). Therefore, it is of paramount importance that appropriate human resource and employee development programs, once relegated to full-time paid employees, also be part of mainstream training for unpaid workers. They must also support the development needs and credibility of volunteer leaders in dealings with paid employees, perhaps through social events, and joint training and development opportunities. Training programs will need to be developed to facilitate the transitional role for those becoming leaders or for those who increase their knowledge and expertise in leadership. Modeling effective leader behavior should be incorporated as part of the training. Self-learning programs can also be incorporated to alleviate the time demands of volunteers that may not be available. Finally, the study highlights the limitations associated with appointing volunteer team leaders to just manage administrative tasks. These volunteer leaders have demonstrated proactive behaviors and personal power bases that could be used to inﬂuence other volunteers, at both mature and junior stages of volunteering. They may demonstrate managerial talents but still require leadership coaching. As such, mentoring programs, coupled with training and feedback, should be instituted. One ﬁnal aspect is a reward system. Through discussions with paid staff, we learned that formal rewards consisted primarily of pins presented at an annual volunteer event, paid lunches, and parking. We consistently heard from volunteer leaders that the accolades they received were enough to keep them motivated and satisﬁed. They made comparisons to other institutions that did not do these things. Therefore, it is important that the hospital continue its stated traditions, but also consider other programs such as volunteer leader of the month selected amongst all other volunteer leaders in the hospital, as well as for volunteer of the month in each of their respective departments. As well, senior management needs to have a presence and conduct department visits on a consistent basis, thanking the volunteers whenever they see the expected behavior.

Limitations and Implications for Future Research This case analysis was conducted at only one healthcare institution, and our results are most appropriately generalized to other institutions within a similar context, speciﬁcally those with a large and formalized group of

44

FRANCINE SCHLOSSER ET AL.

volunteers and a unionized workforce. Both of these conditions might inﬂuence the scope and content of volunteer duties and the need for volunteer leaders. Measurement of effectiveness in other volunteering contexts may pose challenges for future research. The exploratory study identiﬁed attitudinal differences between paid staff and volunteers. Paid staff were more likely to assess volunteer effectiveness based on monetary contribution or supervisory evaluation. A similar perspective has been used in previous research, for example, Farmer and Fedor (2001) measured effectiveness by how much money and time were given by the volunteer and using overall assessments provided by the volunteer administrator. These measures are limited because they do not assess the level of client satisfaction. Indeed, volunteers in the exploratory study noted that their effectiveness should be measured by customer appreciation. The qualitative nature of this study is inherently subjective. Qualitative analysis always entails making choices on explanatory models and options, and alternative explanations might apply. However, we have used our best interpretation of the relevant literature to choose the models and frameworks that will give this chapter a strong conceptual anchor. Future research should build upon the ﬁndings of this qualitative research to investigate methods that might be used to measure the effectiveness of volunteer leaders and other volunteers across North America. It would be important to understand the differences between a publicly funded healthcare system and one that is not. Furthermore, the notion of working in a unionized setting presents many constraints on a volunteer’s ability to do their work, and therefore, it would be advantageous to understand how volunteer leaders and other volunteers function in non-unionized settings. Additionally, it would be important to understand the motivations of younger volunteers, who may engage in this activity beyond just educational credit requirements. This would help to create a sustainable stable of volunteers.

CONCLUSION To conclude, this research has examined how volunteer leaders and other volunteers differ in their volunteer self-concept. Self-concept theory has been applied within a volunteer context, and the views of different stakeholders triangulated to form an understanding of the attitudes and behaviors of volunteer leaders.

Identifying Volunteer Leader Potential in Healthcare

45

In summary, our study proﬁles a volunteer leader self-concept that includes a proactive, learning-oriented attitude, capitalizing on signiﬁcant prior work experience to fulﬁll a sense of obligation to the institution and its patients, and demands a high level of respect from paid employees. Without effective volunteers, healthcare institutions would not successfully fulﬁll their societal obligations to high-quality healthcare. Therefore, volunteerism, with volunteer leaders, is an important program requiring further research to help these organizations reach their stated goals.

REFERENCES American Society of Training and Development. (2007). Healthcare searches for retention cure. TþD, 61(6), 16. Bandura, A. (1997). Self-efﬁcacy: The exercise of control. New York: W.H. Freeman. Barrick, M. R., & Mount, M. K. (1993). Autonomy as a moderator of the relationships between the big ﬁve personality dimensions and job performance. Journal of Applied Psychology, 78, 111–118. Bass, B. M. (1990). Bass & Stogdill’s handbook of leadership. New York: The Free Press. Bekkers, R. (2005). Participation in voluntary associations: Relations with resources, personality, and political values. Political Psychology, 26(3), 439–454. Bingham, W. (1927). Leadership. In: H. Metcalf (Ed.), The psychological foundations of management. New York: Shaw. Bowden, A. (1926). A study of the personality of student leaders in the United States. Journal of Abnormal and Social Psychology (47), 534–539. Bowman, W. (2004). Conﬁdence in charitable institutions and volunteering. Nonproﬁt and Voluntary Sector Quarterly, 33(2), 247–270. Brewer, M. B., & Gardner, W. L. (1996). Who is this ‘‘We’’? Levels of collective identity and self-representations. Journal of Personality and Social Psychology, 71, 83–93. Brudney, J. L., & Kellough, J. E. (2000). Volunteers in state government: Involvement, management and beneﬁts. Nonproﬁt and Voluntary Sector Quarterly, 29(10), 111–130. Burden, J. (2000). Community building, volunteering and action research. Loisir & SocieteSociety and Leisure, 23(2), 353–370. Cooley, C. (1902). Human nature and the social order. New York: Scribners. Creswell, J. W. (1998). Qualitative inquiry and research design: Choosing among ﬁve traditions. Thousand Oaks, CA: Sage. Davis, M. H., Hall, J. A., & Meyer, M. (2003). The ﬁrst year: Inﬂuences on the satisfaction, involvement, and persistence of new community volunteers. Personality and Social Psychology Bulletin, 29(2), 248–260. Dein, S., & Abbas, S. Q. (2005). The stresses of volunteering in a hospice: A qualitative study. Palliative Medicine, 19(1), 58–64. Edwards, B., Mooney, L., & Heald, C. (2001). Who is being served? The impact of student volunteering on local community organizations. Nonproﬁt and Voluntary Sector Quarterly, 30(3), 444–461.

46

FRANCINE SCHLOSSER ET AL.

Farmer, S. M., & Fedor, D. B. (2001). Changing the focus on volunteering: An investigation of volunteers’ multiple contributions to a charitable organization. Journal of Management, 27(2), 191–211. Greenﬁeld, E. A., & Marks, N. F. (2004). Formal volunteering as a protective factor for older adults’ psychological well-being. The Journals of Gerontology, 59B(5), S258–S264. Handy, F., & Srinivasan, N. (2004). Valuing volunteers: An economic evaluation of the net beneﬁts of hospital volunteers. Nonproﬁt and Voluntary Sector Quarterly, 33(1), 28–54. Handy, F., & Srinivasan, N. (2005). The demand for volunteer labor: A study of hospital workers. Nonproﬁt and Voluntary Sector Quarterly, 34(4), 491–501. Hartenian, L. S. (2007). Nonproﬁt agency dependence on direct service and indirect support volunteers: An empirical investigation. Nonproﬁt Management and Leadership, 17(3), 319–334. Katz, D., & Kahn, R. (1966/1978). The social psychology of organizations. New York: Wiley. Keith, P. M. (2003). Interests and skills of volunteers in an ombudsman program: Opportunities for participation. International Journal of Aging & Human Development, 57(1), 1–20. Keith, P. M. (2005). Correlates of change in perceptions of nursing facilities among volunteers. Journal of Applied Gerontology, 24(2), 125–141. Le, H., Oh, I.-S., Shaffer, J., & Schmidt, F. (2007). Implications of methodological advances for the practice of personnel selection: How practitioners beneﬁt from meta-analysis. Academy of Management Perspectives, 21(3), 6–15. Liao-Troth, M. A. (2005). Are they here for the long-haul? The effects of functional motives and personality factors on the psychological contracts of volunteers. Nonproﬁt and Voluntary Sector Quarterly, 34(4), 510–530. Lord, R. G., & Brown, D. J. (2001). Leadership, values and subordinate self-concepts. The Leadership Quarterly, 12, 133–152. Lord, R. G., Brown, D. J., & Freiberg, S. J. (1999). Understanding the dynamics of leadership: The role of follower self-concepts in the leader/follower relationship. Organizational Behavior and Human Decision Processes, 78, 167–203. Lum, T. Y., & Lightfoot, E. (2005). The effects of volunteering on the physical and mental health of older people. Research on Aging, 27(1), 31–55. Markus, H., & Wurf, E. (1986). The dynamic self-concept: A social psychological perspective. Annual Review of Psychology, 3, 299–337. Mowen, J. C., & Sujan, H. (2005). Volunteer behavior: A hierarchical model approach for investigating its trait and functional motive antecedents. Journal of Consumer Psychology, 15(2), 170–182. Mumford, E. (1906/1907). Origins of leadership. American Journal of Sociology, 12, 216–240. Neuman, W. L. (2004). Basics of social research (Vol. 138–139). Boston: Allyn and Bacon/Pearson. Olmstead, J. (2007). Predict future success with structured interviews. Nursing Management (March), 52–53. Omoto, A. M., Snyder, M., & Martino, S. (2000). Volunteerism and the life course: Investigating age-related agendas for action. Basic and Applied Social Psychology, 22(3), 181–197. Pearce, J. L. (1993). Volunteers: The organizational behavior of unpaid workers. London: Routledge. Reeser, J. C., Berg, R. L., Rhea, D., & Willick, S. (2005). Motivation and satisfaction among polyclinic volunteers at the 2002 Winter Olympic and Paralympic Games. British Journal of Sports Medicine, 39(4), e20, Accessed June 14, 2009 at http://bjsm.bmj.com.ezproxy. uwindsor.ca/content/vol39/issue4/#ORIGINAL_ARTICLES

Identifying Volunteer Leader Potential in Healthcare

47

Stogdill, R. (1950). Leadership, membership and organization. Psychological Bulletin, 47, 1–14. Stolinski, A. M., Ryan, C. S., Hausmann, L. R. M., & Wernli, M. A. (2004). Empathy, guilt, volunteer experiences, and intentions to continue volunteering among buddy volunteers in an AIDS organization. Journal of Applied Biobehavioral Research, 9(1), 1–22. Van Willigen, M. (2000). Differential beneﬁts of volunteering across the life course. Journals of Gerontology Series B-Psychological Sciences and Social Sciences, 55(5), S308–S318. Warburton, J., & Terry, D. J. (2000). Volunteer decision making by older people: A test of a revised theory of planned behavior. Basic and Applied Social Psychology, 22, 245–257. Wilson, A., & Pimm, G. (1996). The tyranny of the volunteer. The care and feeding of voluntary workforces. Management Decision, 34(4), 24–35. Wilson, J. (2000). Volunteering. Annual Review of Sociology, 26, 215–240. Yin, R. K. (1984). Case study research: Design and methods (1st ed.). Beverley Hills, CA: Sage.

This page intentionally left blank

LEADERSHIP STRATEGIES FOR BIOTECHNOLOGY ORGANIZATIONS: A LITERATURE REVIEW Lynn Johnson Langer ABSTRACT This research explored the literature regarding successful leadership practices and how these practices form the organizational context that leads to success in the biotechnology industry. Dominate themes emerged in general leadership strategies, leading research and development scientists, moving ideas from research to the consumer and the culture of research versus practice. Themes include leaders must be adaptable and able to lead effectively in a dynamic environment. Leaders need to consistently articulate the vision throughout the organization. Leaders need to be strategic decision-makers and ﬂexible enough to allow the vision to adjust to the culture and the environment. Leaders need to communicate effectively and create an organization where communication ﬂows efﬁciently at all levels. Leaders need to recognize clear cultural differences between functional groups, and they need to empower employees at all levels to make strategic decisions. Leaders need to know which decisions must be retained as his or her sole responsibility.

Biennial Review of Health Care Management: Meso Perspectives Advances in Health Care Management, Volume 8, 49–80 Copyright r 2009 by Emerald Group Publishing Limited All rights of reproduction in any form reserved ISSN: 1474-8231/doi:10.1108/S1474-8231(2009)0000008007

49

50

LYNN JOHNSON LANGER

INTRODUCTION Background This literature review investigates the published research regarding the exploratory question how to lead biotechnology organizations to success. There is sparse speciﬁc literature that addresses this important question. Biotechnology is a relatively young industry that has evolved rapidly since its commercial beginnings in the early 1970s. Since then, it has developed into a multibillion dollar industry with many successful products that have contributed to society with new biopharmaceuticals, food, and energy products and processes. However, many products do not make it to market because of extremely high development costs and the length of time to develop and commercialize new products. Biotechnology is the use of genetically modiﬁed organisms to make products or solve problems. Genentech, founded in 1976, was the ﬁrst biotechnology organization created on the idea that genes could be spliced into bacteria that would then produce therapeutic proteins under highly controlled conditions. Genentech received the ﬁrst United States Food and Drug Administration (USFDA), (n.d.) approval for a biotechnology therapy for recombinant human insulin in 1983. Since that time, biotechnology has produced ‘‘254 drugs approved for 392 indications, including treatments for cancer, diabetes, HIV/AIDS and autoimmune diseases. There are now more than 400 biotech drug products and vaccines currently in clinical trials targeting more than 200 diseases’’ (Biotechnology Industry Organization, 2007). By early 2006, there were 1,415 biotechnology companies in the United States with revenues over $51 billion. Biotechnology is the most researchintensive industry in the world; the US biotech industry spent $19.8 billion on research and development (R&D) in 2005 alone (Biotechnology Industry Organization, 2007). Although biotechnology R&D has provided numerous beneﬁts to society, many promising discoveries have not reached consumers because of the difﬁculties involved in bringing new technologies to market. The high costs and resources required in early-stage research, discovery, and development of new biotechnology ideas, especially for biopharmaceuticals, can be prohibitive for all but the most promising projects. The Tufts Center for Drug Development estimates that it costs more than $800 million to bring a biopharmaceutical product from the research bench to the consumer (Tufts E-News, 2007). This number includes the amortized costs of multiple failures and hundreds of millions in marketing costs.

Leadership Strategies for Biotechnology Organizations

51

Economics of the Biopharmaceutical Industry The biopharmaceutical industry is an extremely high-risk market segment because of the extraordinary time and expense involved in getting a new drug approved. The potential revenue of more than $1 billion for a blockbuster drug draws many venture capitalists to the industry. It is the science, though, that makes the industry possible. Most biotechnology ﬁrms are started by biomedical scientists. In fact, ‘‘academic entrepreneurs account for 43.1 percent of biotech founders, far greater than any other industry’’ (Zhang & Patel, 2005, p. 58). For every start-up biotech ﬁrm that succeeds, 15–20 fail and eight of 10 drugs fail in clinical trials (Federal Reserve Bank of Dallas, 2007; Stanford, Graduate School of Business, n.d.). Zhang & Patel (2005) examined 351 biotechnology ﬁrms that had founder information and determined that 626 entrepreneurs were involved in starting ﬁrms. Of the 351 ﬁrms, they examined 58.2 percent were founded by scientists; founders of the remaining ﬁrms were not described. Zhang and Patel examined the career paths of the founders and discovered that 43.1 percent of biotechnology ﬁrm founders were university-based professors or researchers and 46.7 percent engaged in nonproﬁt research. Given that the majority of biotechnology ﬁrms are started by research or academic scientists, the success of these ﬁrms is dependent on either the founder/scientist intuitively knowing how to successfully run an organization or business talent being brought in. Pisano (2006) deﬁnes success in the biotechnology industry as, ‘‘how effectively an organization or industry uses the capital it raises, and more speciﬁcally, how well it creates true value from its activities. Ultimately, value creation and capture are what matters’’ (p. 162). Pisano argues that existing approaches to business and organizations are not satisfactory to biotechnology. ‘‘Organizational and institutional innovations are needed in order to unlock the potential of biotechnology’’ (p. 202). Given that very few biotechnology organizations succeed and most organizations are founded by scientists, questions arise about what conditions in biotechnology organizations are necessary for success (Zhang & Patel, 2005). There are some highly successful biotechnology organizations with scientist founders. Notable examples include Genentech, Amgen, and Genzyme. What are the leaders of these organizations doing right given that so many fail? The purpose of this literature review was to examine success and does not evaluate the conditions that lead to failure.

52

LYNN JOHNSON LANGER

Clearly, the nation’s health is affected by the success of biotechnology and particularly the biopharmaceutical industry. Very little research has been done on effective methods for helping scientists succeed when they start a new biopharmaceutical company. Some research indicates that opportunities for communication between business people and scientists are important (Allen, 1984; Argote & Ingram, 2000; Hirst & Mann, 2004; Kivimaki et al., 2000; Langer, 2007, 2008; Nonaka, 1995; Rogers, 2003; Yin and Gwaltney, 1981). Other studies have examined leadership in R&D organizations (Elkins & Keller, 2003; Hirst & Mann, 2004; Sapienza, 2005; Shim & Lee, 2001). Tidd, Bessant, and Pavitt (2005) studied 94 biotechnology start-up businesses and found that three factors were associated with success: location within a signiﬁcant concentration of similar ﬁrms, quality of scientiﬁc staff (measured by citations), and the commercial experience of the founder.

Overview of Literature and Framework for Methodology Although most biotechnology organizations are founded and run by scientists and most fail, there is very little research that directly addresses leadership in successful biotechnology or biopharmaceutical organizations. To explore this subject, a strategy was used to examine various topics that are tangentially related in an effort to build a framework of common themes. The themes would provide some understanding of what scientist leaders must do to ensure success. This framework consists of leadership strategies for successfully leading any organization, and more speciﬁcally, leadership strategies for how to lead R&D scientists; most biotechnology organizations are research intensive. Furthermore, the literature regarding how scientists move ideas from the laboratory was included in the framework to better understand what scientists must speciﬁcally do to promote the product development process. To best understand how successful biopharmaceutical organizations are led, it is important to examine what the literature states that scientists themselves do to move ideas from the laboratory to commercialization. Literature is reviewed that investigates the culture of R&D scientists as it relates to business. Key themes in moving ideas from the laboratory to engage business leaders to take action that are supported by the literature include communication, networking, culture, and leadership. Lastly, literature was reviewed to understand the culture of research versus a business culture to better understand the personal transition scientists must make to become

Leadership Strategies for Biotechnology Organizations

53

successful business leaders in the biotechnology industry. On the basis of these research ﬁndings from the literature reviewed, a number of requirements have been identiﬁed that may be important for achieving success for newly emerging biotechnology organizations. These requirements include not just general leadership strategies broadly described by leadership scholars listed later, but speciﬁc strategies for leading scientists that center on the leader’s ability to continuously learn and adapt in various stages of the organization’s growth and to maintain a high level of communication at all levels of the organization. Although communication was a common theme in all the literature, to understand what happens in organizations that are successfully developing biotechnology products, one must also consider the collective processes that include networking, leadership, and organizational culture. The typical organizational culture within biotechnology companies tends to be more oriented toward individual research and scientiﬁc discovery, and less on human interactions. The following review summarizes the current state of the literature as it relates to the framework described earlier.

LITERATURE REVIEW Overview of Literature Requirements for managerial and technical capabilities within an organization differ from one stage of drug development to the next (Greiner, 1998). To be successful, organizations must adapt to the environment, and leaders must understand how to lead in uncertain circumstances. Because of the long cycle time to commercialize new drug discoveries, leaders, and the organization itself, must regularly transform. However, successfully moving the fruits of scientiﬁc discovery from the laboratory to commercialization requires a greater focus on leadership, organizational behavior, and ﬁnance than on other functional or operational areas of the organization (Langer, 2008). Financial strategies for commercializing technologies are well researched and reported on by large consulting ﬁrms such as Ernst and Young, McKinsey, PricewaterhouseCoopers and others. There is a need, however, to understand from the literature the leadership practices in successful biotechnology organizations. A large body of literature exists on strategies for leading change in organizations. There is also literature available on how

54

LYNN JOHNSON LANGER

to lead R&D scientists. However, the organizational behavior and what scientists themselves must do to move ideas from the laboratory to engage business leaders to take action are not as well researched, although it was recently reported on by Langer (2007). Most biotechnology organizations are in the early stages of the company’s lifecycle (Biotechnology Industry Organization, 2007). Lifecycles of organizations are also well-researched, but none of the current research speciﬁcally relates to the biotechnology industry; although recent literature does mention the need for new organizational structures to meet the needs of the modern organizations (Greiner, 1998). The following sections review the seminal leadership literature regarding the theoretical framework of practices and strategies for leading change, organization change literature, research regarding leading R&D scientists, and literature about what scientists need to do to successfully move ideas from the laboratory. Table 1 includes a summary of the strategies discussed later.

General Leadership Practices and Strategies To successfully negotiate critical organizational changes, leaders work within a framework of responsibilities. However, these responsibilities need to be balanced against individual and organizational needs to ensure changes are effectively implemented. The publications were chosen because they offer speciﬁc strategies for the leader, as opposed to studies of leadership traits or personalities. Most scholars mention vision as a crucial component of leading any organizational change (Bennis & Nanus, 1997; Brown, 1995; Covey, 1996; Dess & Picken, 2005; Drath, 2001; Farkas & De Backer, 1996; Gardner, 1995; Geller, 2002; Gill, 2003; Greenleaf, 1991; Greiner, 1998; Heifetz, 1994; Hemp & Stewart, 2004; Kanter, 2004; Kellerman, 2004; Kotter, 1996, 1999; Kouzes & Posner, 2002; Langer, 2008; Miller, 2002; O’Toole, 1995; Rooke & Torbert, 2005; Ruvolo & Bullis, 2003; Senge, 1990; Skipton, 2003; Wheatley, 1999). Most also mention that the organization needs to become a learning organization and be able to adapt to new problems and challenges (Bennis & Nanus, 1997; Brown, 1995; Dess & Picken, 2005; Drath, 2001; Greiner, 1998; Handy, 1995; Heifetz, 1994; Heifetz & Laurie, 2001; Kellerman, 2004; Kotter, 1996; Kouzes & Posner, 2002; Langer, 2008; McGregor, 1957; Miller, 2002; Pascale & Sternin, 2005; Priestland & Hanig, 2005; Ruvolo & Bullis, 2003; Senge, 1990; Vaill, 1996; Wheatley, 1999). Another strategy frequently mentioned

Table 1.

X X X X X X X X X X X X X X X X X X

X X

X X

X

X X X

X X X

X X X X X

X X X

X X X

X X

Other

Trust X

X-E X-E

X Servant Expert

X X X

X X X Servant

X X X

X

X

X

X X X

X

X

X X X X

X X X X

X X

X X

X X

X X X

X X

X X X

X X-E X-E

X

Accountability Limit Tenure

X X Theory X, Y

X X X

X

X

X

X

X X X

X X X X

X

X X

X X X

Theory Y Facilitator Conﬁdence

X X X

X X

X-E X

X X

X X

X

55

Bennis Brown Covey Dess Drath Farkas Gardner Geller Gill Greenleaf Greiner Handy Heifetz Hemp Kanter Kellerman Kotter Kouzes Langer McGregor Mezirow Miller O’Toole Pascale Priestland Rooke Ruvolo Senge Skipton Vaill Wheatley

Goal Follower Motivation Power (P)/ Politics Small Vision/ Learning Communication/ Field Theory/ Disorienting Embody Wins Empower Dilemma/ the Story Alignment Active Listening Systems Values Org/ (E) Sense of Culture/ Adaptive Urgency Holding Challenge Environment

Leadership Strategies for Biotechnology Organizations

Author

Leadership Strategies.

56

LYNN JOHNSON LANGER

in the literature is the importance of good communication and active listening by the leader. Creating a sense of urgency or a disorienting dilemma was also cited as important in changing leadership (Kotter, 1996; Mezirow, 1991). Some scholars mention that the leader needs to embody his, or her, own story, view the organization as an entire system, and recognize that what affects one area of an organization can also impact others (Bennis & Nanus, 1997; Gardner, 1995; Senge, 1990). The necessity of a culture that is sensitive to the needs of the members in the organization and creates a holding environment for employees is also ranked as important (Heifetz, 1994; Senge, 1990). These strategies, although intended for leaders of any organization, do not speciﬁcally address the needs of scientists as leaders.

Organization Phase General leadership strategies alone are insufﬁcient to successfully lead biotechnology organizations. Greiner (1998), Handy (1995), and Skipton (2003) discuss the importance of the organizational phase to the leader and leadership strategy. In the 21st century, new types of organizations are required. Because of the rapidly accelerating rate of change within our economic systems, organizations will need to be increasingly more ﬂuid and adaptable, as organizations have evolved from the Industrial and Modern Age to Postmodern organizations. Greiner (1998) describes the leadership and followership requirements of the organization as it proceeds through its life cycle. It should be noted that biopharmaceutical organizations progress somewhat more slowly through an organization life cycle, partly because of the long time required to develop new drugs. Greiner’s (1998) Phase 1 of the organization life cycle requires an entrepreneurial leader who is charismatic and communicates frequently and informally. (Phase 1 is similar to the startup phase of a new biopharmaceutical organization.) Greiner’s Phase 2 describes the more mature organization that requires leaders who are more directive and able to put formal processes in place that better handle company functional areas such as ﬁnance and marketing. (Because biopharmaceutical companies have products in the pipeline, speciﬁc processes that adhere to USFDA standards and regulation must be in place. For some scientists, it is difﬁcult to adjust to these highly structured requirements.) As the company continues to mature, into Phase 3, it requires the leader to delegate more responsibility and become more decentralized. ‘‘Much greater responsibility is given to the

Leadership Strategies for Biotechnology Organizations

57

managers of plants and market territories’’ (p. 62). Communication from the senior executive to the lower levels in the hierarchy is now formal and infrequent. In Phase 4, a formal structure is required with more formal planning procedures. Product groups replace decentralized units. The ﬁnal or most mature phase of the organization is Phase 5, where collaboration is required to overcome the red-tape of the now large bureaucracy. A matrixstyle structure is implemented where cross-departmental groups are able to interact and solve problems. Organizations generally pass from one stage of growth into another by what Greiner (1998) describes as crises. A crisis of leadership is what propels an organization into Phase 2, when the entrepreneurial leader is no longer able to single-handedly run the organization. Phase 3 is brought about by a crisis of autonomy. ‘‘Lower-level employees ﬁnd themselves restricted by a cumbersome and centralized hierarchy. They have come to possess more direct knowledge about markets and machinery than do their leaders at the top; consequently, they feel torn between following procedures and taking initiative on their own’’ (Greiner, 1998, p. 60). This crisis is resolved by more delegation to lower-level managers. Particularly with lower-level scientists, many may feel frustrated as they may feel they know more about their duties and responsibilities than their doctoral-educated PhD scientist supervisors. The third crisis occurs at the end of Phase 3 when ‘‘top-management teams attempt to return to centralized management, which usually fails because of the organization’s new vast scope of operations. Those companies that move ahead ﬁnd a new solution in the use of special coordination techniques’’ (Greiner, 1998, p. 62). The attempt to return to centralized management ultimately gives way to the crisis of red-tape. Managers become frustrated with direction from off-site senior management direction not in tune with local conditions. Both senior and lower-level managers are critical of the bureaucratic system. Leaders must carefully analyze the stages of organizational development and be prepared, in advance, to put managers and systems into place before the organization devolves toward entropy. Followers must be prepared to understand the changing organizational structure and needs and be willing to accept changing responsibility and new job requirements. Leaders are responsible for educating their followers in this regard. Greiner (1998) argues that learning from the history of changing organizations, leaders and followers can be better prepared to grow with the company. Both leaders and employees can better determine whether they wish to continue with the changing organization or ﬁnd a company in the organizational stage to

58

LYNN JOHNSON LANGER

which they are most happy and best-suited. (Successful biopharmaceutical organizations are able to successfully navigate the changes required at each stage.) Handy (1995) describes four main types of organizations using Greek Gods as an analogy. A Zeus organization is deﬁned as a group with one powerful leader. Those closest to the leader are the ones most trusted and who have conferred power. A Zeus organization is similar to Greiner’s Phase 1 leader (Greiner, 1998). Apollo is the typical hierarchical situation where the organization is divided by functional areas and the workers are deﬁned by their role in the area. This organization type is typiﬁed by the government bureaucracy or a government laboratory. Handy claims that in the Apollo organization, workers are the ‘‘servant of technologyyand are hired to operate, service or often just watch increasingly sophisticated equipmentyand be in many senses its servants’’ (p. 191). This organizational structure is similar to Greiner’s Phase 4 organization, or in the Coordination phase. In an Athena organization, short-term project groups are assembled to solve problems and consist of whoever is needed from different functional areas. Lastly, the Dionysus organization is a group of independent individuals loosely joined by a common need, such as a group of lawyers sharing a suite with a shared ofﬁce manager. University departments are often formed in the Dionysus style. According to Handy (1995), the university is a Dionysus organization: ‘‘The organization exists to help the individual achieve his purpose’’ (p. 24). The Athena organization is similar to Greiner’s Phase 5, or Collaboration phase, and is often how biopharmaceutical companies begin. Relatively few people are suited to employment at all stages of a company’s development. Yet for leaders to be constantly growing a company from one phase to another is regarded in much the same way as promotions within American companies. Success is based on title, position, and salary. To be considered successful, scientists may be expected to leave the laboratory and manage people; a job responsibility in which they may have no training or support. ‘‘Organizations must respond continually to their environment, even if they do not themselves set out to change it’’ (Handy, 1995, p. 81). New organizations in the 21st century, according to Handy are moving away from this paradigm to newer organizational structures that are more short-term and team-project oriented. Handy predicts that the factory or the ofﬁce will ‘‘give wayyto a more contractual, dispersed and federal organization. It will lead to more small businesses, particularly in the services and more self employment’’ (p. 7).

Leadership Strategies for Biotechnology Organizations

59

Leonard Skipton (2003) describes the evolution of leadership theory over the past 50 years, but also discusses what the postmodern organization must do to foster new types of leaders. ‘‘The requirements for leadership change in contemporary organizations have changed signiﬁcantly in the past several decades’’ (p. 3). Additionally, Skipton states that there must be ‘‘a clear lineof-sight between leadership training and participant goals and objectives must be demonstrated in order to motivate many participants effectively’’ (p. 11). Program developers must also ‘‘place a high priority on shaping the work environment that participants go back to so that (a) they have an opportunity to practice their new skills on real-world task and problems and get useful feedback and (b) the organization will support and reward them for not only taking the course, but also demonstrating the new capabilities on the job’’ (p. 11). Leadership programs ‘‘must incorporate personal organizational missions, strategic goals, cultural values and assumptions, and core challenges for the sponsoring organization in order to engage both the participant and the larger organization in the developmental process’’ (p. 12). Are traditional leadership strategies sufﬁcient to successfully lead biotechnology organizations? Because biotechnology companies are generally heavily focused on research (Biotechnology Industry Organization, 2007), the next section of this review examines the literature speciﬁcally concerning ‘‘How to lead Research & Development (R&D) Scientists?’’

Leading R&D Scientists The unique challenges involved in leading scientists and engineers have been extensively researched during the past decade. This literature provides insight that scientist/founders may use as they lead heavily research-oriented organizations. More recently, scientists in the biotechnology industry have speciﬁcally been the focus of inquiry with regard to differences in leadership, and the effectiveness of leadership approaches. The biotechnology industry is relatively young and is experiencing rapid growth. Many of the discoveries in biotechnology have come from scientists and researchers in government and academic laboratories. Often, highly intelligent scientists leave the laboratory to join commercial organizations, or newly formed non-proﬁt product development partnerships (PDPs). PDPs are new organizational structures whose mission is to develop and distribute drug products for under-served diseases. PDPs are funded by contributions from major foundations, such as the Bill and Melinda Gates

60

LYNN JOHNSON LANGER

Foundation, and governments and have been created to ﬁll a gap left by traditional biopharmaceutical companies. Traditional for-proﬁt companies often do not focus on diseases with small populations or diseases that mainly affect third-world countries because the proﬁt margins are too low to justify the very large investments required for R&D for new drugs. PDPs are often unique in that they can have a more traditional business focus and results-orientation that are not typical for non-proﬁt organizations. This hybrid-type organization may well become a key player in the future of biopharmaceutical drug development, especially for major diseases such as HIV/AIDS, Tuberculosis, and Malaria that ravage poor countries. This shift may become more common as the relatively large wealth that young, technology-oriented entrepreneurs with altruistic values have accumulated, such as Bill and Melinda Gates. Because the concept of the PDP has only recently been created, there is currently no scholarly literature on leadership and PDPs. However, scientists who work in PDPs face the same difﬁculties as other scientists that may leave academia. These scientists may bring with them cultures and attitudes that are not effective in a highly goal-oriented environment, regardless of whether it is a for-proﬁt, or otherwise. For biotechnology discoveries to ultimately reach the consumer, scientists must frequently shift from a purely research mentality to one of rapid development, production, and commercialization. This shift in perspective is a major issue as many biotechnology organizations are led by former researchers who lack business experience. The question examined in the literature is, ‘‘What leadership practices enhance performance of scientists in R&D groups?’’ More than 250 articles were reviewed and categorized into four speciﬁc themes that relate to the research question. These include supervisory practice, performance, autonomy, and creativity. Researchers used several types of inquiry, including qualitative, quantitative, and mixed methods. The literature cited was chosen because it is a good representation of the themes discovered and because it represents a variety of relevant research methods. Supervisory Practice Supervision of scientists varies by individual, but because scientists tend to be creative and prefer autonomy, leading scientists can be different than leading individuals in the general population. Supervisory practice may include both leaders and managers. Transformational or facilitative leaders can balance the negative effect of obstacles in the climate of R&D teams

Leadership Strategies for Biotechnology Organizations

61

(Pirola-Merlo, Ha¨rtel, Mann, & Hirst, 2002). Transformational leadership is related to employee satisfaction in R&D teams, but transformational leadership is unrelated to creativity because it distracted people from activities at hand (Berson & Linton, 2005; Sosik, Kahai, & Avolio, 1999). The traditional leadership strategy of clearly deﬁning goals may actually reduce intrinsic motivation in R&D scientists (Hennessey & Amabile, 1988; Mullin & Sherman, 1993). Sapienza (2005) studied how scientists deﬁne effective leaders, and Cordero, Farris, and DiTomaso (2004) examined supervisory practice as it relates to performance and the work environment. As scientists transition from individual contributor to team player, the skills required to perform also change. Scientists, whose leadership responsibilities require 25% or more of their time, practice leadership skills, as deﬁned by Kouzes and Posner (2002), more frequently than those scientists who spend less than 25% of their time leading. These practices include inspiring a shared vision and encouraging the heart, by recognizing contributions and celebrating values. Scientists are more effective when they have a variety of tasks rather than a single task (Andrews & Pelz, 1976). This ﬁnding suggests that leadership identiﬁed as a role may actually increase the overall effectiveness of the scientist. Scientists who are trained speciﬁcally in leadership may not only increase their use of leadership practices but also increase their effectiveness as a scientist (Day, 2003). Day’s research leads to questions as to whether identifying leadership as a role for all employees would increase work effectiveness in general and could be the subject of future research. Scientists deﬁne an effective leader as being caring and compassionate, technically accomplished, a good role model, and possessing managerial skills. Scientists view a bad leader as one who is abusive, exploitative, or unable to deal with conﬂict (Sapienza, 2005). Supervisors who attempt to help employees by using only technical skills actually decrease the employee’s job satisfaction and job performance if the environment is already stimulating. In other words, in a stimulating environment, supervisors should not emphasize the use of technical skills in an attempt to improve job performance and job satisfaction. Concurrently, supervisors should not overuse administrative skills in an unstimulating environment (Cordero et al., 2004). In these situations, Cordero claims that people skills are most helpful in creating a stimulating environment in which to work. Performance Scientiﬁc performance is directly related to supervisory practice (Jabri, 1992; Jordan, 2005; Keller, 2006). Effective leaders create an increase in job

62

LYNN JOHNSON LANGER

satisfaction and performance in their subordinates (Keller, 2006). Job satisfaction and performance increase when scientists feel that the task allocation by the leader is appropriate (Jabri, 1992). ‘‘Effective leaders in R&D project groups tend to inspire a sense of mission and purpose about the importance of the work being done, stimulate new ways of thinking and problem solving, and encourage group members to do more than what might normally be expected’’ (Keller, p. 498). Transformational leadership may be useful in studying professional organizations. There is a correlation between job satisfaction and job performance when task allocation is seen as appropriate, but no correlation between satisfaction and performance when allocation was seen as inappropriate (Jabri, 1992). Future research could incorporate more variables, including ‘‘conditions of goal setting, the processes that underlie task allocation decisions, and issues relating to control over task, control over others, and control over the work environment’’ (Jabri, 1992, p. 98). A longitudinal study by the US Department of Energy found 37 factors that R&D workers ﬁnd important that may increase their performance (Jordan, 2005). R&D workers cited 8 of the 37 factors to be most important. These factors include having a ‘‘clear research vision,’’ investing ‘‘in future capabilities,’’ making ‘‘sure staff have challenging work,’’ championing ‘‘long term foundational research,’’ having a ‘‘systematic way to identify new partnerships,’’ measuring ‘‘the success of each project appropriately,’’ ensuring ‘‘managers are technically competent,’’ and researchers appreciate non-monetary rewards’’ (p. 32). Autonomy, Creativity, and Innovation Scientists are generally thought to prefer a high level of autonomy (Bailyn, 1987; Trevelyan, 2001). However, the type of autonomy is important in understanding satisfaction levels of the scientists. Bailyn describes two types of autonomy: strategic autonomy (the freedom to set one’s own research agenda) and operational autonomy (the freedom, once a problem has been set to attack it by means determined by oneself, within given resource constraints)yTechnical careers in the R&D lab should start lower on strategic than on operational autonomy, that operational autonomy show initial fairly rapid increase, which should be followed by increases in strategic autonomy, and that thereafter a number of different career paths should be available for technical employees. (p. 129)

Trevelyan (2001) researched academic researchers as opposed to scientists in R&D laboratories in non-academic organizations. Trevelyan described two different types of autonomy, one where the leader is highly directive, yet

Leadership Strategies for Biotechnology Organizations

63

not involved and the other where the leader is not highly directive, but very involved. Leaders who are highly directive with research scientists cause a decrease in satisfaction, but leaders who are non-directive and highly involved with valued input creates an increase in satisfaction which may also lead to an increase in creativity and innovation (Cardinal, 2001; Judge, Fryxell, & Dooley, 1997; Trevelyan, 2001). A high degree of operational autonomy increases innovation. Scientists are typically goal-directed in biotechnology organizations and work best when the organization’s management develops strategic objectives and context, but they require freedom to work independently within that context (Judge et al., 1997; Essex & Kusy, 2004). According to Essex & Kusy, ‘‘Create the loosest boundaries your organizational culture will allow, then let the mavericks out of the corral’’ (p. 128). ‘‘Balancing autonomy, personalized recognition systems, integrated sociotechnical systems, and continuity of slack’’ have a major inﬂuence on whether or not there was a ‘‘goal-directed community in the R&D unit. Those ﬁrms that were the most innovative emphasized the importance of operational autonomy for the researchers, but retain strategic autonomy for top management’’ (Judge et al., pp. 76–77). Scientists exhibit many of the traits described by Essex & Kusy as ‘‘mavericks.’’ Essex & Kusy point out that ‘‘the most important thing a leader of mavericks can do is to run interference for them, giving them some protection from the obstacles inherent in organizational life’’ (p. 128). Cardinal (2001) studied 57 pharmaceutical ﬁrms using ‘‘incremental innovations in the form of drug enhancements and radical innovations in the form of new drugs’’ (p. 19) as dependent variables. The use of input, output, and behavioral controls in R&D groups in US pharmaceutical companies enhances radical innovation. Input and output controls enhance incremental innovation when behavior controls are not in place. One reason that controls may be necessary for enhanced innovation in the pharmaceutical industry is that ‘‘in the case of more uncertain technologies, learning occurs through error-induced discoveries. Looking at the evolution of several critical drugs and their corollariesyfeedback loops created by both incremental and radical innovations are reciprocal, with both leading to subsequent learning’’ (p. 29). Summary To summarize, supervisory practice, performance, autonomy, and creativity are all important considerations when leading scientists. Scientists deﬁne effective leaders as being caring and compassionate, technically

64

LYNN JOHNSON LANGER

accomplished, good role models, and possessing managerial skills (Sapienza, 2005). Effective leaders increase job satisfaction and performance (Keller, 2006). Job satisfaction and performance increase when scientists feel that the task allocation by the leader is appropriate, the leader has inspired a sense of mission and purpose, stimulated new ways of thinking and problem solving, and have been encouraged to do more than what might normally be expected (Jabri, 1992; Jordan, 2005; Keller, 2006). Leaders should be highly involved, but not highly directive with research scientists and should support an environment conducive to creativity (Cardinal, 2001; Essex & Kusy, 2004; Judge et al., 1997; Trevelyan, 2001). People skills are most helpful in creating a stimulating environment in which to work. Supervisors should not use technical skills in an attempt to improve job performance and job satisfaction and should not over-use administrative skills in an unstimulating environment (Cordero et al., 2004). Traditional leadership strategies such as deﬁning goals may reduce intrinsic motivation in R&D scientists (Hennessey & Amabile, 1988; Mullin & Sherman, 1993). Scientists are more effective when they have various tasks rather than a single task (Andrews & Pelz, 1976). Leadership identiﬁed as a role may increase the overall effectiveness of the scientist. Scientists who are trained speciﬁcally in leadership may increase their effectiveness as a scientist (Day, 2003). Scientists may not see the need for leadership, and they may use skepticism as a subtle form of control over younger scientists and may overidealize technology (Feldman, 1989; Owen-Smith, 2001). The use of input, output, and behavioral controls in R&D groups enhances innovation (Cardinal, 2001). Technical careers in the R&D laboratory should start lower on strategic autonomy than on operational autonomy, operational autonomy may show initial fairly rapid increase, and a number of different career paths should be available (Bailyn, 1987). Conventional wisdom about leadership such as the importance of communication is also true for leading scientists. However, scientists prefer a higher degree of autonomy in setting goals, so that conventional leadership strategies are not always directly appropriate to scientists. Because of the importance of creativity in research, leaders need to help foster an atmosphere and culture of shared communication and autonomy. The role scientists themselves play in moving ideas from the laboratory to engage business leaders to take action is directly impacted by the leadership strategies used within the organization.

Leadership Strategies for Biotechnology Organizations

65

Moving Ideas: Research Bench to Consumer To best understand how successful biotechnology organizations are led, it is important to examine what scientists themselves do to move ideas from the laboratory to commercialization. This section of the chapter will examine the literature on how scientists move ideas from the laboratory to engage business leaders to take action. Key themes that arose from the literature include communication, networking, culture, and leadership. These themes have various levels of inﬂuence on the process, but communication and networking were speciﬁcally identiﬁed as having the most signiﬁcant inﬂuence. Biotechnology companies are based on the application of research discoveries. The application of research discoveries requires effective interdisciplinary teamwork and communication. Unfortunately, many scientists lack training in these skills. Although the majority of authors state that communication is one of the critical elements required to move ideas out of the laboratory, most scientists are not trained in such communication skills, and even worse, the R&D culture may subtly inhibit their ability and desire to effectively communicate. Scho¨n (1983) offers insight into the evolution of theory versus practice and academia that is important in understanding scientists in the biotechnology arena. He states that the ‘‘concept of ‘application’ leads to a view of professional as a hierarchy in which ‘general principles’ occupy the highest level and ‘concrete problem solving’ the lowestyThe application of basic science yields applied science. Applied science yields diagnostic and problem-solving techniques which are applied in turn to the actual delivery of services. The order of application is also an order of derivation and dependence. Applied science is said to ‘rest on’ the foundation of basic science. And the more basic and general the knowledge, the higher the status of its producer’’ (p. 24). Because most scientists in biotechnology were originally trained as researchers, Scho¨n’s conception may be the foundation of the divide between science and business. On either side of this divide, the highest status belongs to basic research. The lowest status belongs to the application of the science. As a result, a subtle cultural discrimination exists in the biotechnology and biopharmaceutical industry that diminishes the value of applied science and its research (Scho¨n, 1983). The foundation of the divide between science and practice began in the ﬁrst research universities in the United States. Scientists who start their careers by studying biochemistry and cell biology enter a culture where ‘‘hard science’’ carries the highest status. Along with

66

LYNN JOHNSON LANGER

this status comes a culture of not preparing scientists for the collaboration necessary to move science out of the laboratory. For the scientist to effectively bring science out of the laboratory, he or she must engage in reﬂective collaboration with practitioners (Scho¨n, 1983). This concept of collaboration is very different for scientists who typically prefer to keep to themselves or to stay involved with other scientists. Interestingly, a study of education levels among managers at life sciences organizations responsible for commercialization showed that 97% of these managers, at all levels, had an undergraduate or graduate life sciences degree (BioPlan Associates, Inc., 2003). This ﬁnding indicates that the great majority of people involved in product commercialization were originally educated as scientists, and as such, they may not have experience in, or the desire, to collaborate with non-scientists. Communication, Networking, and the Transfer of Knowledge Communication is critical in moving ideas out of the laboratory (Allen, 1984; Argote & Ingram, 2000; Hirst & Mann, 2004; Hoegl & Gemeuden, 2001; Langer, 2007; Kivimaki et al., 2000; Nonaka, 1995; Wainer & Rubin, 1969). High levels of communication and interaction between departments is an important factor in innovation (Kivimaki et al., 2000). However, interaction between groups predicts innovation less strongly than collaboration and too much interaction may actually interfere with innovation, if there are too many meetings and an overload of information. Collaboration means working together in a joint intellectual effort, whereas interaction is merely a mutual or reciprocal action. For scientists who may prefer to work alone, too many meetings may stiﬂe creativity, and the leader may want to help create an environment that limits the amount of meetings, but encourages collaboration. Social interaction is important in the dissemination of knowledge (Argote & Ingram, 2000; Nonaka, 1995; Yin & Gwaltney, 1981). The topic of communication between researchers and others in the organization does not need to be speciﬁc to the research at hand. Rather, the communication between researchers and practitioners raises each other’s consciousness about the other and may have serendipitous effects (Yin & Gwaltney, 1981). Researchers may change the focus of their research based on dialogue with end-users. In addition, users may change future projections to reﬂect ongoing research. In this environment, acceptance of a new technology ultimately occurs (or does not occur) based on the effectiveness of ongoing communications. A better understanding of other’s ways of thinking is important in building trust and further increasing communication.

Leadership Strategies for Biotechnology Organizations

67

‘‘Knowledge held by individuals, organizations, and societies can be simultaneously enlarged and enriched through the spiral, interactive ampliﬁcation of tacit and explicit knowledge held by individuals, organizations, and societies’’ (Nonaka, 1995, p. 34). Organization knowledge is more than organizational learning. It is the continuous incorporation and feedback of tacit to explicit knowledge throughout the organization and is increased through continuous dialogue between cross-functional teams. Cross-functional teams (teams composed of individuals from different functional areas within an organization, such as marketing and manufacturing) may lead to success in moving ideas out of the laboratory. Knowledge transfer occurs through communication (Argote & Ingram, 2000). Much of the knowledge an organization has is ‘‘tacit and may not be easily articulated’’ (p. 152). Therefore, to facilitate knowledge transfer, organizations need to provide multiple means of communication within and between groups. This concept may help explain why it is difﬁcult to move science from the laboratory because it is almost inevitably complex. Other studies focus on leadership and communication and its effect on team performance (Hirst & Mann, 2004; Waldman & Atwater, 1994). Hirst and Mann (2004) reported on a longitudinal study to discover what communication variables most affected team performance. They evaluated 350 employees from 56 teams in four organizations. The study surveyed and evaluated both research managers and project customers and discovered that organization stakeholders have different perspectives of which factors inﬂuenced performance most signiﬁcantly. Project customers’ ratings of performance were most inﬂuenced by communication safety. (In the study, communication safety was deﬁned as participative decision making, power conﬂict, and open discussion. Power conﬂict was deﬁned as ‘‘destructive rivalry between members of a team’’.) Task communication was the strongest predictor of both team member and research managers’ ratings of project performance. ‘‘Boundary spanning is most effective when performed by the project leader not the team’’ (p. 147). Although leadership training programs may rightly teach leaders to stimulate debate and act as ‘‘devil’s advocate,’’ more attention to systemic issues is critical if there is to be a long-lasting cultural and behavioral change. Leaders must be held accountable, and they must provide evidence of innovations being developed. ‘‘Organizations can enhance a leader’s boundary spanning ability by implementing mentoring programs which socialize leaders to organizational norms, practices and develop leaders’ inﬂuencing and championing skills’’ (p. 156). Important decisions will continue to be made informally through networks and leaders should be encouraged to build their networks. ‘‘In-house research forums

68

LYNN JOHNSON LANGER

and networking events particularly for peer groups provide an efﬁcient means of developing leaders’ intra-company links and the potential pool of knowledge sharing resources’’ (p. 156). Success in R&D is highly dependent on leadership and championing. ‘‘Championing behavior is y an essential element of project success y championing seems to be even more necessary at higher management levels’’ (Waldman & Atwater, 1994, p. 242). Championing behavior involves key individuals who promote an idea throughout the organization. This ﬁnding is consistent with the work of Scho¨n (1983). ‘‘Representation is also present as a related theme in that the effective leader is one who appreciates project members’ scientiﬁc expertise and takes care of paperwork and other administrative details by upper management’’ (Waldman & Atwater, 1994, p. 242). This concept is supported by Bennis and Beiderman in their book Organizing Genius (2005) who also argue that for highly creative people to be successful, they need strong administrative support. The opportunity to work closely to each other, or in proximity of each other, is important for scientists in networking and communication (Allen, 1984; Tidd et al., 2005). ‘‘Increased communication between R&D projects and other elements of the laboratory staff were in every case strongly related to project performance’’ (Allen, 1984, p. 123). Proximity may be one way to increase communication between scientists including having laboratories built around a central point to increase the chance of discussions, rather than in a linear hallway. At the same time, laboratories arranged in this central point fashion could decrease communication with other departments; this could be highly detrimental to moving ideas from the bench to consumer. Tidd et al. (2005) found three factors associated with success in a study of 94 biotechnology start-ups: ‘‘location within a signiﬁcant concentration of similar ﬁrms, quality of scientiﬁc staff (measured by citations), and the commercial experience of the founder. The number of alliances had no signiﬁcant effect on success, and the number of scientiﬁc staff in the top management team had a negative association, suggesting that the scientists are best kept in the laboratory. Other studies of biotechnology start-ups conﬁrm this pattern, and suggest that maintaining close links with universities reduces the level of R&D expenditure needed to increase the number of patents produced, and moderately increases the number of new products under development. However, as with more general alliances, the number of university links has no effect on the success or performance of biotechnology start-ups, but the quality of such relationships does’’ (p. 551). As mentioned previously, most biotechnology companies are started by scientist/entrepreneurs and most of them fail. In-group favoritism may be

Leadership Strategies for Biotechnology Organizations

69

particularly true with researchers versus other groups within an organization. The authors could have been describing researchers in biotech organizations. For science to be successfully transferred from the laboratory, it is important for laboratory scientists to identify with the larger organization and not just other researchers. Elements of Leadership and a Culture of Success Boundary spanning and commercial experience of the founder have been shown to be elements of leadership important to moving ideas from the laboratory (Hirst & Mann, 2004; Tidd et al., 2005). Others research shows that management action directly impacts project performance (Elkins & Keller, 2003; Shim & Lee, 2001). Innovation and movement of ideas out the laboratory are part of a process that is culture-dependent, but is also directly related to actions of the leader. ‘‘The leaders should also boundary span with important constituents outside the project group, such as managers and personnel in marketing, manufacturing, and operating divisions, as well as with customers from outside the ﬁrm. This kind of activity to champion the project can be critical to the survival and success of the project’’ (Elkins & Keller, 2003, p. 601). Rational tactics for inﬂuencing team members by R&D project leaders have a positive effect on the projects (Shim & Lee, 2001) These tactics include providing a clear understanding of why an action or a goal is important. Project leader activities may frequently include championing ideas across organizational boundaries to garner support and resources. Empowered employees may be a means of overcoming some of the problems of managing knowledge workers. However, empowered employees need direction and leadership if their efforts are not to be wasted. On the other hand, they also need a level of autonomy in how their work is actually carried out (Judge et al., 1997). Tampoe (1993) researched motivation and found that ﬁnancial, personal growth, operational autonomy and task achievement are all key motivators of scientists. However, even when ﬁnancial motivators are met, the need for personal growth, operational autonomy, and task achievement still exist. For biotechnology companies to succeed, managers and leaders must stay closely in tune with the scientists to best understand the employee’s motivation. Another study found that the most successful entrepreneurial scientists have a high need for achievement and have a high degree of self-efﬁcacy (Wainer & Rubin, 1969). Entrepreneurs with the highest need to achieve are often the most successful based on sales. For biotechnology companies to have a higher probability of success, the leadership should have a high need to achieve.

70

LYNN JOHNSON LANGER

Teamwork is also critical success factor in innovation (Hoegl & Gemeuden, 2001). ‘‘Managing innovation to move ideas from the laboratory is more than direct leadership, but also of orchestrating collective action of teams’’ (Stoker, Looise, Fisscher, & de Jong, 2001, p. 1141). Very little research has been done on the role of the leader in biotechnology organizations and the leader’s effect on the success of organization in terms of R&D of new ideas. ‘‘Self-managing teams are responsible for some part of a production process, and they function with signiﬁcant autonomy’’ (Stoker et al., p. 1143). Those scientists with a high degree of self-efﬁcacy believe they can achieve results and are often high performers. In summary, scientist-leaders should boundary-span outside the project group to champion the project, empower their employees, yet also provide clear direction and expectations of their project group have a high need to achieve, understand what motivates the individuals in their project group, and recognize the importance of self-managing teams.

Culture of Research versus Practice Key themes in moving ideas from the laboratory to engage business leaders to take action that are supported by the literature are communication, networking, culture, and leadership. However, there is such a strong cultural bias that favors research over practice, for scientists to effectively move ideas from the laboratory they must work within a culture that supports this effort. Because the culture of an organization is directly inﬂuenced by its leadership, the nature of the organization’s leadership is a critical component in the ability of scientists to move ideas from the laboratory and in developing a culture that supports or inhibits effective communication across boundaries. An organization’s culture is certainly inﬂuenced by factors other than leadership, but the core of company culture comes from its leadership. Therefore, further study is needed to understand the role of the leadership and how it inﬂuences cross-boundary communication within the organization. Moving ideas from the laboratory to commercialization involves complex interactions within the organization. To truly understand what happens in successful biotechnology organizations, additional new research should focus not only on the scientists but also the collective organization processes that include leadership, communication, networking, and organizational culture (Langer, 2007, 2008). These processes, in particular the organization’s culture, are affected by the practices of the leader. The culture of the

Leadership Strategies for Biotechnology Organizations

71

organization determines the level of communication and networking required to support a successful environment. Leadership practices have not, as yet, been explored in-depth in biotechnology companies. Because most biotechnology companies are founded by scientists, the research should speciﬁcally look at those companies with a scientist/founder/leader. Langer suggests multiple case studies involving interviews and archival data would provide important knowledge for the industry. Allen (1984) stated that when doing research on research scientists one must note that ‘‘each piece of work is unique’’ (p. 12). Because there can be no exact replication in the research, it ‘‘makes performance more difﬁcult to measure because there is no common denominator among projects to provide a basis for comparison’’ (p. 12). Allen points out that this lack of comparability makes the case study method a good choice for future research.

Summary of Findings The literature clearly deﬁnes important strategies that leaders should consider to help guide their company to success. These strategies include a guiding vision, creating a learning organization, excellent communication, active listening, a sense of urgency, embodying the story, viewing the organization as an entire system, and creating an atmosphere or culture that is a holding environment for employees. However, while leading scientists and biotechnology organizations, effective leaders also inspire a sense of mission and purpose. They are caring, compassionate, technically accomplished, good role models, stimulate new ways of thinking and problem solving, inspire followers to do more than they might otherwise do, and have good managerial skills. Task allocation should be appropriate; leaders should be highly involved, but not directive and have good people skills. Scientists should be given various tasks, including leadership opportunities and training, because this may increase their effectiveness as scientists. Leaders should be aware that scientists may not see the need for leadership and they may use skepticism as a form of control. However, management action directly affects project performance. Input, output, and behavioral controls should be in place to enhance innovation and scientists should be given various career paths. Moving ideas from the laboratory to the customer requires high levels of communication. The higher the level of interaction between departments the greater the rate of knowledge transfer. Prior commercial experience of the

72

LYNN JOHNSON LANGER

leader has also been shown to be important. Innovation and movement of ideas out the laboratory are part of a process that is culture-dependent, but is also directly related to actions of the leader.

Requirements for Achieving Success On the basis of these research ﬁndings from the literature reviewed, a number of requirements have been identiﬁed that may be important for achieving success for newly emerging biopharmaceutical organizations. These requirements center on the leader’s ability to continuously learn and adapt to various organization growth stages and organization needs. It appears that before a scientist even begins to establish a new organization, it is very important for him or her to have a realistic understanding of the evolving requirements involved in starting and running a biopharmaceutical company. A perception that the initial managerial, commercial, and scientiﬁc requirements will remain static may persuade scientists to ﬁnd organizations that will be at risk of failure as they grow beyond their initial stages. It is extremely important that scientist/founders be willing and able to continuously learn, adapt, and change. Because of the dynamic nature of this industry, scientist/founders need to learn how to suppress their ego and even be willing to give up complete control of the organization. As the company grows from its initial founding, it is very difﬁcult for the scientist/founders to know all aspects of running a vital organization, and they need to surround themselves with experienced people whose opinions they trust and to whom they can delegate responsibility. In addition, running a successful company will require scientist/founders to create an organization with high levels of communication and a culture of learning for all employees.

The Nature of Leadership Styles at Biotechnology Organizations Greiner (1998) argues that organizations go through ﬁve stages of growth that cause a signiﬁcant change in leadership requirements and structure of the organization. However, Greiner’s ﬁndings do not go far enough to describe the needs of the biotechnology organization. Biotechnology companies evolve differently than other organizations due to the regulated nature of the pharmaceutical industry (Langer, 2008). They reach key crisis points that result in the requirement for major changes in leadership and structure. For example, early-stage biotechnology organizations with scientist/founders can

Leadership Strategies for Biotechnology Organizations

73

initially be highly collaborative and entrepreneurial, but as the company grows, the decision-making styles and processes of communication must also change. This need for change is particularly true as biotechnology organizations begin clinical testing of their drug products in humans. As organizations move into the clinical testing phase, the leader now needs to work within a more stringent USFDA-regulated framework. Before human testing, the company is typically more research focused, and a collaborative style is expected and accepted. However, once a protocol for testing in humans has been established, changes can only be made with USFDA approval. Drug development companies require a decisive leader capable of delegating responsibility to expert senior executives in various, but interrelated, functional areas. At this point in the organization’s development, decisions need to be made by the leader that cannot be easily changed due to the greater regulatory oversight. The step to human clinical trials moves the biopharmaceutical company from a research orientation to a more operational, or product development orientation. Regulatory bodies such as the USFDA oversee and regulate all testing in humans and require substantial documentation. Maturing organizations need focused decision making from their scientist/founders regarding testing in humans, but this also continues to be a need for collaborative leadership in other aspects of the organization. This is one of the paradoxes of leadership in biopharmaceutical companies. Collaborative leadership continues to be critical in areas such as marketing, manufacturing, and the regulatory department as they work together to plan for the launch of a new product. This paradoxical challenge involves dealing with the complexity of simultaneously being both a directive and collaborative leader. Biopharmaceutical organizations experience ever-changing leadership needs within the context of ‘‘permanent whitewater’’ (Vaill, 1996). The essence of leaders who are learners and learning organizations is the ability to adapt and adjust as a complex and uncertain environment evolves. This evolution is not a linear process, but one that continuously circles back on itself while continuously moving forward. Furthermore, successful leaders recognize the requirement for different styles of leadership and decision making within different functional areas of the organization. Researchers, for example, may prefer a collaborative style of decision making, whereas clinical and regulatory staff may prefer a more focused, decisive style. In summary, scientist/founders need to be prepared to adapt their leadership style to the situation. At times leaders need to be collaborative, and at other times they need to be more directive. It is not unusual for scientists in general to have a highly collaborative management style because

74

LYNN JOHNSON LANGER

they are trained to collaborate and seek the opinion of others. This style of leadership may work well in the early stages of the organization. However, once a company advances its research to the product development stage and begins pre-clinical and clinical trials, a new style of management is often needed. As the product moves from pure R&D to a regulatory environment that requires stringent oversight for testing in human subjects, the organization’s managerial framework tends to become more rigid. Midstream shifts in business strategy, which may involve multiple organizational functions, can be tolerated early on, but not later on, when the drug begins to be tested in humans. For example, shifts in how, or to whom, the product may be marketed can no longer be easily dealt with as a collaborative decision once a product reaches a certain point. Such mid-stream changes are notoriously difﬁcult to manage once a product moves out of R&D. Leaders of organizations with products in clinical trials need to trust and, in some circumstances, even defer to the opinion of other experienced senior leaders. If the CEO is unable to trust or delegate some decisions to senior executives, a conﬂict may arise and company executives may become frustrated. This requirement does not mean that the CEO must give up all control but there must be staff in place that can be trusted to make the correct decision. The leader must be simultaneously tough and directive, while being collaborative and compassionate, consistent and predictable, and adaptable and open to creativity and dissent. If the leader can understand that the success of the organization depends on the leader’s ability to make strategic decisions that almost necessarily require input from others who have had prior success, then the leader may be more likely to accept advice from others. As companies evolve and change, so must their leaders. This adaptation requires that leaders ensure communication processes are in place as these changes transform the organization over time. There will always be a need for high-level, regular communication to decrease fear and increase trust and synergy among employees. However, as the company evolves, informal communication becomes less effective and more formal processes need to be established. As companies become global and geographically and culturally dispersed, communication issues can become an even greater concern.

CONCLUSION Biotechnology and biopharmaceutical companies require leaders who are able to continually learn and adapt to the continuous change of permanent

Leadership Strategies for Biotechnology Organizations

75

organizational whitewater likely to be present as technology organizations mature and develop (Heifetz, 1994; Langer, 2008; Vaill, 1996). This ability to continuously learn and adapt may be the single most important requirement to lead biotechnology companies to success. Without such a perspective, the leader and the organization will likely fail. To continuously learn, however, leaders must often suppress their own egos and relinquish control. Furthermore, leaders need to embody the story and the vision of the organization. Part of that vision should include creating a learning organization where all employees are encouraged and expected to learn continuously (Bennis & Nanus, 1997; Gardner, 1995; Senge, 1990; Vaill, 1996). Such a learning organization should support and give conﬁdence to employees, so that they will take on leadership roles within their own jobs (Wergin, 2007). A number of important attributes are required for establishing a learning organization that deal with the paradoxical nature of leading biotechnology companies in the 21st century. First, the leader should be a visionary manager who is able to consistently articulate his or her vision throughout the organization. Second, the leader needs to be a strategic decision-maker and be ﬂexible enough to allow the strategic vision to adjust to the culture and the environment. Third, the leader needs to be able to communicate effectively and create an organization where communication ﬂows efﬁciently at all levels. Such communication can be extremely difﬁcult in fast-growing organizations where effective communication is needed across cultural, geographic, or functional boundaries. Fourth, the leader needs to recognize that clear cultural differences exist between functional groups. The leader must not give in to the common temptation among both scientists and business people to downplay the importance of these differences. Within the organization, cultural differences need to be respected, whether they are between people from different countries or people with different functional backgrounds, such as science and business. Finally, organizational leaders need to empower their employees at all levels to make strategic decisions; but at the same time, the leader needs to know which decisions must be retained as his or her sole responsibility.

FUTURE RESEARCH This research was exploratory and examined the literature as it relates speciﬁcally to leadership and the success of biotechnology and biopharmaceutical organizations. Because the current literatures does not speciﬁcally,

76

LYNN JOHNSON LANGER

or directly, address the research question of how to successfully lead biotechnology and biopharmaceutical organizations, this literature review provides a framework for a basis of future research. Although a recent, exploratory study by Langer (2008) describes three exploratory case studies that address the speciﬁc question of how scientist founders lead successful biopharmaceutical organizations, additional exploratory qualitative research needs to be done. Exploratory research is needed to investigate successful, biotechnology and biopharmaceutical organizations that include companies at different stages of growth and success to help determine whether these ﬁndings are true for other biotechnology and biopharmaceutical companies. Questions arise about company size and type that may affect the transferability of these results. Additional research is needed to explore how the new organizations of PDPs are similar and different to other non-proﬁts and to for-proﬁt organizations. Also, research is needed that includes companies in different phases of organizational growth and companies that have failed. Furthermore, quantitative research is needed to help identify generalizable themes for biotechnology leaders and organizations.

REFERENCES Allen, T. J. (1984). Managing the ﬂow of technology: Technology transfer and the dissemination of technological information with the R&D organization (2nd ed.). Cambridge, MA: MIT Press. Andrews, F. M., & Pelz, D. C. (1976). Scientists in organizations: Productive climates for research and development (rev. ed.). Ann Arbor: Institute for Social Research, University of Michigan. Argote, L., & Ingram, P. (2000). Knowledge transfer: A basis for competitive advantage in ﬁrms. Organizational Behavior and Human Decision Processes, 82(1), 150–169. Bailyn, L. (1987). Experiencing technical work-a comparison of male and female engineers. Human Relations, 40(5), 299–312. Bennis, W. G., & Nanus, B. (1997). Leaders: Strategies for taking charge (2nd ed.). New York: Harper Business. Berson, Y., & Linton, J. (2005). An examination of the relationships between leadership style, quality, and employee satisfaction in R&D versus administrative environments. R&D Management, 35(1), 51–60. Bioplan Associates, Inc. (2003). Commercializing biotechnology and life sciences, a survey of education and training life sciences commercialization. Rockville, MD: Author. Biotechnology Industry Organization. (2007). Available at http://www.bio.org/speeches/pubs/ er/technology_collection.asp. Retrieved on June 1, 2007.

Leadership Strategies for Biotechnology Organizations

77

Brown, D. B. (1995). Technimanagement, the human side of the technical organization. Englewood Cliffs, NJ: Prentice-Hall. Cardinal, L. B. (2001). Technological innovation in the pharmaceutical industry: The use of organizational control in managing research and development. Organization Science, 12(1), 19–36. Cordero, R., Farris, G. F., & DiTomaso, N. (2004). Supervisors in R&D laboratories: Using technical, people, and administrative skills effectively. IEEE Transactions on Engineering Management, 51(1), 19–30. Covey, S. (1996). Three roles of the leader in the new paradigm. In: F. Hesselbein, M. Goldsmith & R. Beckhard (Eds), The leader of the future: New visions, strategies, and practices for the next era (pp. 149–159). San Francisco: Jossey-Bass. Day, S. (2003). Leadership practices of project scientists at the United States National Aeronautic and Space Administration, a dissertation. Unpublished doctoral dissertation, Fielding Institute, Santa Barbara, CA. Available at http://proquest.umi.com/pqdweb?did ¼ 765187601&sid ¼ 1&Fmt ¼ 6&clientId ¼ 5241&RQT ¼ 309&VName ¼ PQD. Retrieved on July 24, 2007. Dess, G. G., & Picken, J. C. (2005). Changing roles: Leadership in the 21st century. Available at http://search.epnet.com/login.aspx?direct ¼ true&db ¼ buh&an ¼ 2896274. Retrieved May 24, 2005. Drath, W. (2001). The deep blue sea rethinking the source of leadership. San Francisco: Jossey-Bass. Elkins, T., & Keller, R. T. (2003). Leadership in research and development organizations: A literature review and conceptual framework. Leadership Quarterly, 14(4–5), 587–606. Essex, L., & Kusy, M. (2004). Fast forward leadership. Lafayette, CO: Moonlight Publishing. Farkas, C. M., & De Backer, P. (1996). Maximum leadership. The world’s leading CEOs share their ﬁve strategies for success. New York: Henry Holt & Company. Federal Reserve Bank of Dallas. (2007, March/April). A conversation with Nancy Chang, taking the pulse of biotech, p. 9. Available at http://www.dallasfed.org/research/swe/2007/ swe0702e.pdf. Retrieved on June 6, 2007. Feldman, S. P. (1989). The idealization of technology: Power relations in an engineering department. Human Relations, 42(7), 575–592. Gardner, H. (1995). Leading minds: An anatomy of leadership. New York: Basic Books. Geller, E. S. (2002). Leadership to overcome resistance to change: It takes more than consequence control. Journal of Organizational Behavior Management, 22(3), 29–49. Gill, R. (2003). Change management–or change leadership? Journal of Change Management, 3(4), 307–318. Greenleaf, R. K. (1991). Servant leadership: A journey into the nature of legitimate power and greatness (1st ed.). Mahwah, NJ: Paulist Press. Greiner, L. (1998). Evolution and revolution as organizations grow. Harvard Business Review, May–June, pp. 55–66. Handy, C. (1995). Gods of management the changing work of organizations (4th ed.). New York: Oxford University Press. Heifetz, R. A. (1994). Leadership without easy answers. Cambridge, MA: Belknap Press of Harvard University Press. Heifetz, R. A., & Laurie, D. L. (2001). The work of leadership. Harvard Business Review, 79(11), 131. Hemp, P., & Stewart, T. A. (2004). Leading change when business is good. Harvard Business Review, 82(12), 61.

78

LYNN JOHNSON LANGER

Hennessey, B. A., & Amabile, T. M. (1988). Reward, intrinsic motivation, and creativity. American Psychologist, 53, 674–675. Hirst, G., & Mann, L. (2004). A model of R&D leadership and team communication: The relationship with project performance. R&D Management, 34(2), 147–160. Hoegl, M., & Gemeuden, H. G. (2001). Teamwork quality and the success of innovative projects: A theoretical concept and empirical review. Organization Science, 12(4), 435–449. Jabri, M. M. (1992). Job satisfaction and job performance among R&D scientists: The moderating inﬂuence of perceived appropriateness of task allocation decisions. Australian Journal of Psychology, 44(2), 95–99. Jordan, G. B. (2005). What matters to R&D workers. Research Technology Management (May–June), 23–32. Judge, W. Q., Fryxell, G. E., & Dooley, R. S. (1997). The new task of R&D management: Creating goal-directed communities for innovation. California Management Review, 39(3), 72–85. Kanter, R. M. (2004). Conﬁdence. New York: Crown Business. Keller, R. T. (2006). Transformational leadership, initiating structure, and substitutes for leadership: A longitudinal study of research and development project team performance. Journal of Applied Psychology, 91(1), 202–210. Kellerman, B. (2004). Bad leadership. Boston: Harvard Business School. Kivimaki, M., Lansisalmi, H., Elovainio, M., Heikkila, A., Lindstrom, K., Harisalo, R., Sipila, K., & Puolimatka, L. (2000). Communication as a determinant of organizational innovation. R&D Management, 30(1), 33–42. Kotter, J. P. (1996). Leading change. Boston: Harvard Business School Press. Kotter, J. P. (1999). John P. Kotter on what leaders really do. Boston: Harvard Business School Press. Kouzes, J. M., & Posner, B. Z. (2002). The leadership challenge (3rd ed.). San Francisco: JosseyBass. Langer, L. (2008). How scientist/founders lead successful biopharmaceutical organizations: A Study of three companies. Retrieved on October 21, 2008, from Ohiolink Dissertation database. Langer, L. J. (2007). Moving ideas from the laboratory to the marketplace: How scientists and business leaders engage to take action. In: K. H. Cohn & D. E. Hough (Eds), The business of healthcare (pp. 209–224). Westport, CT: Praeger Publishers. McGregor, D. M. (1957). The human side of enterprise. Management Review, 46(11), 22–28. Mezirow, J. (1991). Transformative dimensions of adult learning. San Francisco: Jossey-Bass. Miller, D. (2002). Successful change leaders: What makes them? What do they do that is different? Journal of Change Management, 2(4), 359. Mullin, R. F., & Sherman, R. (1993). Creativity and performance appraisal: Shall never the twin meet. Creativity Journal, 6, 425–434. Nonaka, I. (1995). A dynamic theory of organizational knowledge creation. Organization Science, 5(1), 14–37. O’Toole, J. (1995). Leading change: Overcoming the ideology of comfort and the tyranny of custom (1st ed.). San Francisco: Jossey-Bass. Owen-Smith, J. (2001). Managing laboratory work through skepticism: Processes of evaluation and control. American Sociological Review, 66(3), 427–452.

Leadership Strategies for Biotechnology Organizations

79

Pascale, R. T., & Sternin, J. (2005). Your company’s secret change agents. Harvard Business Review, 83(5), 73. Pirola-Merlo, A., Ha¨rtel, C., Mann, L., & Hirst, G. (2002). How leaders inﬂuence the impact of affective events on team climate and performance in R&D teams. Leadership Quarterly, 13(5), 561–581. Pisano, G. (2006). Science business: The promise, the reality, and the future of biotech. Boston: Harvard Business School Press. Priestland, A., & Hanig, R. (2005). Developing ﬁrst-level leaders. Harvard Business Review, 83(6), 113. Rogers, E. (2003). Diffusion of innovation. New York: Simon & Schuster. Rooke, D., & Torbert, W. (2005). Transformations of leadership. Harvard Business Review (April), 67. Ruvolo, C. M., & Bullis, R. (2003). Essentials of culture change: Lessons learned the hard way. Consulting Psychology Journal: Practice and Research, 55(3), 155–168. Sapienza, A. M. (2005). From the inside: Scientists’ own experience of good (and bad) management. R&D Management, 35(5), 473–482. Scho¨n, D. (1983). The reﬂective practitioner, How professionals think in action. New York: Basic Books. Senge, P. (1990). The ﬁfth discipline. New York: Doubleday. Shim, D., & Lee, M. (2001). Upward inﬂuence styles of R&D project leaders. IEEE Transactions on Engineering Management, 48(4), 394. Skipton, L. H. (2003). Leadership development for the postindustrial, postmodern information age. Consulting Psychology Journal: Practice and Research, 55(1), 3–14. Sosik, J. J., Kahai, S. S., & Avolio, B. J. (1999). Leadership style, anonymity, and creativity in group decision support systems: The mediating role of optimal ﬂow. Journal of Creative Behavior, 33, 227–256. Stanford, Graduate School of Business. (n.d.). Biotech innovators and investors assess challenges, opportunities. Available at http://www.gsb.stanford.edu/NEWS/headlines/biotechnology. shtml. Retrieved on June 6, 2007. Stoker, J. I., Looise, J. C., Fisscher, O. A. M., & de Jong, R. D. (2001). Leadership and innovation: Relations between leadership, individual characteristics and the functioning of R&D teams. International Journal of Human Resource Management, 12(7), 1141–1151. Tampoe, M. (1993). Motivating knowledge workers-the challenge for the 1990s. Long Range Planning, 26(3), 49–55. Tidd, J., Bessant, J., & Pavitt, K. (2005). Managing innovation, integrating technological, market and organizational change. West Sussex, UK: Wiley. Trevelyan, R. (2001). The paradox of autonomy: A case of academic research scientists. Human Relations, 54(4), 495–525. Tufts E-News. (2007). The ballooning pricetag. Available at http://www.tufts.edu/communications/ stories/120401BallooningCosts.htm. Retrieved on February 2, 2007. United States Food and Drug Administration. (n.d.). The new drug development process: Steps from test tube to new drug application review. Available at http://www.fda.gov/cder/ handbook/develop.htm. Retrieved on June 3, 2007. Vaill, P. B. (1996). Learning as a way of being. San Francisco: Jossey-Boss. Wainer, H. A., & Rubin, I. M. (1969). Motivation of research and development entrepreneurs: Determinants of company success. Journal of Applied Psychology, 53(3), 178–184.

80

LYNN JOHNSON LANGER

Waldman, D. A., & Atwater, L. E. (1994). The nature of effective leadership and championing processes at different levels in R&D hierarchy. Journal of High Technology Management Research, 5(2), 233–245. Wergin, J. (Ed.) (2007). Leadership in place: How academic professionals can ﬁnd their leadership voice. Bolton, MA: Anker. Wheatley, M. J. (1999). Leadership and the new science: Discovering order in a chaotic world (2nd ed.). San Francisco: Berrett-Koehler. Yin, R. K., & Gwaltney, M. K. (1981). Knowledge utilization as a networking process. Knowledge: Creation, Diffusion, Utilization, 2, 555–580. Zhang, J., & Patel, N. (2005) The dynamics of California’s biotechnology industry. Available at http://www.ppic.org/content/pubs/report/R_405JZR.pdf. Retrieved on May 29, 2007.

ATTRIBUTION THEORY AND HEALTHCARE CULTURE: TRANSLATIONAL MANAGEMENT SCIENCE CONTRIBUTES A FRAMEWORK TO IDENTIFY THE ETIOLOGY OF PUNITIVE CLINICAL ENVIRONMENTS Patrick A. Palmieri and Lori T. Peterson ABSTRACT The Institute of Medicine’s seminal report, To err is human: Building a safer health system, established the national patient safety framework and initiated interest in changing the traditionally punitive healthcare culture. This paper reviews a multidisciplinary literature and offers an attribution framework to explicate the organizational processes that contribute to an industry-wide culture where clinicians are routinely blamed for adverse patient events. Attribution theory is concerned with the manner in which people explain the behaviors of others or themselves by assigning causality for events. To date, attribution theory, though well established in the management literature, has yet to be translated to Biennial Review of Health Care Management: Meso Perspectives Advances in Health Care Management, Volume 8, 81–111 Copyright r 2009 by Emerald Group Publishing Limited All rights of reproduction in any form reserved ISSN: 1474-8231/doi:10.1108/S1474-8231(2009)0000008008

81

82

PATRICK A. PALMIERI AND LORI T. PETERSON

healthcare. In this paper, we ﬁrst describe the historical evolution of attribution theory in relation to human behavior in clinical practice and healthcare management and then discuss the work environments in contemporary healthcare organizations. Next, we demonstrate the applicability of attribution theory to healthcare by providing two adverse event exemplar cases. Then, the Healthcare Attribution Error Model is offered to demonstrate how concepts from attribution theory serve as antecedents to the employee cynicism, learned helplessness, organizational inertia, and the emerging Just Culture perspective. We conclude by suggesting attribution theory offers an important theoretical framework that warrants further conceptual development and empirical research. In the quest to produce exceptional healthcare environments where safety and quality are fundamental employee concerns, healthcare managers and clinical professionals need theoretically supported knowledge and evidence-based insights.

INTRODUCTION The release of the Institute of Medicine’s (IOM) seminal report, To err is human: Building a safer health system (Kohn, Corrigan, & Donaldson, 2000), stimulated national interest in understanding the traditional punitive healthcare culture (Ruchlin, Subbs, & Callahan, 2004). Healthcare is a system that frequently harms (Davis et al., 2002) and routinely fails to deliver the appropriate standard of care (IOM, 2004). Recently, the World Health Organization (2008) proclaimed poor safety and suboptimal care is endemic, citing 10% of all patients are impacted by medical errors. Yet, the pressures that slowly force healthcare organizations to identify and reduce their propensity for adverse events (Berta & Baker, 2004) have been ineffective in creating cultures of safety (Page, 2004). Poorly designed systems, not the actions of well-intended clinicians, are responsible for adverse events (IOM, 2001; Reason, 2000). Attribution theory provides a cognitive rationale to explicate the organizational processes that contribute to an industry-wide culture where clinicians are frequently blamed for adverse patient events. To date, this traditional management theory has not been translated to healthcare to offer a theoretically anchored explanation for the etiology of punitive hospital cultures. Attribution theory is concerned with the manner in which people explain the behaviors of others or themselves by assigning causality for events

Attribution Theory

83

(Heider, 1958). In an organizational context, attributions are rapid perceptions about situations or events. These perceptions often lead to attributing events to personal dispositions and intentions rather than to account for other plausible explanations (Jones, 1979). The causal assignment rapidly explains situations but impacts the usefulness and the accuracy of the appraisal. Inaccurate or incomplete causal associations can lead to serious organizational consequences such as employee distrust (Tucker & Edmondson, 2003). These ﬂawed appraisals have caused healthcare safety improvement efforts to be slow and incremental (Devers, Pham, & Lui, 2004), as error reporting systems remain inadequate and poorly supported by clinicians (Edmondson, 2004; Reason, Carthey, & de Leval, 2001; Tucker, 2004). Consequently, substantial and unrealized deﬁcits in organizational knowledge (Devers et al., 2004; Gray, 2001) about system failures (Rasmussen, 1999) are perpetuated by misplaced attributions (Reason, 2000).

Healthcare Organizations and Theoretical Knowledge Application Despite calls for comprehensive system reforms, Millenson (2003) criticized the continued focus on improvement efforts that ignores the repeated reluctance of healthcare leaders to modify systems and to avoid personal attributions for failures. Dysfunctional organizational behavior has an extensive and rich history in the management literature (Argyris, 1957; March & Simon, 1993); however, there has been modest application to the behaviors present in complex healthcare organizations (Kohn et al., 2000). In healthcare, this situation parallels the ‘‘display culture’’ where British Naval ofﬁcers ordered watertight doors on warships polished until they were no longer watertight (Reason, 1998). The polishing was not intended to cause leaky doors, but door shinning provided the appearance of a wellmaintained ship, therefore, a safe ship. Unfortunately, leaky doors were an unanticipated consequence, much like the distrust caused by misplaced attributions to hold clinicians ‘‘accountable’’ for their mistakes. Healthcare leaders continue to shine hospital cultures, but the system remains just as leaky today as in years past (Devers et al., 2004; Leape & Berwick, 2005). Continuing the analogy, the current healthcare polishing activities should be discontinued in favor of translating theoretical and empirical management science to clinical healthcare operations. We examine attribution theory in the context of hospital work environments. Physicians, nurses, pharmacists, and others experience blame and even suffer serious consequences for mistakes due to management’s

84

PATRICK A. PALMIERI AND LORI T. PETERSON

perception of events. While this paper is not intended to be an ‘‘us versus them’’ examination of attribution, the discussion and the case examples emphasize the clinician perspective. In many hospital organizations, physicians are not direct employees of the organization, but the nursing and ancillary staff are directly employed. As such, clinicians are commonly subjected to misattributions when it comes to errors and adverse events that occur while delivering patient care (Page, 2004; Reason, 2000). In this paper, we provide an overview of attribution theory. Then, we discuss attribution theory as a natural human behavior, noting its inﬂuence on healthcare managers and clinicians. To support our theoretical translation and application of attribution theory to healthcare, we then discuss two recent sentinel event cases where attribution processes led to different causal perspectives. Next, we offer the Healthcare Attribution Error Model and apply attribution theory constructs to healthcare work environments. Finally, we consider the emerging concept of Just Culture and conclude with a brief discussion about the future implications for attribution theory in healthcare. This paper contributes to the literature by demonstrating attribution theory is relevant to healthcare and by illustrating the speciﬁc organizational processes that sustain the traditionally punitive industry culture.

ATTRIBUTION THEORY From psychology and sociology, attribution theory emerged in the management literature as a characteristic of organizational behavior by which individual assessments are formed. The theory explains the process whereby people seek to understand the cause of a situation (Heider, 1958), assess responsibility for the outcome (Martinko & Gardner, 1982), and appraise the personal attributes of the people involved (Weiner, 1995a). Furthermore, attribution is synonymous with explanation; placing the discussion of ‘‘why did something occur’’ in the context of leader–member relations (Green & Mitchell, 1979). As healthcare is delivered in a dynamic and complex environment (McDaniel & Driebe, 2001), managers may unknowingly engage in ‘‘attribution-like’’ behaviors to make sense of multifarious situations especially when time is limited (Kelley, 1972). In many cases, these attributions are not purposeful, nor are they intended to misrepresent; however, attributions often lead to judgments about a person or situation. This is similar to the analogy about judging a book by the appearance of its cover. Healthcare managers, like any person, seek to explain their behaviors and the behaviors of others; attribution theory explicates how people form

85

Attribution Theory

judgments (Heider, 1958). These causal associations, when focused on clinician behavior, have serious organizational consequences in the healthcare industry’s approach to address performance fallibility (Edmondson, 2004). For example, increased exposure for adverse patient events and the inclination to attribute such failures to clinicians engenders an adversarial organizational safety climate in many healthcare organizations (IOM, 2001, 2004). Attribution theory provides a framework to identify mechanisms that contribute to blame subsequent to non-routine harmful patient events. Whether cognitively assembled or formally conducted, adverse events are followed by both immediate and planned investigations to identify causality. As conclusions about event causality develop (Martinko, 1995), two attribution types emerge, internal and external (Weiner, 1985). Internal attribution is the causality factor that falls within an individual’s control, also known as dispositional attribution (Heider, 1958). External attribution deﬁnes the factors residing outside an individual’s control, also known as situational attribution (Kelley, 1967). We will refer to internal attributions as dispositional and external attributions as situational. Both dispositional and situational attributions are described in the literature as the silent hands that guide sensemaking (Weick, 1995). Sensemaking depends on the availability of information, but more importantly, speed yields an absence of critical information. In such cases, fundamental attribution error (Ross & Nisbett, 1991) often results. Fundamental attribution error (Fig. 1) explains how perceptual tendencies, rather than situational or contextual, result in ﬂawed conclusions (Weiner, 1995b) about event causality. To illustrate this point, we provide a real example of attributions from an operating room adverse event (Box 1). This example provides a basic examination of factors that contribute to the complexity surrounding a sentinel event, and how easy it is for outsiders

Fig. 1.

Fundamantal Attribution Error.

86

PATRICK A. PALMIERI AND LORI T. PETERSON

Box 1. A neurosurgeon mistakenly injects alcohol into a patient’s spinal column instead of spinal ﬂuid at the end of a surgical procedure. The surgeon is immediately blamed by most people, including management, for this outrageous and critical error. Here, the surgeon encounters dispositional attributions of poor performance, perhaps incompetence, since the error was within his immediate control. Then, additional information surfaces as the investigation continues indicating the alcohol syringe was actually prepared by the circulating nurse and drawn from an unlabeled container in the surgical ﬁeld. The nurse then handed the syringe to the surgeon. Now, the event may be viewed as removed from surgeon’s control and the nurse becomes the culprit (situational attribution to the surgeon, dispositional attribution to the nurse). Next, an investigator learns the spinal ﬂuid container was replaced with the alcohol by a surgical technician in his preparation to end the procedure, unbeknownst to the nurse or the surgeon. At this point, the technician would be blamed as the incompetent person responsible for this event. As the investigation concludes, we learn the surgical technician is a new graduate, has only trained in gynecological surgery, and remains on orientation, ‘‘approved’’ for certain gynecological procedures. The technician was assigned to the spinal case to ﬁll a vacancy as directed by the operating room manager. Now, the operating room manager is also implicated in the event.

to assign rash attributions to people involved in the incident. As demonstrated in this example and Fig. 1, situational and dispositional attributions depend on information availability and accuracy, time constraints and explanation urgency, and the person responsible for making the causal associations. More speciﬁcally, fundamental attribution error is probable when time constraints, often created by adverse patient events, lead to incomplete knowledge about event circumstances. These concepts will be revisited and further developed in the sections to follow. Attribution in Healthcare Attribution theory is derived from a relatively parsimonious multidisciplinary literature (Heider, 1958; Kelley, 1967; Martinko, 1995). Cognitively,

Attribution Theory

87

humans automatically respond to explain the behavior of others by trying to construct rationales for these behaviors without much consideration of situational inﬂuences. More recently, attribution has been shown to have a biological foundation. Neuroimaging provides evidence that attribution is an organic processes that can be predictably mapped in human brains (Harris, Todorov, & Fiske, 2005) and attribution placement, self versus others, shifts with damage to the brain (e.g., strokes in the frontal region). These recent medical ﬁndings deserve additional attention as clinicians and managers could be predisposed to speciﬁc types of attributions. Using ‘‘problemistic search’’ (Cyert & March, 1963, p. 120), managers scan their environments seeking rapid solutions for urgent problems such as the cause of an adverse event. This cognitive strategy seeks solutions that reﬂect casual schemas, or assumed patterns of behavior, usually forced by superﬁcial inquiry and time constraints (Martinko & Gardner, 1987). The process to develop a perception is sought through an evaluation of undertaken action. Collectively, managers of organizational processes are often disinterested in isolating perfect decisions when problems are illdeﬁned and difﬁcult to solve. As such, attributing situations to ‘‘others’’ in a perceived causation, at the conclusion of complex events, produces relatively quick, easy, and efﬁcient solutions. This process is historically preferred and often advocated in healthcare management (Reason, 2000). Concepts Attribution simpliﬁes event causality (DeJoy, 1999; Heider, 1958; Kelley, 1967; Weiner, 1986) for people processing information and making assessments about observed situations or outcomes (Weiner, 1985) as they attempt to understand their environment (Berger & Luckmann, 1967; Standing, Guilfoyle, Lin, & Love, 2006). Therefore, attribution is a cognitive instrument healthcare professionals use to make sense of the world around them and the interactions that engage them (DeJoy, 1999). When managers encounter unexpected events, circumstances are processed to construct a mental image that normalizes the event (Weick & Sutcliffe, 2001). This process of sensemaking is an inherent tendency (Weick, 1995), not necessarily reﬂective of objective reality (Heider, 1958). Explained by attribution theory, the antecedents information, beliefs, and motivations lead to causal attributions (DeJoy, 1985; Weiner, 1985). These factors contribute to inferences consequential to people (Green & Mitchell, 1979) in directing or changing behavior, affect, and expectancy (DeJoy, 1999). As described in the literature, the dimensions causality, controllability, intentionality, stability, and globality represent the major attribution

88

PATRICK A. PALMIERI AND LORI T. PETERSON

Table 1. Dimension

Dimensions (Locus) for Attributions. Deﬁnition

Example

Causality

The personal internality or external relationships that shapes event causation (Weiner, 1985)

Controllability

The degree to which a person believes they or another entity is capable of controlling the causality of an outcome (Weiner, 1995a) The perception of intention as the cause of an event (Langdrige & Butt, 2004)

The nurse was derelict in her performance, whereas, the pharmacist was sure unlucky in that situation The physician should have given the patient more blood, whereas, the blood bank was completely out of the patient’s blood type

Intentionality

Stability

The context of time and situation as causative variables (Kent & Martinko, 1995)

Globality

The distinct cultural and social contextual variations in which events or outcomes occur (Langdrige & Butt, 2004)

The physician euthanized the patient, whereas, the patient died during surgery as the surgeon accidently nicked the aorta Personality (e.g., bedside manner) is considered a stable causation, whereas, effort (e.g., hard work) varies and is dependant on time/ situation Cultures have distinctively internal or external attribution when compared against each other (e.g., Hofstede & Hofstede, 2005; House, Hanges, Javidan, Dorfman, & Gupta, 2004)

concepts that provide structural and functional ﬁtness to the theory (DeJoy, 1985; Weiner, 1985). The manifestation of speciﬁc dimensions, summarized in Table 1, contribute the description of relationships that link the humanenvironment-healthcare process through schematic interfaces. A thorough discussion about these dimensions is beyond the scope of this paper. Attribution Assignment for Adverse Events and Clinical Errors Dispositional Attribution Dispositional attribution is a cognitively associated state where feelings of human fallibility manifest in practice. Decreased self-efﬁcacy can cause clinicians to feel culpable for unfortunate situations that often do not reside within their immediate control or inﬂuence. In the organizational context related to patient safety, the term ‘‘dispositional attribution’’ resembles the

Attribution Theory

89

person approach to error causation (Palmieri, Delucia, Peterson, Ott, & Green, 2008; Reason, 1997). Internal attribution in the case of error generally results from organizational blame and punishment ensuing from investigations that attempt to unearth ‘‘people’’ as the root cause of an adverse event (Reason, 1997) versus objectively examining all plausible causes. That is, this approach focuses on ‘‘Who did something?’’ (Rasmussen, 1999) versus ‘‘Why did something happen?’’ (Reason, 2000). Causality is an important consideration in the assignment of attribution (Kent & Martinko, 1995; Weiner, 1995a) much like being presumed innocent until proven guilty. Similar to breakdowns in due process, appropriate causality is not always properly distributed (Runciman, Merry, & Tito, 2003), due to ﬂawed attribution schemas. This process will become more evident with the case examples later in the paper. Martinko and Gardner (1987) indicate attribution processes are moderated by biases and individual differences. This particular point is readily experienced by most clinicians, especially nurses (Page, 2004) working in healthcare. Blame in healthcare is derived from an attribution-like process (Reason, 1997; Sasou & Reason, 1999) as managers perceive lack of effort (Reason, 2000), inability (Vincent, 2003), incompetence (Rasmussen, 1999), and absent vigilance (Reason & Hobbs, 2003) as explanations for clinician error. As such, dispositional attributions result in disciplinary actions (Martinko & Gardner, 1987). This person-centered approach most often results from high-visibility errors; those errors manifesting as patient harm (Kohn et al., 2000; Rasmussen, 1999; Reason, 2000). Perceived employee performance inﬂuences management attributions (Martinko & Gardner, 1987). As such, employee performance is modulated by manager attributions as evidenced by the concealment of practice mistakes. Accordingly, clinician reluctance to report errors is quite common (IOM, 2007; Kohn et al., 2000). This failure to report occurs as clinicians understand that ‘‘blaming individuals is emotionally more satisfying than targeting institutions’’ (Reason, 2000, p. 70), where people are seen as straightforward causative agents. Therefore, clinicians realize that admissions of fallibility or error result in reprimands instead of appreciation. Mistakes leading to undesirable outcomes are not usually related to incompetent or substandard care (e.g., Cook & Woods, 1994), a frequently misunderstood and unappreciated fact (Page, 2004; Reason, 2000). Rather, errors often reﬂect a clinician’s inability to cope with gaps produced by poorly designed (Reason, 2000) and exceedingly complex systems (Woods & Cook, 2002). Martinko and Gardner (1987, p. 240) stated, ‘‘behaviors that yield adverse consequences may lead to unfavorable

90

PATRICK A. PALMIERI AND LORI T. PETERSON

inferences,’’ linking human error guided by ﬂawed processes to people instead of system characteristics and inﬂuences. In addition, Bandura (1999) described the typical management role as a detrimental schema where managers deliberately remain uninformed by utilizing indirect and ambiguous practice consent. This concept is supported by Reason and colleagues (2001) as they explain how managers avoid controllable system problems. The management approval concept is similar to the attribution of stable and unstable causes (Martinko & Gardner, 1987). When errors arise, managers usually focus on subordinate performance while subordinates, such as pharmacists and nurses, focus on environmental elements (Bernardin, 1989; Martinko & Gardner, 1987; Mitchell & Woods, 1980). For example, when a medication arrives on the nursing unit labeled with the wrong patient name, the nurse manager might instantly attribute this issue to poor pharmacist practice. However, the pharmacist might see the unreliable labeling machine and frequent understafﬁng in the pharmacy to be the causative agents. Martinko and Gardner (1987, p. 239) realize ‘‘the subordinate attribution process is complex and interactive.’’ Later, Martinko (1995, 2004) discusses how attribution can be confusing when considering the variety of manners in which the concept can ﬂow. This ﬂow could also be different between industries or professions. We recognize the complex and interactive nature of attribution in our healthcare discussion. Martinko’s point is especially important because attribution theory and research matured in the organization behavior and management literature and the knowledge has not been transferred to healthcare. As such, it is critical for the consistency of attribution theory application, in both the language and in its logical ﬂow, to provide a sensible approach for viewing management attributions. In addition, the clinician attributions related to their own practice and performance are important considerations. For example, the experienced healthcare professional may be more likely to attribute error to the less-experienced individual, but the less-experienced individual may attribute error to the external causes in the environment. This incongruence in belief structures presents an opportunity for health services researchers to study. Situational Attribution Situational attribution describes the causes of behavior credited to the direct inﬂuences of situation factors. These attributions resemble the systems approach to considering errors (Palmieri et al., 2008). In the hospital industry, clinicians tend to place blame on the environment, and even management, when issues or adverse events arise. The typical clinician,

Attribution Theory

91

similar to other professionals described in the literature (Martinko & Gardner, 1987), usually favor situational attributions to the organization (Page, 2004), whereas managers often blame clinician incompetency (Reason et al., 2001). Of course, not all clinicians have an adversarial relationship with management. Supportive and constructive clinical relationships with management (or when clinicians are managers) can contribute to positive work environments; yet, these institutions represent the minority in the industry (IOM, 2004; Kohn et al., 2000; Page, 2004). Experiential knowledge inﬂuences the likelihood that clinicians will make self-serving, external attributions despite evidence of poor performance. By applying empirical evidence from sports, personal awareness derived from experience motivates individuals to improve their performance. Roesch and Amirkham (1997) reported top athletes deal better with genuine causes of poor performance and, over time, tend to improve their overall performance by making fewer self-serving external attributions. However, these athletes are not performing in punitive environments and are motivated through reward systems to improve their performance. This is an important distinction for healthcare managers to consider. Following a sincere evaluation and discussion about performance, replacing the risk of personal harm with the promise of organizational learning, clinicians ought to be able to discern where environmental and personal improvement efforts should be focused. When managers want to understand adverse event causality, hidden system features, or latent factors (Dorner, 1996; Perrow, 1984; Reason, 2000), they must act with a purposeful focus to uncover facts and not perceptions (Reason, 1990; Dekker, 2007). Most managers realize no system can completely eliminate human error (Perrow, 1984; Rasmussen, 1990), but extending their perspective to action will lead to routine examinations of situational causes – an instrumental philosophical shift. This change is essential to reduce the tendency for dispositional blame and to increase organizational learning. Situational attributions result from a deep understanding about care delivery systems (Kohn et al., 2000) and the realization that the best clinicians make the worst mistakes (Reason, 2000). Therefore, reducing risk (Reason, 1998), increasing reliability (Flin, Burns, Mearns, Yule, & Robertson, 2006), and improving quality (Kohn et al., 2000; Reason, 2000; Smetzer & Cohen, 1998) are associated with careful situational appraisals (Weick, Sutcliffe, & Obstfeld, 1999), not immediate dispositional attributions (Rasmussen, 1990). Root Cause Analysis Situational attributions often result from deliberate assessments similar to the event speciﬁc root cause analysis (RCA). The RCA is an investigatory

92

PATRICK A. PALMIERI AND LORI T. PETERSON

mechanism adopted by healthcare organizations to uncover the situation ‘‘why’’ and to determine the root cause ‘‘how’’ for an adverse event (Battles, Kaplan, van der Schaaf, & Shea, 1998; IOM, 2007; Joint Commission Resources, 2007). Both attributions and RCAs help managers recognize the contributory factors that culminate in untoward events. However, attribution and RCA can place emphasis on vastly different outcomes. While attribution is an immediate appraisal that often assigns personal accountability (dispositional), the RCA seeks causality within system properties and processes (situational) while avoiding individual blame (Gano, 2007; Reising & Portwood, 2007). The RCA process emphasizes the system, or situational, approach to viewing events, while management attributions usually focus on the dispositional or person approach (Palmieri et al., 2008). Time-constrained management attributions occur most frequently in healthcare since RCA is a time-consuming process, undertaken with many participants and multiple meetings to examine facts and consider possible scenarios. Indeed, the assignment of personal responsibility for adverse events often results from purposeful attributions to individuals directly linked to the event (Weiner, 1985), not necessarily the process that led to the poor outcome (Reason, 1997; Reason, 2000). Altering the ‘‘name and shame’’ mind-set of many healthcare managers is challenging (Runciman et al., 2003) despite their understanding about the important contributions RCA provide to process improvement efforts. Next, we move to the exemplar cases, followed by the presentation of the Healthcare Attribution Error Model (Fig. 2).

EXEMPLAR CASES Recognizing healthcare environment complexity, there are a number of single-event attributions (including slips, lapses, mistakes, and error) that can contribute to management attributions and to employee responses. The associated emergent system properties, cynicism, or improved employee conﬁdence reside in organizational culture, such as organizational inertia or Just Culture. These properties might also include external inﬂuences such as the licensing boards that govern clinical practices within the organization. The examples to follow include inﬂuences external to the organization; however, the focus for this discussion is on how the organizations handled the events. The next two cases exemplify dispositional and situational attribution responses to publically reported clinical errors that resulted in patient deaths. Both cases involve similar mistakes that resulted in patient harm;

93

Attribution Theory

Fig. 2.

The Healthcare Attribution Error Model.

however, the outcomes for the involved clinicians are remarkably dissimilar. The ﬁrst case resulted in one death and immediate public dispositional attributions that ended with a criminal prosecution. The second case involved multiple infant deaths; the catalyst for dramatic hospital improvement as dispositional attributions were withheld, situational factors were studied, and organizational learning resulted.

Dispositional Attribution Case Dispositional attributions in response to an adverse patient event were exempliﬁed at a Midwestern hospital. An experienced labor and delivery nurse mistakenly administered an epidural drug to a young late-term pregnant patient through an intravenous line (Institute for Safe Medication Practices, 2006). Unfortunately, the patient died as a direct result of this fatal medication administration error. With closer analysis, however, the

94

PATRICK A. PALMIERI AND LORI T. PETERSON

facts revealed the fatal error was only one in a cascade of events in a system plagued with ‘‘hidden’’ or contributory errors. Typically observed in medication delivery systems, errors originate proximal to system level and cascade through the process distally to clinical practice (Smetzer & Cohen, 2006). In this case, causality was immediately and publically reported as an incompetent nurse derelict in her duties. Speciﬁcally, her duty to make certain the right patient, received the right dose, of the right drug, at the right time, and by the right route – the ﬁve rights of nursing mediation administration – was the focus. Management was paralyzed by the publically reported sentinel event and was unable to recognize that the nursing error resulted from interactions within a complex adaptive system with unfavorable and dormant conditions. Following the event, the state health agency cited the hospital for a number of system deﬁciencies. In this case, the nurse was working a double shift, the pharmacy did not secure the medication on the unit per policy, and management was not appropriately monitoring the effectiveness of the new automated medication veriﬁcation system. Dispositional attributions combined with local and national media attention subjected the nurse to signiﬁcant public blame, which created humiliation and hardship. Multifaceted dispositional attributions caused the nurse to face serious punishment despite the presence of other signiﬁcant contributory factors (Institute for Safe Medication Practices, 2006). In the end, purposed to protect the public from future harm, criminal prosecution was the chosen response to the event (Institute for Safe Medication Practices, 2006; State of Wisconsin, 2006). Numerous professional and quality-improvement organizations reacted with position statements objecting to this ‘‘miscarriage of justice’’ (Institute for Safe Medication Practices, 2006; Wisconsin Hospital Association, 2006; Wisconsin Medical Society, 2006) as the actions taken against the nurse were not in the spirit set forth by To Err is Human (Institute for Safe Medication Practices, 2006; Kohn et al., 2000; Wisconsin Medical Society, 2006). Cases such as this one are familiar to most clinical professionals (IOM, 2003) as the punitive healthcare industry is perpetuated by aggressive dispositional attributions (Kohn et al., 2000; Reason, 2000) despite continued calls for reform (IOM, 2007; Page, 2004).

Situational Attribution Case Deliberate and methodical management consideration, when facing critical organizational failures where patients are seriously harmed, is likely to result

Attribution Theory

95

in strong situational attributions with minimal dispositional types (Joint Commission Resources, 2005). At a Southeastern hospital, ﬁve nurses committed the same medication error causing three infant deaths and three other infant injuries. The hospital found ‘‘human and procedural errors account for the administration of inappropriate doses of Heparin to six infants in the Methodist NICU’’ (Clarian Health Partners, 2006). This case differs from the previous case example, as management made a closer analysis, before immediately attributing blame to the nurses, and determined the fatal error was merely one in a cascade of events from a system plagued with contributory factors. Actor Dennis Quaid’s experience began in 2007 when his then newborn twins encountered near death (Institute for Safe Medication Practices, 2008), as nurses mistakenly administered heparin (‘‘blood thinner’’) in an adult dose (10,000 units) instead of infant dose (100 units) to prematurely born infants in a neonatal intensive care unit (NICU). The 2006 Southeastern hospital event was similar in root cause. Both of these errors resulted from pharmacy issues and nursing issues. The pharmacy technician restocked the automated medication dispensing unit with the adult dose of heparin instead of the infant dose of heparin. The nursing issues surrounded the experiences of the ﬁve nurses who varied from routine protocols for medication veriﬁcation and administration due to both situational and human factors. A critical issue implicated in the infant deaths is the nearly indiscernible difference between the labeling (virtually identical color and lettering) on the adult and infant heparin vials (Institute for Safe Medication Practices, 2006). Important to this discussion, the literature has numerous reports about uniform heparin vial appearance contributing to patient harm (Institute for Safe Medication Practices, 2008, 2009); yet, the Food and Drug Administration and the manufacturer were unresponsive in correcting this potential sentinel event opportunity. Furthermore, multiple hospital policies, procedures, and practices were not adequate to address this problem due to operational impediments. In this case, and even the previous example, the nurses’ actions simply continued a cascade of errors, not only the manifestation of a single but lethal error (Shalo, 2007). Hospital management recognized the event could have been prevented at a number of critical points and implemented processes and training to address these concerns. For example, pharmacy operations for reﬁlling automated dispensing cabinets were overhauled, medication policies and procedures were reviewed and improved, and a mass re-education process for medication delivery was provided. Speciﬁcally, the ‘‘ﬁve rights’’ of drug administration was the focus for educational efforts,

96

PATRICK A. PALMIERI AND LORI T. PETERSON

instead of the basis for a criminal complaint (as noted in the previous example). In summary, the Southeastern hospital management attributed the event to be one where ‘‘an institutional error – one in which our procedures failed’’ was the causal agent (Clarian Health Partners, 2006). Management responses similar to this case facilitate the movement of punitive cultures to Just Cultures (Dekker, 2007); a concept addressed later in the paper.

HEALTHCARE ATTRIBUTION ERROR MODEL In this section, we present the Healthcare Attribution Error Model (Fig. 2) to support the application of attribution theory to healthcare. The Healthcare Attribution Error Model is the ﬁrst theoretically derived framework proposed for healthcare. Building upon the exemplar cases, we discuss learned helplessness, clinician cynicism, and organizational inertia as consequential outcomes in cultures where managers prefer dispositional attributions to explain their hospital environment. Just Culture is presented as an emerging concept as some organizations begin to focus on reducing dispositional attributions for clinical errors; preferring to focus on the system-related problems. Although we realize attribution is not a linear process, the representation of three-dimensional concepts in a two-dimensional diagram is challenging. In this model, all clinical practices can result in attributions (positive and negative); however, we are most concerned with those clinical circumstances that result in failures (the most prevalent and damaging to the individual and the organization) and adverse events. We also recognize there are other inﬂuences within the hospital environment, including malpractice legislation, licensing boards, regulatory agencies, and patient advocacy groups, but these issues are outside the scope of this discussion.

Attribution and Employees Learned Helplessness One unfortunate consequence of dispositional attributions is the linkage to cultures predominately derived from fear (Kohn et al., 2000; Reason, 2000). Resulting from the drive for accountability in the delivery of healthcare, organization leaders, professional boards, and society, in general, consider clinicians involved in harming patients as being ‘‘at fault.’’ Attributing blame to clinicians, such as physicians and nurses, is premised on the need to protect the public from harm. Furthermore, clinician punishment for mistakes is

Attribution Theory

97

expected by the majority of leaders regardless of causality (Vincent, 1997). The consequence of this dispositional attribution laden culture results in clinicians hiding problems and mistakes (Lawton & Parker, 2002) to avoid reprisal and punishment (Gibson & Singh, 2003). By hiding problems, the issue of employee silence is brought to the forefront. Attribution theory provides an appropriate framework to expose this under-recognized and unappreciated relationship that clinicians often experience in the delivery of care. Martinko and Gardner (1982) describe learned helplessness as the development of passive professionals consequential to repeated punishment for mistakes that makes success unlikely, even following organizational change. Learned helplessness is created by excessive dispositional attributions. Clinicians witness the adjudication of errors in the media, before licensing boards and courts, and within their own organizations, which often result in public chastisement and character assassination. These punitive experiences can be internalized by clinicians and lead to employee silence to avoid public humiliation for unfortunate but consequential mistakes. The present healthcare situation broadly parallels the ﬁndings Seligman reported in his famous canine experiment. In Seligman’s research, dogs were electrically shocked under varied conditions (Haggbloom et al., 2002). First, the dogs were chained to a wall where with each shock they were unable to escape. After repeated shocks and attempts to escape these shocks, the dog chain was disconnected and the dogs were shocked again. Despite the new freedom to escape, the dogs simply remained in place (Seligman & Maier, 1967). Dispositional attributions may prompt an analogous environment where clinicians’ are apathetic to incidents over which they lack control. Learned helplessness also illustrates perceptions regarding event controllability in considering how people feel about their ability to act (Walker, 1979). As mistakes usually manifest from perceptual (Wiegmann & Shappell, 1997), decision-making (Norman, 1988; Rasmussen, 1999), and skill-based system misalignments (Norman, 1988; Reason, 1990), the control jurisdiction is systemic in nature and not a dispositional factor. This is evident in a recent review of nursing performance where hospital nurses were found to experience high levels of stress and fatigue due to overwhelming workloads, cognitive overload, and excessive interruptions (Delucia, Ott, & Palmieri, in press). Subjected to these stressful working conditions, coupled with emergent and unexpected events, clinical professionals are likely to believe dispositional attributions they might normally perceive as false or erroneous (Gilbert, 1998). Through experiment, Gilbert, Tafarodi, and Malone (1993) demonstrated a phenomenon of ‘‘automatic believing’’ can occur under stressful conditions.

98

PATRICK A. PALMIERI AND LORI T. PETERSON

Cynicism Workplace cynicism can be created when clinicians encounter dispositional attributions, but the system is really the culpable agent (Andersson & Bateman, 1997). In addition, helplessness can lead to cynicism (Dorner, 1996). Cynical knowledge is generated by management actions (Goldner, Ritti, & Ference, 1977). For example, inadequate supervision, lack of management support, and deﬁciencies in resource availability help to build clinician cynicism about the organization’s goals (Palmieri et al., 2008; Reason, 1998). A high level of employee cynicism depicts an organization in crisis (Reichers, Wanous, & Austin, 1997). Cynicism is also derived from preserving and protecting organizational cultures from clinician-initiated change. When managers engage in frequent communications that focus on dispositional attributions, maladjusted workplace performance typically results (Martinko & Gardner, 1987). Managers further contribute to cynicism (Prussia, Brown, & Willis, 2003) when clinicians attempt to attribute causation for operational issues to the organizational work processes and managers repeatedly ignore these attempts (Reason, 1997; Reason et al., 2001). Managers appearing uninterested and uninvolved in work environments also engender employee cynicism (Reason & Hobbs, 2003) about management competence (Vincent, 1997) and the potentially consequential situations that remain hazardous (Reason et al., 2001). With heightened cynicism, attribution further contributes to considerable environmental stress (Rasmussen, 1999) and the stress leads to additional cynicism, as evidenced by the increasingly dissatisﬁed, and cynical, hospital nurses (Page, 2004; Stone et al., 2006). Clinician involvement in operations is absolutely critical to maintain protective care delivery systems (Reason & Hobbs, 2003). When clinicians who attempt to correct known issues are ignored by management, dysfunctional artifacts are collected. Known issues often include those related to work environments. Workload pressures, such as understafﬁng or stressful work assignments, only exacerbate cynicism in situations where dispositional attribution processes may already ﬂourish. For example, Tucker and Spear (2006) reported that inadequate nursing task times, multiple unplanned alterations to tasks, and frequent interruptions mid-task throughout a typical shift contribute to work overload and fatigue. While seldom discussed in the literature, stressful situations and work overload may cause departures from policies and procedures (Page, 2004). These deviations contribute to workplace attribution and worker cynicism (Vaughn, 1999). In reviews of medical error (Morrow, North, & Wickens, 2005) and nursing performance (Delucia et al., in press), cognitive overload

99

Attribution Theory

and environmental factors (such as cumbersome work processes and technologies) were found to hinder clinical performance. For example, an organizational issue, such as unpaid overtime at the end of the shift, is a wellknown phenomenon (Tucker & Spear, 2006). Nurses, unable to manage their heavy workload within a scheduled shift, report overtime lasting, on average, an hour per shift (Rogers, Wang, Scott, Aiken, & Dinges, 2004). Collectively, organizational characteristics can create stress and contribute to clinician cynicism. Hospital work environments are difﬁcult for clinicians to navigate (Kohn et al., 2000; Page, 2004) and overload is no longer an exception; instead, burdensome workloads have become normalized. Argyris and Scho¨n (1996) acknowledge in their discussion of organizational learning that there is often a substantial disconnect between the organization and the individual. Furthermore, people frequently know what needs to be done, however they are pessimistic, and potentially cynical regarding the outcome (Cook & Yanow, 1993). Cynicism limits both clinician practice and patient advocacy, while the organization struggles to move forward, reverse course, or even to overhaul operational practices. Management avoidance of attribution bias is an important moderator to reduce the likelihood that negative emotions and expectancies will contribute to the existence of workplace cynicism (Harvey, Martinko, & Douglas, 2006).

Attribution and Organizations Over time, the products of cynicism (Andersson, 1996; Harvey et al., 2006) and learned helplessness (Martinko & Gardner, 1982) can lead to the phenomena called organizational inertia (Amburgey, Kelly, & Barnett, 1993; Argyris, 1990; Kelly & Amburgey, 1991; Proehl, 2001; Rumelt, 1995). In this section, we discuss the manifestation and consequences of inertia and describe the emerging concept of the Just Culture. Inertia Healthcare organizations are inherently complex systems that adapt to change (Anderson, Crabtree, Steele, & McDaniel, 2005; Anderson & McDaniel, 2000) and exhibit dynamic network characteristics (McDaniel & Driebe, 2001) where multiple agents continuously act and react to other agents’ behavior and actions. Managers should recognize that hospitals function with a high degree of dispersed and decentralized organizational control (McDaniel & Driebe, 2001). Within an organization network, such as a hospital, multitudes of disparate decision-making actions are performed by

100

PATRICK A. PALMIERI AND LORI T. PETERSON

numerous agents (Waldrop, 1992). However, leaders may attribute group failures or system instability (Reason, 1998) to individual behavior (Bandura, 1999). Therefore, evaluating error etiology in a complex adaptive systems creates substantial, and potentially irreconcilable, situations for organizations to balance (Anderson, Issel, & McDaniel, 2003; Cilliers, 1998). This leads to the question of who (or what) is to blame for this failure? As a manager arrives at judgment, the formal decision following evaluation of a poor outcome, conﬂict between the organization and the clinician is certain when people are blamed for unintended failures. These judgments manifest in cynicism directed at the organization or lead to individuals feeling helpless about their ability to contribution to the organization. Indeed, poor group performance is a noted problem for organizations when management attributions are improperly directed (Brown, 1984; Kelley, 1967). The development of clinician knowledge speciﬁc to group dynamics and interpersonal relationships serves a critical role in healthcare team performance. Studies demonstrate that clinical practice is performed best in complimentary group settings where team work is embraced (Edmondson, 2004; Leggat, 2007). Dispositional attributions can manifest as inertia and seriously disrupt group synergy, contributing to suboptimal patient care and increased clinical failures. McElroy (1982, p. 416) suggests ‘‘leader-member conﬂict may be the directed result of a leader taking action based on his/her own casual analysis of the situation, a causal analysis potentially quite different from that of his/ her subordinates.’’ Similarly, Martinko and Gardner (1987) explain how managers often blame subordinate characteristics as the cause of poor performance while subordinates fault leaders for environmental ﬂaws leading to an inability to appropriately perform. Through repeated attribution cycles that manifest in blame, organizational effectiveness and the ability to change processes substantially diminish. This repetitive process eventually leads to an inertial state. Avoiding negative attribution cycles serves a positive role in facilitating the foundation for a Just Culture (Dekker, 2007). Just Culture Just Culture, a relatively new concept for hospitals (Dekker, 2007), is a culture where organizational norms support and encourage learning from mistakes rather than focusing on blaming and punishing those involved in error (Reason, 1997; Tucker & Edmondson, 2003). Just Culture, focused on trust (e.g., Dirks & Ferrin, 2002) and supportive management practices (e.g., Konteh, Mannion, & Davies, 2008), is an essential component for creating cultures where safety is valued and recognized as an organizational priority

Attribution Theory

101

(Dekker, 2007; Page, 2004; Tucker & Edmondson, 2003). Management support is important to encourage error reporting (IOM, 2007) and to identify acceptable and unacceptable behaviors (Reason, 1990). Just Culture resides within the organization safety culture (Wiegmann, Zhang, von Thaden, Sharma, & Gibbons, 2004). From research, Tucker and Edmondson (2003, p. 67) note ‘‘To learn from failures, people need to be able to talk about them without fear of ridicule or punishment.’’ Initially, healthcare focused on human error (Kohn et al., 2000) and those errors associated with perceptual limitations (Rasmussen, 1999) and employee behavior (van Vuuren, 1999). This approach, although helpful in understanding human error, did little to shift the focus from the traditional person-centered philosophy to a system perspective when addressing errors that result in adverse events (Palmieri et al., 2008). However, there has been a recent shift to scrutinize more often organizational factors, such as management practices, organization structures, and system processes (Flin et al., 2006; IOM, 2004), which impact employee performance (Singla, Kitch, Weissman, & Campbell, 2006) and cause error cascades that result in patient harm (Reason, 1997; IOM, 2004). The Just Culture is believed to be an essential prerequisite for healthcare safety and quality (Dekker, 2007; Khatri, Halbesleben, Petroski, & Meyer, 2007). This culture is postulated to offer an atmosphere of trust and encouragement for clinicians to openly engage in safety-related information exchanges (IOM, 2007, Kohn et al., 2000; Page, 2004). The key determinant to build and sustain a Just Culture is the ability to focus on situational attributions when evaluating failures. High levels of dispositional attribution for clinical errors detract from clinician acceptance of the Just Culture as legitimate. Therefore, clinician trust and the willingness to improve error communication will remain nominal when individual clinicians are the primary causative focus for sentinel events. The model presented in Fig. 2, recognizes how the frequent use of situational attributions leads to the Just Culture. Further discussion about Just Culture is beyond the focus of this paper.

Attribution Theory Complexity and Hospital Application Theory complexity relates to the application of attributional processes to practice when compared to the theoretical basis for attribution. In healthcare, the use of multiple theories to address complex trajectories of chronic illness and care systems management is prudent, perhaps even necessary to study, describe, explain, and intervene in phenomena. By

102

PATRICK A. PALMIERI AND LORI T. PETERSON

including well-deﬁned and situational explanatory theories, such as attribution theory, healthcare professionals can better understand seemingly unique experiences by studying the common elements across situations and between environments. With adaptation and empirical testing for applicability to healthcare management, attribution theory may begin to theoretically explain commonly experienced, frequently discussed, but rarely published, clinical hardships related to system failures attributed to individuals. Attribution theory not only provides a framework to study hospital managers and clinicians, but it also is applicable to the broader healthcare industry, especially given the emergence of the Just Culture. The complexity in attribution theory is driven by the scope and variety of potential problem solving applications. For example, attribution theory addresses mistakes and errors leading to accidents in a unique manner by seeking to understand the ascription from the management viewpoint, usually dispositional, versus the system perspective. As a result, causes that ought to be ascribed to poorly designed care delivery systems (Kohn et al., 2000) are often overlooked in favor of the belief that clinicians are carelessness or negligent (Cook & O’Connor, 2005; Helmreich & Davies, 2004; Kohn et al., 2000; Rasmussen, 1999; Reason, 2000). Therefore, attribution theory literally can be ‘‘brought to the bedside’’ along with the physician and nurse in order to answer important questions related to both the people and systems involved in errors. The root cause of patient care system failure precludes traditional approaches to change management as effective alternatives for process improvement. Turning to the disciplines of organizational behavior and theory, cognitive and behavioral psychology, and organizational sociology in order to learn more about attribution theory can offer an alternative approach to consider consequential phenomena.

Attribution and Healthcare Management Interventions The purpose of this paper was to introduce attribution theory to healthcare with a conceptual framework focused on dispositional attributions by managers. Although we will not address interventions in this paper, we would like to draw your attention to previous work in Advances in Health Care Management (AHCM) that provides further context about appropriate organization and management interventions to improve clinical cultures. Most notably, in AHCM Patient Safety and Health Care Management (vol. 7), four papers address interventions and strategies to: 1) improve the likelihood for

103

Attribution Theory

accurate attributions of causality (Batten, Goodman, & Distefano, 2008); 2) facilitate positive changes in organizational culture that support employee performance (McAlearney, 2008); 3) assist managers and supervisors learn how to balance safety culture promotion with performance accountability (Tamuz, Russell, & Thomas, 2008); and 4) appreciate the etiology of errors in care delivery systems that lead to consequential adverse events (Palmieri et al., 2008). Finally, an important contribution by McDaniel and Driebe (2001) in AHCM (vol. 2) frames healthcare organizations as complex adaptive systems that necessitate contemporary management practices in order to improve organizational processes and system outcomes.

CONCLUSION This paper has described attribution theory and demonstrated its suitability to healthcare. First, the evolution of attribution theory was described, followed by a discussion about the translation of attribution theory to healthcare. Then, concepts related to attribution theory were addressed. Next, the relevant work and contemporary knowledge that provides for the hardiness of attribution theory was evaluated to support the application as an important, but unrealized feature of the hospital work environment. This was accomplished with actual examples from adverse events that occurred in two hospitals where causes for these events were directed either to the organization and its complexity or to the practitioner and their competency. Then, the Healthcare Attribution Error Model was introduced to address dispositional and situational attribution, cynicism, learned helplessness, organizational inertia, and Just Culture. We next addressed complexity issues related to attribution theoretical translation to healthcare clinical practice and management. Finally, we suggest attribution theory provides an important theoretical framework that warrants further conceptual development and empirical research. The IOM (2001) report, Crossing the Quality Chasm, nonspeciﬁcally refers to attribution by advocating the reaction to healthcare error must change since the result of clinicians simply ‘‘trying harder’’ will only perpetuate the current ‘‘blame and shame’’ environment. However, by positively changing the hospital environment improvement can be realized (Runciman et al., 2003). Achievement will be apparent not only in the creation of an environment where more satisﬁed patients and fewer negative events occur, but also by creating the environment for a more satisﬁed work force.

104

PATRICK A. PALMIERI AND LORI T. PETERSON

Attribution theory supports the organizational discussion by explaining the need to focus on causal relationships instead of individual clinicians. In most cases, leaders are drawn toward the lowest common denominator, the clinician, with the focus of blame, fault and the assignment of responsibility (Taylor, 1981). Therefore, this form of attribution results in placing causation on the shoulders of a clinician, while permitting leaders to avoid personal responsibility (Burgoyne, 1982). When repeatedly manifested, this form of attribution ultimately results in the organization being held harmless (Green & Mitchell, 1979) and clinicians growing more fearful (Kohn et al., 2000; Page, 2004; Reason, 2000). Managers should begin to actively work with clinicians to deconstruct the punitive healthcare culture, while simultaneously building process improvements into the clinical work environment. Speciﬁcally, increased management attention to understanding the complexity of care delivery systems, to actively engage in enhancing the clinical work environment, and to seek system causes while avoiding personal attributions for human errors are essential beginning steps. Indeed, the fundamental philosophy advocated by To Err is Human (Kohn et al., 2000) strives to reduce the level of attribution directed at clinicians by promoting system solutions for critical failures. Attribution theory as applied to the hospital practitioners can unlock the many obstructed doors necessary to unearth meaningful solutions to challenging problems.

ACKNOWLEDGMENTS We would like to acknowledge the assistance of Alan Lind at the Duke University Fuqua School of Business, William Gardener at the Texas Tech University Rawls College of Business, Debra Brandon & Sharon Docherty, faculty at the Duke University School of Nursing, and Alexia Green, Dean at the Texas Tech University Health Sciences Center School of Nursing, for their helpful comments and suggestions on earlier versions of this ﬁnal paper. In addition, we would like to acknowledge the comments from reviewers at the Academy of Management, speciﬁcally for the theoretical guidance, and the Western Academy of Management, particularly for the suggested improvements for the model, in our earlier work. Furthermore, support was provided by the Duke University School of Nursing, the Duke Health Technology Solutions and the Texas Tech University Center for Healthcare Innovation, Education, and Research. Last, we thank the anonymous reviewers for their critical assessments and suggested improvements.

105

Attribution Theory

REFERENCES Amburgey, T. L., Kelly, D., & Barnett, W. P. (1993). Resetting the clock: The dynamics of organizational change and failure. Administrative Science Quarterly, 38, 51–73. Anderson, R. A., Crabtree, B. F., Steele, D. J., & McDaniel, R. R. (2005). Case study research: The view from complexity science. Qualitative Health Research, 15(5), 669–685. Anderson, R. A., Issel, M. L., & McDaniel, R. R. (2003). Nursing homes as complex adaptive systems: Relationship between management practice and resident outcomes. Nursing Research, 52(1), 12–21. Anderson, R. A., & McDaniel, R. R. (2000). Managing health care organizations: Where professionalism meets complexity science. Health Care Management Review, 25(1), 83–92. Andersson, L. M. (1996). Employee cynicism: An examination using a contract violation framework. Human Relations, 49, 1395–1418. Andersson, L. M., & Bateman, T. S. (1997). Cynicism in the workplace: Some causes and effects. Journal of Organizational Behavior, 18(5), 449–469. Argyris, C. (1957). Personality and organizations. New York: Harper & Row. Argyris, C. (1990). Overcoming organizational defenses: Facilitating organizational learning. Needham, MA: Allyn & Bacon. Argyris, C., & Scho¨n, D. (1996). Organizational learning: Theory, method and practice. Reading, MA: Addison Wesley. Bandura, A. (1999). Moral disengagement in the perpetration of inhumanities. Personality and Social Psychology Review, 3(3), 193–209. Batten, D., Goodman, G., & Distefano, S. M. (2008). Protecting the patient: Collaborating to achieve the ideal hospital work environment. In: G. T. Savage & E. W. Ford (Eds), Advances in health care management (Vol. 7, pp. 149–161). Bingley, England: Emerald Publishing Group. Battles, J. B., Kaplan, H. S., van der Schaaf, T. W., & Shea, C. E. (1998). The attributes of medical event reporting systems for transfusion medicine. Archives of Pathology and Laboratory Medicine, 122, 231–238. Berger, P. L., & Luckmann, T. (1967). The social construction of reality: A treatise in the sociology of knowledge. New York: Anchor Books. Bernardin, H. J. (1989). Innovative approaches to personnel selection and performance appraisal. Journal of Management Systems, 1, 25–36. Berta, W. B., & Baker, R. (2004). Factors that impact the transfer and retention of best practices for reducing error in hospitals. Health Care Management Review, 29(2), 90–97. Brown, K. A. (1984). Explaining poor group performance: An attributional analysis. Academy of Management Review, 9(1), 54–63. Burgoyne, J. H. (1982). Accident investigation. Journal of Occupational Accidents, 3, 289–297. Cilliers, P. (1998). Complexity and post modernism: Understanding complex systems. New York: Routledge. Clarian Health Partners. (2006). Heparin error backgrounder. Available at http://medicine.iu. edu/documents/Scope%20Content/Heparin%20error%20backgrounder.pdf. Retrieved February 2, 2009. Cook, R. I., & O’Connor, M. F. (2005). Thinking about accidents and systems. In: H. Manasse & K. Thompson (Eds), Improving medication safety. Mathesda, MD: American Society for Health-System Pharmacists.

106

PATRICK A. PALMIERI AND LORI T. PETERSON

Cook, R. I., & Woods, D. D. (1994). Operating at the sharp end: The complexity of human error. In: M. S. Bogner (Ed.), Human error in medicine (pp. 285–310). Hinsdale, NJ: Lawrence Erlbaum. Cook, S. D. N., & Yanow, D. (1993). Culture and organizational learning. Journal of Management Inquiry, 2(4), 373–390. Cyert, R., & March, J. G. (1963). A behavorial theory of the ﬁrm. Englewood Cliffs, NJ: Prentice-Hall. Davis, K., Schoenbaum, S. C., Collins, K. S., Tenney, K., Hughes, D. L., & Audet, A. J. (2002). Room for improvement: Patients report on the quality of their health care. New York: The Commonwealth Fund. DeJoy, D. M. (1985). Attributional processes and hazard control management in industry. Journal of Safety Research, 16, 61–71. DeJoy, D. M. (1999). Motivation. In: M. S. Wogalter, D. M. DeJoy & K. R. Laughery (Eds), Warnings and risk communication (pp. 221–244). Philadelphia: Taylor & Francis. Dekker, S. (2007). Just culture: Balancing safety and accountability. Aldershot, England: Ashgate Publishing. Delucia, P. R., Ott, T. E., & Palmieri, P. A. (in press). Performance in nursing. In: F. F. Durso (Ed.), Reviews of human factors and ergonomics (Vol. 5). Santa Monica, CA: Human Factors and Ergonomics Society. Devers, K. J., Pham, H. H., & Lui, G. (2004). What is driving hospitals’ patient-safety efforts? Health Affairs, 23(3), 103–115. Dirks, K. T., & Ferrin, D. L. (2002). Trust in leadership: Meta-analytic ﬁndings and implications for research and practice. Journal of Applied Psychology, 87(4), 611–628. Dorner, D. (1996). The logic of failure: Recognizing and avoiding error in complex situations. New York: Metropolitan Books. Edmondson, A. C. (2004). Learning from failure in health care: Frequent opportunities, pervasive barriers. Quality & Safety in Health Care, 13(S2), 3–9. Flin, R., Burns, C., Mearns, K., Yule, S., & Robertson, E. M. (2006). Measuring safety climate in health care. Quality & Safety in Health Care, 15(2), 109–115. Gano, D. (2007). Apollo root cause analysis: A new way of thinking (2nd ed.). Yakima, WA: Apollonian Publications. Gibson, R., & Singh, J. P. (2003). Wall of silence: The untold story of the medical mistakes that kill and injure millions of Americans. Washington, DC: Lifeline Press. Gilbert, D. T. (1998). Ordinary personology. In: D. T. Gilbert, S. T. Fiske & G. Lindzey (Eds), The handbook of social psychology (4th ed., pp. 89–150). New York: McGraw-Hill. Gilbert, D. T., Tafarodi, R. W., & Malone, P. S. (1993). You can’t not believe everything you read. Journal of Personality and Social Psychology, 65, 221–233. Goldner, F. H., Ritti, R. R., & Ference, T. P. (1977). The production of cynical knowledge in organizations. American Sociological Review, 42(4), 539–551. Gray, J. A. M. (2001). Evidence-based healthcare: How to make health policy and management decisions (2nd ed.). London: Churchill Livingstone. Green, S. G., & Mitchell, T. R. (1979). Attributional processes of leaders in leader-member interaction. Organizational Behavior and Human Performance, 23, 429–458. Haggbloom, S. J., Warnick, R., Warnick, J. E., Jones, V. K., Yarbrough, G. L., Russell, T. M., Borecky, C. M., McGahhey, R., Powell, J. L., Beavers, J., & Monte, E. (2002). The 100 most eminent psychologists of the 20th century. Review of General Psychology, 6(2), 139–215.

Attribution Theory

107

Harris, L. T., Todorov, A., & Fiske, S. T. (2005). Attributions on the brain: Neuro-imaging dispositional inferences, beyond the theory of mind. NeuroImage, 28, 763–769. Harvey, P., Martinko, M. J., & Douglas, S. C. (2006). Causal reasoning in dysfunctional leadermember interactions. Journal of Managerial Psychology, 21(8), 747–762. Heider, F. (1958). The psychology of interpersonal relations. New York: Wiley. Helmreich, R. L., & Davies, J. M. (2004). Culture, threat, and error: Lessons from aviation. Canadian Journal of Anesthesia, 51(6), R1–R6. Hofstede, G., & Hofstede, G. J. (2005). Cultures and organizations: Software of the mind (2nd ed.). New York: McGraw-Hill. House, R. J., Hanges, P. J., Javidan, M., Dorfman, P. W., & Gupta, V. (2004). Culture leadership and organizations: The globe study of 62 societies. Thousand Oaks, CA: Sage. Institute for Safe Medication Practices. (2006). Since when it is a crime to be human? http:// www.ismp.org/pressroom/viewpoints/julie.asp; November 30. Institute for Safe Medication Practices. (2008). Heparin errors continue despite prior highproﬁle fatal events. Medication Safety Alert, July 17. Available at http://www.ismp.org/ newsletters/acutecare/articles/20080717.asp Institute for Safe Medication Practices. (2009). Inattentional blindness: What captures your attention. Medication Safety Alert, February 26. Available at http://www.ismp.org/ newsletters/acutecare/articles/20090226.asp Institute of Medicine (IOM). (2001). Crossing the quality chasm: A new health system for the 21st century. Washington, DC: National Academy Press. Institute of Medicine. (2003). Health professions education: A bridge to quality. Washington, DC: National Academy Press. Institute of Medicine. (2004). Patient safety: Achieving a new standard for care. Washington, DC: National Academy Press. Institute of Medicine. (2007). Preventing medication errors: Quality chasm series. Washington DC: National Academies Press. Joint Commission Resources. (2005). What every health organization should know about sentinel events. Oakbrook Terrace, IL: Joint Commission Resources. Joint Commission Resources. (2007). Front line of defense: The role of nurses in preventing sentinel events (2nd ed.). Oakbrook Terrace, IL: Joint Commission Resources. Jones, E. E. (1979). The rocky road from acts to dispositions. American Psychologist, 34, 107–117. Kelley, E. (1972). Attribution in a social interactions. In: E. E. Jones, D. E. Kanouse, R. E. Nisbett, S. Valins & B. Weiner (Eds), Attribution: Perceiving the cause of behavior (pp. 1–26). Hillsdale, NJ: Lawrence Elbaum and Associates. Kelley, H. H. (1967). Attribution theory in social psychology. In: D. Levine (Ed.), Nebraska symposium on motivation (Vol. 15). Lincoln, NE: University of Nebraska Press. Kelly, D., & Amburgey, T. L. (1991). Organizational inertia and momentum: A dynamic model of strategic change. Academy of Management Journal, 34, 591–612. Kent, R., & Martinko, M. J. (1995). The measurement of attribution in organizational research. In: M. J. Martinko (Ed.), Attribution theory: An organizational perspective (pp. 17–34). Delray, FL: St. Lucie Press. Khatri, N., Halbesleben, J. R., Petroski, G. F., & Meyer, W. (2007). Relationship between management philosophy and clinical outcomes. Health Care Management Review, 32(2), 128–139. Kohn, L. T., Corrigan, J. M., & Donaldson, M. S. (Eds). (2000). To err is human: Building a safer health system. Washington, DC: National Academy Press.

108

PATRICK A. PALMIERI AND LORI T. PETERSON

Konteh, F. H., Mannion, R., & Davies, H. T. O. (2008). Clinical governance views on culture and quality improvement. Clinical Governance, 13(3), 200–207. Langdrige, D., & Butt, T. (2004). The fundamental attribution error: A phenomenological critique. British Journal of Social Psychology, 43, 357–369. Lawton, R., & Parker, D. (2002). Barriers to incident reporting in a healthcare system. Quality and Safety in Health Care, 11(1), 15–18. Leape, L. L., & Berwick, D. M. (2005). Five years after ‘‘To err is human’’: What have we learned? Journal of the American Medical Association, 293(19), 2384–2390. Leggat, S. G. (2007). Effective healthcare teams require effective team members: Deﬁning teamwork competencies. BMC Health Services Research, 7(1), 7–17. March, J. G., & Simon, H. A. (1993). Organizations (2nd ed.). Indianapolis, IN: Wiley. Martinko, M. J. (1995). The nature and function of attribution theory within the organizational sciences. In: M. J. Martinko (Ed.), Attribution theory: An organizational perspective (pp. 273–288). Delray Beach, FL: St Lucie Press. Martinko, M. J. (2004). Parting thoughts: Current issues and future directions. In: M. J. Martinko (Ed.), Attribution theory in the organizational sciences: Theoretical and empirical contributions: (pp. 297–306). Greenwich, CT: Information Age Publishing. Martinko, M. J., & Gardner, W. L. (1982). Learned helplessness: An alternative explanation for performance deﬁciencies. Academy of Management Review, 7(2), 195–204. Martinko, M. J., & Gardner, W. L. (1987). The leader/member attribution process. Academy of Management Review, 12(2), 235–249. McAlearney, A. S. (2008). Improving patient safety through organizational development: Considering the opportunities. In: G. T. Savage & E. W. Ford (Eds), Advances in health care management (Vol. 7, pp. 213–239). Bingley, England: Emerald Publishing Group. McDaniel, R. R., & Driebe, D. J. (2001). Complexity science and health care management. In: J. D. Blair, M. D. Fottler & G. T. Savage (Eds), Advances in health care management: (Vol. 2, pp. 11–36). Stamford, CT: JAI Press. McElroy, J. C. (1982). A typology of attribution leadership research. Academy of Management Review, 7(3), 413–417. Millenson, M. L. (2003). The silence. Health Affairs, 22(2), 103–112. Mitchell, T. R., & Woods, R. E. (1980). Supervisor’s responses to subordinate poor performance: A test of an attributional model. Organizational Behavior and Human Performance, 25, 123–138. Morrow, D., North, R., & Wickens, C. D. (2005). Reducing and mitigating human error in medicine. Reviews of Human Factors and Ergonomics, 1, 254–296. Norman, D. (1988). The design of everyday things. New York: Doubleday. Page, A. (Ed.) (2004). Keeping patients safe: Transforming the work environment of nurses. Washington, DC: National Academy Press. Palmieri, P. A., Delucia, P. R., Peterson, L. T., Ott, T. E., & Green, A. (2008). The anatomy and physiology of error in adverse healthcare events. In: G. T. Savage & E. W. Ford (Eds), Advances in health care management (Vol. 7, pp. 33–68). Bingley, United Kingdom: Emerald. Perrow, C. (1984). Normal accidents: Living with high risk systems. New York: Basic Books. Proehl, R. A. (2001). Organizational change in the human services. Thousand Oaks, CA: Sage. Prussia, G. E., Brown, K. A., & Willis, P. G. (2003). Mental models of safety: Do managers and employees see eye to eye. Journal of Safety Research, 34(2), 143–156. Rasmussen, J. (1990). The role of error in organizing behavior. Ergonomics, 33, 1185–1199.

Attribution Theory

109

Rasmussen, J. (1999). The concept of human error: Is it useful for the design of safe systems in health care? In: C. Vincent & B. deMoll (Eds), Risk and safety in medicine (pp. 31–47). London: Elsevier. Reason, J. T. (1990). Human error. New York: Cambridge University Press. Reason, J. T. (1997). Managing the risks of organizational accidents. Aldershot, England: Ashgate. Reason, J. T. (1998). Achieving a safe culture: Theory and practice. Work and Stress, 12, 293–306. Reason, J. T. (2000). Human error: Models and management. British Medical Journal, 320(7237), 768–770. Reason, J. T., Carthey, J., & de Leval, M. R. (2001). Diagnosing ‘‘Vulnerable system syndrome’’: An essential prerequisite to effective risk management. Quality & Safety in Health Care, 10(S2), 21–25. Reason, J. T., & Hobbs, A. (2003). Managing maintenance error: A practical guide. Aldershot, England: Ashgate. Reichers, A. E., Wanous, J. P., & Austin, J. T. (1997). Understanding and managing cynicism about organizational change. Academy of Management Executive, 11(1), 48–59. Reising, L., & Portwood, B. (2007). Root cause analysis and quantitative methods: Yin and yang? Paper presented at the 25th International System Safety Conference, August 17–23, Baltimore, MD. Roesch, S. C., & Amirkham, J. H. (1997). Boundary conditions for self-serving attributions: Another look at the sports pages. Journal of Applied Social Psychology, 27, 245–261. Rogers, A., Wang, W. T., Scott, L. D., Aiken, L. H., & Dinges, D. F. (2004). The working house of hospital staff nurses and patient safety. Health Affairs, 23(4), 202–212. Ross, L., & Nisbett, R. E. (1991). The person and the situation. New York: McGraw-Hill. Ruchlin, H. S., Subbs, N. L., & Callahan, M. A. (2004). The role of leadership in instilling a culture of safety: Lessons from the literature. Journal of Healthcare Management, 19(1), 47–58. Rumelt, R. P. (1995). Inertia and transformation. In: C. A. Montgomery (Ed.), Resource-based and evolutionary theories of the ﬁrm: (pp. 101–132). Boston: Kluwer Academic Publishers. Runciman, W. B., Merry, A. F., & Tito, F. (2003). Error, blame and the law in healthcare: An antipodean perspective. Annals of Internal Medicine, 138(12), 974–979. Sasou, K., & Reason, J. T. (1999). Team errors: Deﬁnitions and taxonomy. Reliability Engineering and System Safety, 65(1), 1–9. Seligman, M. E. P., & Maier, S. F. (1967). Failure to escape traumatic shock. Journal of Experimental Psychology, 74, 1–9. Shalo, S. (2007). To err is human: But for some nurses, a crime. American Journal of Nursing, 107(3), 20–21. Singla, A. K., Kitch, B. T., Weissman, J. S., & Campbell, E. G. (2006). Assessing patient safety culture: A review and synthesis of the measurement tools. Journal of Patient Safety, 2(3), 105–115. Smetzer, J. L., & Cohen, M. R. (1998). Lessons from Denver medication error/criminal negligence case: Look beyond blaming individuals. Hospital Pharmacy, 33, 640–656. Smetzer, J. L., & Cohen, M. R. (2006). Lessons from Denver. In: P. Aspden, J. Wolcott, J. L. Bootman & L. R. Cronenwett (Eds), Preventing medication errors (pp. 43–104). Washington, DC: National Academy Press.

110

PATRICK A. PALMIERI AND LORI T. PETERSON

Standing, C., Guilfoyle, A., Lin, C., & Love, P. (2006). The attribution of success and failure in it projects. Industrial Management & Data Systems, 106(8), 1148–1165. State of Wisconsin. (2006). Criminal complaint: State of Wisconsin versus Julie Thai: Wisconsin: Circuit court of Dane County. Stone, P. W., Larson, E. L., Mooney-Kane, C., Smolowitz, J., Lin, S. X., & Dick, A. W. (2006). Organizational climate and intensive care unit nurses’ intention to leave. Critical Care Medicine, 34(7), 1907–1912. Tamuz, M., Russell, C. K., & Thomas, E. J. (2008). Promoting patient safety by monitoring errors: A view from the middle. In: G. T. Savage & E. W. Ford (Eds), Advances in Health Care Management (Vol. 7, pp. 69–99). Bingley, England: Emerald Publishing Group. Taylor, D. H. (1981). The hermeneutics of accidents and safety. Ergonomics, 24, 487–495. Tucker, A. L. (2004). The impact of organizational failures on hospital nurses and their patients. Journal of Operations Management, 22(2), 151–169. Tucker, A. L., & Edmondson, A. C. (2003). Why hospitals don’t learn from failures: Organizational and psychological dynamics that inhibit system change. California Management Review, 45(2), 55. Tucker, A. L., & Spear, S. J. (2006). Operational failures and interruptions in hospital nursing. Health Services Research, 41(3), 643–662. van Vuuren, W. (1999). Organisational failure: Lessons from industry applied in the medical domain. Safety Science, 33(1–2), 13–29. Vaughn, D. (1999). The dark side of organizations: Mistake, misconduct, and disaster. In: J. Hagan & K. S. Cook (Eds), Annual review of sociology (pp. 271–305). Palo Alto, CA: Annual Reviews. Vincent, C. (1997). Risk, safety and the dark side of quality. British Medical Journal, 314, 1175–1176. Vincent, C. (2003). Understanding and responding to adverse events. New England Journal of Medicine, 348(11), 1051–1056. Waldrop, M. M. (1992). Complexity: The emerging science at the edge of order and chaos. New York: Simon & Schuster. Walker, L. E. (1979). How battering happens and how to stop it. In: L. E. Walker (Ed.), The battered woman (pp. 59–78). New York: Harper & Row. Weick, K. E. (1995). Sensemaking in organizations. Thousand Oaks, CA: Sage. Weick, K. E., & Sutcliffe, K. M. (2001). Managing the unexpected: Assuring high performance in a range of complexity. San Francisco: Jossey-Bass. Weick, K. E., Sutcliffe, K. M., & Obstfeld, D. (1999). Organizing for high reliability: Processes of collective mindfulness. Research in Organizational Behavior, 21, 81–123. Weiner, B. (1985). An attribution theory of achievement motivation and emotion. Psychological Review, 92(4), 548–573. Weiner, B. (1986). An attributional theory of motivation and emotion. New York: Springer-Verlag. Weiner, B. (1995a). Attribution theory in organizational behavior: A relationship of mutual beneﬁt. In: M. J. Martinko (Ed.), Attributional theory in organizational perspective (pp. 3–6). Del Ray Beach, FL: St. Lucie Press. Weiner, B. (1995b). Judgment of responsibility: A foundation for a theory of social conduct. New York: Guilford Press. Wiegmann, D. A., & Shappell, S. A. (1997). Human factor analysis of post-accident data: Applying theoretical taxonomies of human error. The International Journal of Aviation Psychology, 7(4), 67–81.

Attribution Theory

111

Wiegmann, D. A., Zhang, H., von Thaden, T. L., Sharma, G., & Gibbons, A. M. (2004). Safety culture: An integrative review. International Journal of Aviation Psychology, 14(2), 117–134. Wisconsin Hospital Association. (2006). Hospital association statement regarding legal actions against nurse. Available at http://www.wha.org/newsCenter/pdf/nr11-2-06Crimchargestmt. pdf. Retrieved November 10. Wisconsin Medical Society. (2006). Position statement regarding attorney general charges ﬁles today. Available at http://www.wha.org/newsCenter/pdf/2006wms11-7.pdf. Retrieved November 7, 2008. Woods, D. D., & Cook, R. I. (2002). Nine steps to move forward from error. Cognition, Technology, and Work, 4(2), 137–144. World Health Organization. (2008). World alliance for patient safety: Progress report 2006–2007. Geneva, Switzerland, WHO Press. Available at http://www.who.int/ patientsafety

This page intentionally left blank

SECTION II ORGANIZATIONAL DEVELOPMENT AND STRATEGIC MANAGEMENT PERSPECTIVES

This page intentionally left blank

MEASURING UP: ARE NURSE STAFFING MEASURES ADEQUATE FOR HEALTH SERVICES RESEARCH?$ Lynn Unruh, C. Allison Russo, H. Joanna Jiang and Carol Stocks ABSTRACT Background – Reliable and valid hospital nurse stafﬁng measures are a major requirement for health services research. As the use of these measures increases, discussion is growing as to whether current nurse stafﬁng measures adequately meet the needs of health services researchers. Objective – This study assesses whether the measures, sampling frameworks, and data sources meet the needs of health services research in areas such as stafﬁng assessment; patient, nurse, and ﬁnancial outcomes; and prediction of stafﬁng.

$

This work was performed under contract with the Agency for Healthcare Research and Quality (AHRQ). The opinions expressed in this chapter are those of the authors and do no necessarily reﬂect the views of AHRQ or the U.S. Department of Health and Human Services.

Biennial Review of Health Care Management: Meso Perspectives Advances in Health Care Management, Volume 8, 115–154 Copyright r 2009 by Emerald Group Publishing Limited All rights of reproduction in any form reserved ISSN: 1474-8231/doi:10.1108/S1474-8231(2009)0000008009

115

116

LYNN UNRUH ET AL.

Methods – We performed a systematic review of articles from 1990 through 2007, which use hospital nurse stafﬁng measures in original research, or which address the validity, reliability, and availability of the measures. Taxonomies of measures, sampling frameworks, and sources were developed. Articles were analyzed to assess what measures, sampling strategies, and sources of data were used and to ascertain whether the measures, samples, and sources meet the needs of researchers. Results – The review identiﬁed 107 articles that use hospital nurse stafﬁng measures for original research. Multiple types of measures, some of which are used more often than others and some of which are more valid than others, exist in each of the following categories: stafﬁng counts, stafﬁng/ patient load ratios, and skill mix. Sampling frameworks range from hospital units to all hospitals nationally, with all hospitals in a state being the most common. Data sources range from small-scale surveys to national databases. The American Hospital Association Annual Survey is the most frequently used data source, but there are limitations with its nurse stafﬁng measures. Arguably, the multiplicity of measures and differences in sampling and data sources are due, in part, to data availability. The limitations noted by other researchers and by this review indicate that stafﬁng measures need improvements in conceptualization, content, scope, and availability. Discussion – Recommendations are made for improvements to research and administrative practice and to data.

BACKGROUND Hospital nurse stafﬁng measures have been used in health services research for several decades. A ‘‘count’’ of nurses is used for workforce planning (O’Brien-Pallas, Cockerill, & Leatt, 1992) and to assess facility stafﬁng (Edwardson & Giovannetti, 1994; Shullanberger, 2000). Nurse stafﬁng measures are often analyzed in association with other variables such as patient outcomes and ﬁnancial costs (Aiken, Clarke, Sloane, Sochalski, & Silber, 2002; Cho, Keteﬁan, Barkauskas, & Smith, 2003). These associative studies have become popular because of concerns about the quality and costs of hospital care. As the utilization of nurse stafﬁng measures increases, discussion is growing around whether current nurse stafﬁng measures adequately meet

Are Nurse Stafﬁng Measures Adequate for Health Services Research?

117

the needs of health services researchers. Primarily at issue are the validity and reliability of the measures. Validity issues involve measurement of the intended nursing personnel and workload. For example, validity may be threatened if nurse stafﬁng measures include nurse administrators and educators when the intent is to count only direct caregivers, or if the nurse stafﬁng measures include nurses caring for long-term care patients and/or outpatients when the study is examining inpatient acute care (Harless & Mark, 2006; Mark, 2006; Unruh, 2001). Validity may also be problematic depending on the measure of patient volume used in stafﬁng ratios, because some measures are better at capturing the patient load than others (Unruh, Fottler, & Talbott, 2003; Unruh & Fottler, 2006). Reliability is problematic when data are collected from many different units or different hospitals with varying deﬁnitions of the measure and collection methods. Any of the validity issues mentioned earlier can contribute to reliability problems if different stafﬁng measures from different data sources are thrown together without regard for the internal content of the measures. As an example, nurse full-time equivalents (FTEs) may be calculated in various ways such as by adding full- and part-time positions together or by adding actual hours of care together. If measures of nurse FTEs from study sources are not calculated in the same way, reliability is low (Spetz, 2004; Unruh, 2001, 2002, 2003; Unruh et al., 2003). Another issue regarding nurse stafﬁng measures is the availability of data (Jiang, Stocks, & Wong, 2006; Mark, 2006). The collection of hospital stafﬁng data is not federally required, and data are often difﬁcult to obtain. One voluntarily collected national data set for hospital nurse stafﬁng measures is the American Hospital Association (AHA) Annual Survey. However, these data have been noted to exclude information on nursing assistants (NAs) from 1994 to 2003, include both indirect care givers and managers in the registered nurse (RN) category, and combine acute and long-term care staff (Jiang et al., 2006; Kovner, Jones, Zhan, Gergen, & Basu, 2002; Mark, 2006). Less is known about the collection of hospitallevel nurse stafﬁng data at the state level. It appears that systematic collection occurs in some states, but not in others (Needleman, Buerhaus, Mattke, Stewart, & Zelevinsky, 2002). It is very likely that problems with validity and reliability of nurse stafﬁng measures are inherently related to data availability. Researchers may have no other choice than to use certain measures because no others are readily available. It is also possible that the sampling frameworks for nurse stafﬁng studies are convenience samples, rather than planned and intended samples, because of limited sources of data. As a result, part of the discussion around

118

LYNN UNRUH ET AL.

nurse stafﬁng measures should focus on identifying the current sampling frameworks and data sources and on assessing their adequacy for the research questions being explored. In short, the issue is whether nurse stafﬁng measures meet the following needs: 1. Nurse stafﬁng measures are valid, reliable, appropriate to the research study, and comparable across studies. It is important that nurses are counted accurately in their separate categories (e.g., RNs, licensed practical nurses (LPNs), and NAs). Moreover, nurses counted in the measure should correspond to the nursing roles being studied (e.g., direct care, non-direct care, both), and they should correspond to the patient populations being cared for in the study (e.g., inpatient, outpatient, acute, long-term care). It is also important that patient volume is adequately captured and reﬂective of the workload of nurses, and measures should be consistent across all observations in the study. 2. Samples are appropriate for the type of study and allow for generalization to intended population. 3. Sources of data enable the aforementioned to occur.

OBJECTIVES Whether one is conducting research involving hospital nurse stafﬁng, interpreting the results of such studies, or using stafﬁng measures for administrative purposes, it is necessary to have information about the nurse stafﬁng measures commonly used in health services research, the samples from which the measures are drawn, the sources of the data, and how well suited the measures, sample, and data are for particular applications. Understanding these issues helps researchers make research design decisions and clarify the signiﬁcance of results obtained from using speciﬁc nurse stafﬁng measures. It assists managers in deciding what type of nurse stafﬁng measure to use and to understand the limitations of those measures. It helps policy makers interpret research results, including conﬂicting results due to the use of different measures. Identifying data needs may also promote improvements in nurse stafﬁng data collection. Only a few studies have assessed aspects of the validity and reliability of nurse stafﬁng measures (Harless & Mark, 2006; Jiang et al., 2006; Mark, 2006; Reiner et al., 2005). In this chapter, we go beyond a summary of those studies by providing a broader exploration of stafﬁng measures. Moreover,

Are Nurse Stafﬁng Measures Adequate for Health Services Research?

119

we explore sampling issues, and more intensely examine data availability issues, which are touched on only brieﬂy in existing publications. Through a systematic review and analysis of the literature, this study summarizes and synthesizes (1) what hospital nurse stafﬁng measures are commonly used in health services research and their frequency of use; (2) what sampling strategies are employed; (3) what are the data sources for those measures; (4) what limitations of the data, samples, and sources have been noted in the literature; and (5) whether the measures, samples, and sources meet the needs and goals of health services research, and if they do not, what would. We conclude with recommendations for improving health services administration practice and research.

THEORETICAL FRAMEWORK Nurse stafﬁng measures generally indicate the level of nursing personnel and the ratio of nursing personnel to patient load. These indicators serve multiple purposes in health services research and administration. They can be used as (1) estimates of stafﬁng adequacy; (2) structural elements related to patient outcomes; (3) structural elements related to nurse outcomes; (4) structural elements related to ﬁnancial outcomes (cost, efﬁciency, productivity, net revenue, or beneﬁts); and (5) outcome elements that are related to structure. As estimates of stafﬁng adequacy, nurse stafﬁng measures are used to assess and plan stafﬁng and workforce needs (Kovner, Jones, & Gergen, 2000; Spetz, 2000; Unruh, 2002). As structural elements related to patient, nurse, and ﬁnancial outcomes, they are part of analyses that utilize the Donabedian structure-process-outcome model (Donabedian, 1969, 1988). Good structure such as good stafﬁng contributes to good process and ultimately to good outcomes such as healthy patients (Needleman et al., 2002), satisﬁed nurses (Aiken et al., 2002), and positive institutional ﬁnances (McCue, Mark, & Harless, 2003). The Donabedian structure-process-outcomes model is also at the basis of the ﬁfth use of nurse stafﬁng measures – as the outcomes that are determined or inﬂuenced by structural and process elements such as hospital and market characteristics and changes. In this case, the stafﬁng measures are the outcomes, and the other variables are the structural or process predictors of stafﬁng levels and ratios (Brewer & Frazier, 1998). Our approach to assess nurse stafﬁng measure takes the multidimensional nature discussed earlier into account. We investigate the multiple uses of the measures and the appropriateness of using the measures for those purposes.

120

LYNN UNRUH ET AL.

METHODS Literature Review Search Strategy A comprehensive literature review of journal articles was performed to assess the use of hospital nurse stafﬁng measures in health services and nursing research in the United States. Additionally, we noted studies addressing the validity, reliability, and availability of these measures. The review was conducted using the following article databases for years 1990 through December 2007: Academic Search Premier, CINAHL, Econlit, Health Source Nursing/Academic Edition, and Medline. Search terms included nurse/licensed nurse (LN)/RN stafﬁng; nurse stafﬁng measures; nurse skill mix; nurse stafﬁng and patient outcomes/adverse events/quality; nurse stafﬁng and costs/ length of stay; nursing workload/patient load/productivity; nurse stafﬁng and work environment; nurse stafﬁng and job satisfaction; nurse stafﬁng and magnet hospitals; nurse stafﬁng and turnover/absenteeism; and nurse stafﬁng measures validity/reliability/availability. Articles were selected for review if authors used hospital nurse stafﬁng measures in their original research, or if they examined the validity, reliability, and availability of the measures. The review was limited to studies involving U.S. hospitals with the aim of examining the underpinnings of U.S. nurse stafﬁng measures. To assess the extent of smaller, facility-level research, study inclusion was not limited to research using large data sets. Studies utilizing data from national, state, multi-state, multiple hospitals, a single hospital, and a single hospital unit were included. Studies involving surveys of nurses were examined if they included counts of the nursing staff. Reviews of the literature involving nurse stafﬁng were used to crosscheck whether all appropriate studies were included in this review, but were not included in this review itself.

Taxonomies of Measures, Sampling Frameworks, and Sources A large number of hospital nurse stafﬁng measures, sampling strategies, and data sources were used in these studies. To systematize this analysis, we looked for consistent ways to refer to these aspects. After reviewing all the studies, we found that authors consistently distinguished among three major categories of stafﬁng measures that we will refer to as ‘‘counts,’’ ‘‘nursing staff/patient load,’’ and ‘‘skill mix.’’ Within ‘‘counts,’’ studies distinguished between FTE positions, nursing hours, and numbers of nurses. Most studies did not provide details

Are Nurse Stafﬁng Measures Adequate for Health Services Research?

121

regarding the content of the measures of counts (e.g., which nursing roles or which areas of work – inpatient, outpatient, long-term care – were included in the count or how FTEs were calculated). Within ‘‘nursing staff/patient load’’ measures, studies ﬁrst distinguished between using FTEs (or occasionally numbers) and nursing hours in the numerator. After noting this distinction in the numerator, we found that researchers used an even larger set of ‘‘patient load’’ denominators, including average daily census, number of occupied beds, number of discharges or admissions, and the number of patients, patient days, and adjusted patient days. The adjusted patient days were ‘‘adjusted’’ for any one or more of the following: outpatient care, case mix, patient severity, nursing intensity weight, discharge day, and patient turnover. All the patient load indicators, except for average daily census and the number of beds, were used in conjunction with both FTEs and hours. Given this, we kept the FTE and hour distinctions where appropriate when categorizing ‘‘nursing staff/patient load’’ measures. An additional measure frequently encountered in this category was ‘‘patient or workload/nurse.’’ This taxonomy produced 11 nursing staff/patient load measures which are listed in the results tables. In the category of skill mix, the primary distinction was the type of nurse in the numerator: RNs, LPNs, NAs, and LNs. Except for a few noted studies that used total hospital employees, the denominator was usually the total number of nurses (RNs, LPNs, NAs). Sampling strategies were categorized by the sampling frameworks (i.e., population targeted and sample sizes), ranging from one hospital unit to nearly all U.S. hospitals. There were ten sampling framework categories: hospital unit; single hospital; multiple hospitals; state sample of hospitals; state, all acute-care general hospitals; multi-state; national sample of hospitals; national, all acute-care general hospitals; federal; and other. These sampling categories are not to be confused with the level of analysis, which for a few studies was at the unit-level. Categories of hospital nurse stafﬁng data sources were established if the source was used in more than one study. Otherwise, the data source was placed in the ‘‘other’’ category.

RESULTS The literature review identiﬁed 107 journal articles using hospital nurse stafﬁng measures for original research. These are listed in the Appendix alphabetically by order of appearance. The papers used nurse stafﬁng

122

LYNN UNRUH ET AL.

measures for research in ﬁve main areas or applications: hospital nurse stafﬁng assessment; hospital nurse stafﬁng related to patient outcomes; hospital nurse stafﬁng related to nurse outcomes; hospital nurse stafﬁng related to ﬁnancial performance; and predictors of hospital nurse stafﬁng. These ﬁve areas ﬁt Donabedian’s structure-process-outcome model as previously discussed. The literature search also identiﬁed four purely methodological studies addressing issues and limitations of nurse stafﬁng measures. Of the 107 articles using hospital nurse stafﬁng measures in original research, several presented results in more than one research application, producing 123 studies involving nurse stafﬁng measures. Twenty-ﬁve of these studies concerned hospital stafﬁng assessment, which predominantly focused on the downsizing of the hospital nursing workforce in the 1990s, the hospital nursing shortage, and the recently mandated minimum stafﬁng ratios in California. The relationship between nurse stafﬁng and patient outcomes (e.g., falls, medication errors, mortality, nosocomial infections, pressure ulcers, and other adverse events) was featured in 53 studies. Seven pertained to the interaction between nurse stafﬁng and nurse outcomes (e.g., needle stick injuries, patient assaults on nurses, and nurse burnout), whereas 27 examined the relationship of nurse stafﬁng to hospital ﬁnancial performance. An emerging area of research examines predictors of nurse stafﬁng, such as managed care, type of ownership, size, support services, and wages (used in 11 studies).

Hospital Nurse Stafﬁng Measures Used in Health Services Research Table 1 speciﬁes the articles using speciﬁc types of nurse stafﬁng measures within each research application, whereas Table 2 indicates the frequency of use of those measures. Except in stafﬁng assessment applications, overall, stafﬁng counts are infrequently used in evaluations of hospital nurse stafﬁng (used in only 24 of 123 studies). When used, FTEs or hours predominate the count measurements within the hospital stafﬁng assessment, patient outcomes, and ﬁnancial performance research applications, with studies nearly equally distributed between these two stafﬁng count categories. Stafﬁng counts are used only once in a nurse outcome study, which measured the number of nurses, and in one study evaluating predictors of nurse stafﬁng, which measured FTEs. Instead of nursing counts, the predominant stafﬁng measures within all ﬁve research applications are nursing staff to patient load ratios (used in 104 of 123 studies). Overall, the ratios used most frequently are hours per patient

Hospital Nurse Stafﬁng Measures, Sampling Frameworks, and Data Sources Used in Health Services Research Studies. Article Reference Stafﬁng assessment studies (N ¼ 25)

Stafﬁng measures (staff categories ¼ RN, LPN, LN, NA, Stafﬁng counts FTEs 1, 12, 17, 18, 20, 21 Hours 4, 5, 6, 9, 13, 14, 16 Numbers 7 Nursing staff (NS)/patient load NS FTEs/average daily 1 census NS FTEs/bed 7 NS FTEs/discharge or None admission NS FTEs or numbers/ 9 patient NS FTEs/patient day 2, 3 NS FTEs/adjusted 2, 3, 12, 17, 18, 19, patient day a 20, 21, 22 NS hours/discharge or 13, 14 admission NS hours/patient 8, 15 NS hours/patient day 9, 23, 24, 25

Nurse outcomes studies (N ¼ 7)

Financial performance studies (N ¼ 27)

Predictors of stafﬁng studies (N ¼ 11)

nurses) 21, 27 31, 77 None

None None 80

87, 88 86 9

104 None None

26, 27, 66

79

None

None

37, 38, 61 57, 61

None None

61 61

100, 101 107

30, 41, 42, 44, 48, 68, 74

None

30, 42, 68, 94, 95, 96

105

34 50, 51, 56, 58, 59

None None

None 59, 89

None 102

71

None

None

None

None 82, 83

8, 92 9, 39, 69, 73, 92, 93

None 97, 98, 99, 105, 106

13, 14

None 31, 35, 36, 39, 43, 67, 69, 72, 73, 75, 76, 77 33, 55, 60, 64

None

None

None

11

28, 45, 46, 47, 49, 53

28, 78

90, 91

None

9, 10, 11, 17, 18, 20

26, 32, 33, 34, 35, 36, 39, 43, 52, 54, 56, 57, 70, 72

81

9, 32, 39, 69, 88, 92

99, 103, 105

123

NS hours/adjusted patient daya Patient or workload/nurse Skill mix (by FTE or hours) RN/nurse

Patient outcomes studies (N ¼ 53)

Are Nurse Stafﬁng Measures Adequate for Health Services Research?

Table 1.

124

Table 1. (Continued ) Article Reference Stafﬁng assessment studies (N ¼ 25)

Patient outcomes studies (N ¼ 53)

Nurse outcomes studies (N ¼ 7)

Financial performance studies (N ¼ 27)

Predictors of stafﬁng studies (N ¼ 11)

4, 10 17, 18, 20, 21 4, 10 4, 9

43, 70 21 43, 67, 70 29, 36, 40, 62, 63, 65, 67

None None None 81

None None None 9, 40, 63, 65, 84, 85

None None None None

Sampling frameworkb,c Single hospital unit Single hospital

None 23, 24, 25

31, 32, 45, 46, 47, 63, 34, 40, 48, 49, 53, 67, 69, 72, 76 35, 43, 62, 75, 77,

None None

32, 63, 84 40, 69, 95, 96

None None

78, 79, 81, 82, 83

93

None 28 None None

9, 68, 89 8, 27, 39, 42, 65, 73, 86, 61, 92, 94 88 59, 85, 87

97, 99, 100, 101, 103, 104, 106 None 98, 105 None 102, 107

None

None

None

Multiple hospitals

None

State, sample State, all acute-care general hospitals Multi-state National, sample

4, 6, 9, 10 11 3, 5, 8, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22 None None

National, all acute-care general hospitals Federal (e.g., VA) Other/unknown

1, 2, 7, 12

29, 36, 41, 68, 74 21, 28, 30, 33, 39, 42, 44, 56, 60, 61, 65, 71, 73 64 26, 27, 38, 50, 51, 52, 54, 55, 57, 58, 59, 66 37, 70

None None

None None

80 None

None 90, 91

None None

1, 2, 7, 12, 17, 18, 19, 20, 21, 22

21, 27, 28, 37, 38, 44, 50, 51, 52, 54, 55, 57, 58, 59, 66, 70, 75

28

59, 87, 89

102, 107

Data source American Hospital Association Annual Survey

LYNN UNRUH ET AL.

LPN/nurse LN/nurse NA/nurse Other:

Survey or primary data collection Other or N/A

4, 6, 9, 10

26, 36

None

9

None

3, 8, 13, 14, 15, 16

39, 56, 60, 71, 73

None

8, 39, 73, 86, 92, 94

105

None

61

None

61

None

None

30, 42

None

27, 42

None

None

64

None

88

None

5

33, 56

None

None

98

None

None

81

None

103, 104

17, 18, 19, 20, 21, 22

21, 28,

28

None

None

23, 24, 25,

31, 32, 34, 45, 46, 47, 48, 49, 53, 62, 67, 69, 72, 76, 77 29, 35, 40, 41, 63, 65, 68, 74 43

79, 80, 82, 83

32, 69, 84, 93, 95, 96

99, 106

78

63, 65, 68, 85, 40

97, 100, 101

None

90, 91

None

11 None

a Adjusted for any one or more of the following: outpatient care, case mix, patient severity, nursing intensity weight, Medstat Resource Demand Scale, discharge day, and patient turnover. b When the sampling framework involves a sampling of data, it may be random, purposive, or convenience. c The level of analysis for these samples may be at the hospital or unit level.

Are Nurse Stafﬁng Measures Adequate for Health Services Research?

California Nursing Outcomes Coalition California Ofﬁce of Statewide Health Planning and Development Florida Agency for Health Care Administration Maryland Hospital Association Survey Multiple state data sources New York Institutional Cost Report Outcomes Research in Nursing Administration Project Pennsylvania Department of Health Stafﬁng records

125

126

Table 2.

LYNN UNRUH ET AL.

Hospital Nurse Stafﬁng Measures by Frequency of Use within Applications and in Total. Stafﬁng Patient Assessment Outcomes Studies Studies (N ¼ 25) (N ¼ 53)

Nurse Financial Predictors Total Outcomes Performance of Stafﬁng Studies Studies Studies Studies (N ¼ 123) (N ¼ 7) (N ¼ 27) (N ¼ 11)

Stafﬁng measures (staff categories ¼ RN, LPN, LN, NA, nurses) Stafﬁng counts FTEs 6 2 0 2 Hours 7 2 0 1 Numbers 1 0 1 1 Total 14 4 1 4 Nursing Staff (NS)/patient load ratios NS FTEs/average 1 3 daily census NS FTEs/bed 1 3 NS FTEs/discharge 0 2 or admission NS FTEs or 1 7 numbers/patient NS FTEs/patient 2 1 day 9 5 NS FTEs/adjusted patient day a NS hours/discharge 2 1 or admission NS hours/patient 2 0 NS hours/patient 4 12 day 2 4 NS hours/adjusted patient daya Patient or 1 6 workload/nurse Total 25 44 Skill mix (by FTE or hours) RN/nurse 6 LPN/nurse 2 LN/nurse 4 NA/nurse 2 Total 14 Other Total a

2

1 0 0 1

11 10 3 24

1

0

0

5

0 0

1 1

2 1

7 4

0

6

1

15

0

0

0

3

0

2

1

17

0

0

0

3

0 2

2 6

0 5

4 29

0

0

0

6

2

2

0

11

5

20

10

104

14 2 1 3 20

1 0 0 0 1

6 0 0 0 6

3 0 0 0 3

30 4 5 5 44

7

1

6

0

16

Adjusted for any one or more of the following: outpatient care, case mix, patient severity, nursing intensity weight, Medstat Resource Demand Scale, discharge day, and patient turnover.

Are Nurse Stafﬁng Measures Adequate for Health Services Research?

127

day (29 studies), FTEs per adjusted patient day (17 studies), and FTEs or numbers per patient day (15 studies). Within the stafﬁng assessment application, the predominant nursing staff to patient load ratio is that of FTEs to adjusted patient days (9 studies), whereas the most frequently used ratio in the patient outcomes application is nurse stafﬁng hours per patient day (12 studies). The most common ratio measure for evaluations of stafﬁng and nurse outcomes is evenly split between hours per patient day and the patient to nurse ratio (2 studies each). Along with FTEs or numbers per patient, the ratio of hours per patient day is commonly used (6 studies each) in evaluations of ﬁnancial performance. The most commonly used ratio in predictors of nurse stafﬁng applications is hours per patient day (5 studies). Overall, skill mix measures are used more frequently than stafﬁng counts in hospital nurse stafﬁng research (used in 44 of 123 studies). As expected, the most frequently used skill mix measure in all applications is the RN to nurse ratio. The ‘‘other’’ category of nurse stafﬁng measures is used in 16 of the 123 total studies evaluated. This measurement category includes the percent of RNs with a certain educational level, contract staff hours, the number of patient admissions per shift per unit, a ratio of provided hours to suggested hours; the number of perioperative RNs per surgical procedure; and the proportion of nurses ‘‘ﬂoating’’ outside their regular unit.

Sampling Frameworks for Studies Using Hospital Nurse Stafﬁng Measures Table 1 speciﬁes the articles utilizing various sampling frameworks for nurse stafﬁng measures. The ﬁrst part of Table 3 shows the frequency of use of each sampling framework. Overall, results show that the most frequent type of sampling is that of all hospitals within one state (used in 39 of 123 studies). Sampling from multiple hospitals (18 studies), national sample databases (17 studies), single hospitals (16 studies), and a sample of state hospitals (13 studies) follow. Single hospital unit frameworks are also fairly common. In comparing sampling across applications, it can be seen that stafﬁng assessment, patient outcomes, and ﬁnancial performance studies heavily utilize state all-hospital samples. Patient outcomes studies use national samples, single hospitals, and single hospital units more than any other application. Predictors of stafﬁng studies use multi-hospital samples more than any other research application.

128

LYNN UNRUH ET AL.

Table 3. Hospital Nurse Stafﬁng Sampling Frameworks and Data Sources by Frequency of Use within Applications and in Total. Stafﬁng Patient Nurse Financial Predictors Total Assessment Outcomes Outcomes Performance of Stafﬁng Studies Studies Studies Studies Studies Studies (N ¼ 123) (N ¼ 25) (N ¼ 53) (N ¼ 7) (N ¼ 27) (N ¼ 11) Sampling frameworka,b Single hospital unit Single hospital Multiple hospitals State, sample State, all acute-care general hospitals Multi-state National, sample National, all acutecare general hospitals Federal (e.g., VA) Other/unknown Total Data source AHA Annual Survey Database California Nursing Outcomes Coalition California Ofﬁce of Statewide Health Planning and Development Florida Agency for Health Care Administration Maryland Hospital Association Survey Multi-state data sources New York Institutional Cost Report Outcomes Research in Nursing Administration Project

0 3 0 5 13

6 9 5 5 13

0 0 5 0 1

3 4 1 3 10

0 0 7 0 2

9 16 18 13 39

0 0 4

1 12 2

0 0 0

1 3 0

0 2 0

2 17 6

0 0 25

0 0 53

1 0 7

0 2 27

0 0 11

1 2 123

10

17

1

2

2

32

4

2

0

1

0

7

6

5

0

6

1

18

0

1

0

1

0

2

0

2

0

2

0

4

0

1

0

1

0

2

1

2

0

0

1

4

0

0

1

0

2

3

129

Are Nurse Stafﬁng Measures Adequate for Health Services Research?

Table 3. (Continued ) Stafﬁng Patient Nurse Financial Predictors Total Assessment Outcomes Outcomes Performance of Stafﬁng Studies Studies Studies Studies Studies Studies (N ¼ 123) (N ¼ 25) (N ¼ 53) (N ¼ 7) (N ¼ 27) (N ¼ 11) Pennsylvania Department of Health Stafﬁng records Survey or primary data collection Other or N/A Total

6

2

1

0

0

9

3 1

15 8

4 1

5 5

2 3

29 18

0 31

1 56

0 8

2 27

0 11

3 131

a

When the sampling framework involves a sampling of data, it may be random, purposive, or convenience. b The level of analysis for these samples may be at the hospital or unit level.

Sources of Hospital Nurse Stafﬁng Data Table 1 speciﬁes the articles using various data sources for nurse stafﬁng measures. The second part of Table 3 shows the frequency of articles using these sources. The AHA Annual Survey data stand out as being the most frequently cited data source (used in 32 of 123 studies). In comparing data sources across applications, AHA was at the top of stafﬁng assessment (10 studies) and patient outcomes studies (17 studies), but was used very little in the other applications. Despite being a national data source, many studies using AHA data were focused at the state-level. The next most frequently used data source for all applications was stafﬁng records from multiple or individual hospitals (used in 29 of 123 studies). Stafﬁng records were used not only in patient outcomes studies but also in nursing outcomes and ﬁnancial outcomes studies. The use of stafﬁng records correlates with sampling frameworks of single-unit, single hospital, and multiple hospitals. State all-hospital data from ﬁve states (California, Florida, Maryland, New York, and Pennsylvania) are also used on a regular basis. The most popular state-level sources for stafﬁng assessment studies include nurse stafﬁng data from the California Ofﬁce of Statewide Health Planning and Development (OSHPD), California Nursing Outcomes Coalition (CalNOC), New York Institutional Cost Report (ICR), and the Pennsylvania Department of Health

130

LYNN UNRUH ET AL.

(PDH). All but the CalNOC data are inclusive of nearly all acute-care hospitals in the respective state. CalNOC is a convenience sample of unit-level data from voluntary reporting, whereas OSHPD is service-level data for all California hospitals. California’s OSHPD was one of the primary sources in ﬁnancial performance studies, a second most used source in stafﬁng assessment studies, but a little used source in other applications. PDH data were used mostly in stafﬁng assessment studies. Limitations of Existing Hospital Nurse Stafﬁng Data Several studies discussed the limitations of hospital nurse stafﬁng data (Table 4). Researchers note that nurse stafﬁng measures are varied and nonstandardized and that individual state-level data are limited in terms of generalizability to the U.S as a whole. Table 4 lists other general concerns related to nurse stafﬁng measures, such as the constructs of hours and FTEs, the roles included in the RN category, the service areas included in the nursing counts, and the calculations of patient load. These concerns show that many of the needs researchers have for nurse stafﬁng measures listed in the introduction are not being met. Several researchers also mentioned the lack of unit-level, shift level or daily measures of nurse stafﬁng. Unit-level measures meet many of the needs of researchers for valid stafﬁng measures in that they allow for separation of nurses by service area. Unit-level measures record more granular ﬂuctuations in stafﬁng than hospital-level data, and they provide a more direct link to patient, nurse, and ﬁnancial outcomes, which occur on a unit, rather than hospital, basis. AHA Annual Survey data limitations were speciﬁcally cited in several studies. Researchers noted that the AHA Annual Survey changed data collection procedures in 1993 and again in 2003. The exclusion of information on NAs between 1994 and 2003 posed a concern. Researchers mentioned numerous problems with the RN category in the AHA Annual Survey as well. Moreover, researchers also noted limitations of speciﬁc data sources used for state and multi-hospital sampling. For example, the CalNOC, NDNQI, and ORNAP data are proprietary data. Adequacy of Nurse Stafﬁng Measures, Samples and Data Sources As Table 1 indicates, there are three main nurse stafﬁng categories (counts, ratios, and skill mix) for measures involving RNs, LPNs, LNs, NAs, and all

Are Nurse Stafﬁng Measures Adequate for Health Services Research?

Table 4.

131

Limitations of Nurse Stafﬁng Measures Noted in the Literature.

Comments General comments Studies using data from one state not readily generalizable Nursing hours may reﬂect a different construct from FTEs FTEs may not reﬂect 40 hour work week FTE deﬁnitions and calculations are not standardized Annual payroll numbers may not reﬂect staffed positions Workload is inﬂuenced by unmeasured factors inﬂuencing nursing care intensity Lack of unit-speciﬁc and/or shift-speciﬁc measures Lack of differentiation of direct care RNs from Administrators and/or educators Lack of empirically-derived measure of nursing work intensity or workload Lack of variability in aggregated (over a year) measures Lack of standard data across states Failure to distinguish between inpatient and outpatient nurses Comments on speciﬁc data sources AHA Annual Survey Database Changed data collection procedures in 1993 and 2003 Information on NAs not available 1994–2003 RN category includes nurse practitioners after 1993 RN category includes indirect care givers and managers RN category does not distinguish between inpatient and outpatient staff RN category includes total facility RNs (e.g., nursing home) after 1993 Adjusted patient days may not reﬂect patient load AHA Annual Survey stafﬁng data is less accurate than that of OSHPD California Nursing Outcomes Coalition (Cal NOC) Not publicly available Non-random sample of California hospitals California Ofﬁce of Statewide Health Planning and Development Use of hours may not translate to FTEs National Database of Nursing Quality Indicators Not publically available Outcomes Research in Nursing Administration Project Not publicly available

Article Reference

3, 17, 18, 20, 22 14, 16, 17, 18, 110 17, 18, 111 111 17, 18, 20 18, 19 17 ,33, 65, 66, 110 17, 24, 18, 19, 20, 108 22, 25 34 64 19, 59, 110

12, 51, 110, 109 12, 51, 66, 110 12 51 59 59, 87, 110 87, 110 109

16 9, 10 14, 16 43 103, 104

132

LYNN UNRUH ET AL.

nursing staff. These categories are used across several research applications, suggesting that all three categories are needed to provide information in each of these areas. Within each of these categories, however, researchers use an array of nurse stafﬁng measures: three types of stafﬁng counts, 11 types of nursing staff/patient load, and four types of skill mix. Why do so many measures exist? Do we need this variety of measures? If not, which measures should be used? The issue of why so many measures exist appears to involve an interaction between research design and data availability. We make this observation because there are numerous data sources, each containing a limited number of speciﬁc stafﬁng measures. The AHA survey, for example, provides RN FTEs and patient days or adjusted patient days, but between 1994 and 2003 did not provide NAs. If, for example, national or multiple state data are desired to research nurse stafﬁng and patient outcomes, researchers seem to reach for the AHA data and limit stafﬁng measures to those in that dataset. If, instead, ﬁnancial outcomes related to nurse stafﬁng is the focus of research, available data may require the use of different nurse stafﬁng measures. Researchers appear to use measures that will reasonably serve their research needs given the data available to them and their need to match the stafﬁng measures to other data. That being said, it may be that not all of these measures are needed and that some would be better to use than others, especially in certain research applications. For example, patient outcomes studies are known for their inconsistent results across studies (Kane, Shamliyan, Mueller, Duval, Wilt, 2007; Unruh, 2008). It may be that the inconsistency is linked, in part, to the use of multiple stafﬁng measures with varying degrees of validity and reliability. In what follows, we review the issues that emerge from these studies with regard to nurse stafﬁng measures. Starting with counts of nurses, we note that these measures are used in all research applications, but predominantly in stafﬁng assessment studies. The most commonly used counts are FTEs, followed by nursing hours. FTEs can be problematic due to potential differences in their calculation (Reinier et al., 2005). They are a reliable count measure only if they are measured uniformly. FTEs also indicate a slightly different construct compared to hours. One FTE assumes a 40-hour week when the hours of work could be more (due to overtime) or less (due to part-time status). Hours, on the other hand, can be misleading if it is assumed that every 40 hours is equal to an FTE, when instead overtime is contributing to the indicated hours. For example, if hours are stable over time, the workforce appears stable when there was actually a reduction in the workforce under conditions of

Are Nurse Stafﬁng Measures Adequate for Health Services Research?

133

overtime. The appropriateness of using either FTEs or hours depends on the research design and methods. The third type of stafﬁng measure in the stafﬁng counts category is ‘‘numbers,’’ used in only 3 of the 123 studies. The validity of numbers as an accurate count of nurses is questionable as it does not take into account the difference between full-time and part-time status. Although the National Quality Forum (NQF) lists numbers as a numerator in one of its system measures standards (NQF, nd), we do not recommend it as a stafﬁng count measure. Of the nurse stafﬁng/patient load ratios used in nurse stafﬁng research, the most common of the eleven ratios encountered are nursing hours/patient day, FTEs/adjusted patient day, FTEs or numbers/patient, and patient or workload/nurse. Are these the best measures to be used? We addressed choice of optimal numerators in the earlier paragraphs. Regarding denominators, some research indicates that ‘‘adjusted patient days’’ is the best indicator of stafﬁng allocation available. Counting patient days (or average daily census) rather than patient stays (number of patients, number of patient admissions or discharges) is preferred because the former includes the variability of lengths of stay whereas the latter does not. When using patient days, it is best to adjust them for the allocation of nursing services between in- and out-patient care (Harless & Mark, 2006). The use of beds in the denominator is the least desirable because it does not account for the number of patients or patient days, which could vary considerably per bed. Skill mix measures are more common in nursing staff assessment, patient outcomes, and hospital ﬁnancial performance studies. RN/nurse is overwhelmingly the most popular skill mix measure. However, the literature indicates legitimate uses for the other skill mix measures and does not criticize the use of any of the skill mix measures we encountered. For the most part, these studies did not provide details about the content of nurse stafﬁng measures, particularly whether they reﬂected the intended nursing roles and nursing care services (e.g., whether they included inpatient, outpatient, or long-term care nurses). However, we know that the stafﬁng measures in the studies using AHA data are subject to the limitations listed in Table 4. Mark, Harless, McCue, and Xu (2004) note that they improved the AHA stafﬁng measures by subtracting the long-term care nurses from the total using the On-line Survey Certiﬁcation and Reporting System (OSCAR) data. Person et al. (2004) note validity concerns with the use of hospital-wide stafﬁng measures from AHA in their study of the impact of nurse stafﬁng on patients with acute myocardial infarction. Other studies using AHA data do not note limitations of, or manipulation

134

LYNN UNRUH ET AL.

of, AHA data. In their study using multi-state data sources, Needleman et al. (2002) discussed the need to standardize stafﬁng across data sets to determine the proportion of inpatient nurses. It is important to match the content of nurse stafﬁng measures with the intended construct because if the content of the stafﬁng measure was not what was intended, the validity of the study or application may be affected. For example, if the measure being used to set adequate stafﬁng levels on a unit is the number of patients per RN without regard for patient acuity, this may under- or overestimate the RNs needed to care for those patients. Similarly, if a quality reporting system that compares nurse stafﬁng levels (as a measure of quality) across hospitals does not adjust the levels of stafﬁng for both the number of patients and patient acuity, certain hospitals could look better or worse than others when they are not due to their different patient populations. In addition, without consistency in the content of the stafﬁng measures used in quality-reporting systems, some hospitals could choose a measure that would make them look better than others. In conducting research, if the intended construct is the number of acute-care hospital nurses, the measure should not include nurses associated with the long-term care section of the hospital. When the study intends to associate nurse stafﬁng with patient outcomes, the stafﬁng measure needs to match the patient population being studied. In research, it is also important to be sure that the content of measures derived from multiple data sources are consistent across sources, otherwise the reliability of the measure may be threatened. Regarding sampling frameworks, it is probable that patient outcomes studies make use of national samples, single hospitals, and single hospital units to match patient outcomes data to the stafﬁng data. For example, the Healthcare Cost and Utilization Project (HCUP) databases are the most commonly used source of state- and national-level patient outcomes data. The HCUP State Inpatient Databases (SID) have state-by-state inpatient data for 40 states, whereas the National Inpatient Sample (NIS) contains a stratiﬁed, random sample of around 1,000 hospitals. Samples are formed by matching hospital discharge data from HCUP with stafﬁng data from either state databases or the AHA database. When the SID is used, the resulting sample can be from just one state or from several. When the NIS is used, the ﬁnal sample can be the entire set of hospitals in the NIS or a sub-sample. Mark et al. (2004) used 422 hospitals in the HCUP NIS, matching the aggregated discharge data to AHA stafﬁng data, whereas Kovner et al. (2002) used samples from 6 to 13 states in the HCUP NIS, also matching the HCUP discharge data to AHA stafﬁng data.

Are Nurse Stafﬁng Measures Adequate for Health Services Research?

135

Samples for patient outcomes studies may also be formed by locating patient outcomes data on a state-by-state basis from sources other than HCUP. Needleman et al. (2002) used hospital discharge data from 11 states, matching them to stafﬁng data from each state. A few data sets such as CalNOC, NDNQI, and ORNAP provide both nurse stafﬁng and patient outcomes in the same data set. On a smaller scale, patient outcomes studies link patient data within one individual hospital or unit to stafﬁng data in the same institution or unit. The samples in these studies often use units within the hospital as the level of analysis. The matching stafﬁng data in these studies usually come from internal stafﬁng records rather than state- or national-level databases. Hospital ﬁnancial performance studies also have a distinctive sampling strategy. These studies overwhelmingly use state all-hospital data. When the sampling strategy is linked to ﬁnancial performance data sources, one reason can be seen: six studies use California’s OSHPD data. This database provides both stafﬁng and cost information and therefore is a popular source for hospital ﬁnancial performance studies. Sampling frameworks, in summary, are as varied as the measures used in nurse stafﬁng research. As with measures, sampling frameworks are linked to available data, including data other than nurse stafﬁng that is needed in the analysis. Ideally, both measures and samples would be determined only by research design needs, but that is not the reality at this time. So for the time being, most sampling strategies appear to be justiﬁed given data availability. Because of their small size, there is less justiﬁcation for the use of single hospital and hospital unit samples, but these are often used with unit-level analyses, which may provide more exact measures of both nurse stafﬁng and other outcomes data. What would best meet researchers’ needs in nurse stafﬁng research would be a national data set, or multiple state data sets, with unit-level data. Regarding sources of nurse stafﬁng data, we note ﬁrst that the AHA Annual Survey data stand out as being the most frequently cited source and were used for sampling on a national all-hospital basis, a national sample basis, and a state all-hospital basis. However, as noted previously, many researchers have expressed concerns with this data source. For example, owing to changes in the AHA Annual Survey data collection procedures in 1993, and again in 2003, caution was advised when using this data set for longitudinal studies. The changed procedures may affect the validity of measures used even in cross-sectional analyses. In addition, the RN category presents many limitations for researchers. First, it includes indirect care givers and managers, whereas researchers typically prefer to exclude, or at

136

LYNN UNRUH ET AL.

least separate, these categories of nurses from their ‘‘RN’’ counts. Second, after 1994, the data include total facility nursing staff (i.e., both acute and long-term care). Acute-care stafﬁng cannot be ascertained unless the researcher employs a creative solution such as obtaining nursing home stafﬁng data from other sources (e.g., OSCAR or the CMS POS ﬁles) and subtracting those numbers from the AHA measures (Mark, 2006). Third, there is no distinction between inpatient and outpatient RN staff. These issues may be the reason why Jiang et al. (2006) found that AHA data for California hospitals were less accurate than California’s OSHPD data. Jiang and colleagues noted that OSHPD data, on the other hand, contain nursing hours that may not translate to FTEs.

DISCUSSION This review reveals that nurse stafﬁng measures are a signiﬁcant part of health services research. Recall that we had ﬁve objectives for this review: (1) identify the hospital nurse stafﬁng measures that are commonly used in health services research and determine their frequency of use; (2) classify the sampling strategies employed; (3) categorize the data sources for stafﬁng measures; (4) consider the limitations of the data, samples, and sources that have been noted in the literature; and (5) determine whether the measures, samples, and sources meet the needs and goals of health services research, and if they do not, what would. We now consider the limitations of the measures, sampling frameworks, and sources of data found in the literature and discuss how they may be improved. Unfortunately, there are a number of limitations that affect the validity, reliability, and availability of the measures. Measures often are not adequate for the intended construct or application. Choice of measures and samples appears to be as much datadriven as design-driven. There are many non-national sources of data, while the only national source used in publications to date – the AHA Annual Survey – has many limitations. A summary of the challenges in this area was made by Mark (2006), who said that ‘‘no single, comprehensive, valid, and reliable database on nurse stafﬁng exists’’ (p. 697). In the previous section, we discussed the negative implications of this situation separately for nurse stafﬁng measures, sampling frames, and data sources. The problems we identiﬁed separately, however, may lead to a cascade of increasingly deﬁcient research and applications. For example, if inadequate stafﬁng measures are combined with inappropriate samples and data sources, then patient load could be misestimated, payments for care

Are Nurse Stafﬁng Measures Adequate for Health Services Research?

137

could lack correspondence to the care delivered, and quality indicators could be misleading. Thus, health services research results that often drive health policy would be erroneous and harm rather than help improve patient safety and outcomes, as well as negatively affect nurses and other health care providers. At least three approaches could be taken to improve this situation. The ﬁrst approach puts the onus on researchers and administrators to be fully informed about the measures they are considering and to carefully choose the measures best suited for their study or application. Sometimes information is not readily available about the measures, and additional research has to be conducted to learn more about the content of the measure. If it is discovered that data sources limit the content of the available measures so that the content does not ﬁt the intended construct, one solution is to adjust the measures to better ﬁt the construct, as Harless and Mark (2006) recommend. If that cannot be done, the construct (and therefore research design, hypotheses, analytical model, or applications) should be altered to ﬁt the measure, or the study should not be completed. An example of these choices can be made with the measure of numbers of RN FTEs. If the intended construct is to establish the number of RNs who do direct patient care in an acute-care setting such as a hospital, but the only available data include the RN educators and administrators, then the choices are to ﬁnd a way to count and subtract the educators and administrators, to change the research study or application to include educators and administrators, or to not do the study or project at all. The second approach is to improve the availability of better data. This could occur through improvements in (a) existing data sets, (b) expanded use of emerging data sets, or the (c) creation of new data sets. Improve existing data sets. Since the AHA Annual Survey is one of the most commonly used data sources, it would be a major improvement for the AHA Annual Survey to undergo changes in operational deﬁnitions and collection procedures. The points in Table 4 concerning the limitations of the AHA data set need to be addressed. For example, the measures should distinguish between RNs, LPNs, and NAs; acute and long-term care staff; inpatient and outpatient care staff; direct and non-direct care staff; regular, contract, and per diem staff; and part-time and full-time staff. In addition to these basic changes, the measures should include, or allow the calculation of, the following NQF System-Centered Measures (NQF, nd): separate nursing staff hours (RN, LPN or LVN, and unlicensed assistive personnel) per patient day; and separate proportions of RN, LPN, or LVN, unlicensed

138

LYNN UNRUH ET AL.

assistive personnel, and contracted nurse care hours to total nursing care hours (we assume that the patient day denominator recommended by NQF distinguishes between inpatients, outpatients, acute care patients, and longterm care patients). Optimally, the data would include an evidence-based nursing intensity weight that could be used to adjust patient days to better reﬂect the quantity of nursing care. Expand use of emerging data sets. Some emerging data sets that could be explored in future research are the Centers for Medicare and Medicaid Services Occupational Mix Survey (CMS OMS) and Provider of Service ﬁle (POS). The CMS OMS ﬁle includes nurse hours and wages every three years on nearly 4,000 hospitals (CMS, 2006a). These data separate RN mangers from staff. The CMS POS ﬁle, extracted from the OSCAR, has annual nurse stafﬁng data for hospitals with nursing homes (CMS, 2006b). To our knowledge, no research involving nurse stafﬁng has been conducted with these data, nor has an assessment been conducted of the validity and reliability of the measures. Create new data sets. Building a new database would be a more difﬁcult option for improving data availability, but there are some indications that a database could be developed from individual state data. This review identiﬁed four states from which hospital-level nurse stafﬁng data are available, plus seven more states mentioned in Needleman et al. (2002). It is possible that nurse stafﬁng data are being collected in other states and that these state data could be merged into a collection of state nurse stafﬁng data, similar to the HCUP SID. The issue would be whether the measures in those states are valid, reliable, and can be standardized. A survey of all states in the United States for the availability and content of nurse stafﬁng data could investigate this possibility. A third approach to improve nurse stafﬁng data would be the development of unit-level data sets. This could be an eventual result of any of the three means for improving the availability of data, but in the near future it would be beneﬁcial if the existing proprietary datasets, the CalNOC, NDNQI, and the ORNAP, all of which contain unit-level data, could be made public. The NDNQI, for example, is a convenience sample of 1,100 hospitals nationally (Montalvo, 2007). The data are unit-level measures of nurse stafﬁng, nurse satisfaction, and patient outcomes. It is important that these types of unit-level data be more readily available to researchers in the future. The three approaches we have recommended – (1) being fully informed about existing measures and their appropriate application, (2) improving the

Are Nurse Stafﬁng Measures Adequate for Health Services Research?

139

availability of better stafﬁng data, and (3) developing unit-level data sets – are not mutually exclusive and they should not be expected to be accomplished overnight. They may be pursued simultaneously and gradually. Although they may seem to be idealistic, they reﬂect the real needs and demands of nurse stafﬁng researchers.

CONTRIBUTIONS, LIMITATIONS, AND FUTURE DIRECTIONS We conclude by mentioning this study’s contributions and limitations and offer suggestions for future directions. By developing and applying a descriptive and comprehensive typology to the extant literature, we were able to assess the face validity and reliability of commonly used nurse stafﬁng measures, sample frames, and data sources. By including sampling and data source issues, we have extended previous reviews of nurse stafﬁng measures and developed three new recommendations for better using existing measures and for improving the quality and type of data available. However, many of the studies using those measures did not describe the content of the measures or reveal known issues with them. As a result, this study did (and could) not provide a rigorous evaluation of the statistical reliability and validity of nurse stafﬁng measures. Hence, we recommend further study of the adequacy of these measures. To further assess reliability, the response rates, data cleaning and other procedures that impact reliability with nurse stafﬁng measures need to be summarized and reported. The content of these measures in the various data sets also needs to be reported and discussed among users. Moreover, existing studies using nurse stafﬁng measures should be used to assess the validity of the measures. A meta-analysis of those studies could be conducted. Since each study that uses nurse stafﬁng measures has a different research design, variables, statistical methods, and other factors that would confound a simple comparison, the meta-analysis would statistically control for those factors that could cause variation in results that are not due to the measures. In conclusion, although we ﬁnd that a certain amount of variability in nurse stafﬁng measures appears to be necessary and appropriate, the use of less valid, less reliable, multiple sets of measures produces biases and errors in research and administrative practice and makes comparisons between organizations and replications of studies difﬁcult. More vigilance in the use

140

LYNN UNRUH ET AL.

of nurse stafﬁng measures and an improvement in the data and their availability would advance research and administrative practices.

REFERENCES Aiken, L. H., Clarke, S. P., Sloane, D. M., Sochalski, J., & Silber, J. H. (2002). Hospital nurse stafﬁng and patient mortality, nurse burnout, and job dissatisfaction. Journal of the American Medical Association, 288(16), 1987–1993. Brewer, C. S., & Frazier, P. (1998). The inﬂuence of structure, staff type, and managedcare indicators on registered nurse stafﬁng. Journal of Nursing Administration, 28(9), 28–36. Cho, S. H., Keteﬁan, S., Barkauskas, V. H., & Smith, D. G. (2003). The effects of nurse stafﬁng on adverse events, morbidity, mortality, and medical costs. Nursing Research, 52(2), 71–79. CMS. (2006a). Medicare wage index: Occupational mix survey. Centers for Medicare and medicaid services. Available at http://www.cms.hhs.gov/AcuteInpatientPPS/downloads/ occmix_survey_06ﬁnal.pdf CMS. (2006b). Provider of services ﬁle. Centers for Medicare and medicaid services. Available at http://www.cms.hhs.gov/NonIdentiﬁableDataFiles/04_ProviderofServicesFile.asp Donabedian, A. (1969). Some issues in evaluating the quality of nursing care. American Journal of Public Health, 59(10), 1833–1836. Donabedian, A. (1988). The quality of care: How can it be assessed? Journal of the American Medical Association, 260(12), 1743–1748. Edwardson, S. R., & Giovannetti, P. B. (1994). Nursing workload measurement systems. Annual Review of Nursing Research, 12, 95–123. Harless, D. W., & Mark, B. A. (2006). Addressing measurement error bias in nurse stafﬁng research. Health Services Research, 41(5), 2006–2024. Jiang, H. J., Stocks, C., & Wong, C. J. (2006). Disparities between two common data sources on hospital nurse stafﬁng. Journal of Nursing Scholarship, 38(2), 187–193. Kane, R. L., Shamliyan, T., Mueller, C., Duval, S., & Wilt, T. J. (2007). Nurse stafﬁng and quality of patient care. Evidence Report/Technology Assessment No. 151, AHRQ, Rockville, MD. Available at http://www.ahrq.gov/clinic/tp/nursesttp.htm Kovner, C., Jones, C., & Gergen, P. J. (2000). Nurse stafﬁng in acute care hospitals, 1990–1996. Policy, Politics & Nursing Practice, 1(3), 194–204. Kovner, C., Jones, C., Zhan, C., Gergen, P. J., & Basu, J. (2002). Nurse stafﬁng and postsurgical adverse events: An analysis of administrative data from a sample of U.S. hospitals, 1990–1996. Health Services Research, 37(3), 611–629. Mark, B. A. (2006). Methodological issues in nurse stafﬁng research. Western Journal of Nursing Research, 28(6), 694–709. Mark, B. A., Harless, D. W., McCue, M., & Xu, Y. (2004). A longitudinal examination of hospital registered nurse stafﬁng and quality of care. Health Services Research, 39(2), 279–300. McCue, M., Mark, B. A., & Harless, D. W. (2003). Nurse stafﬁng, quality, and ﬁnancial performance. Journal of Health Care Finance, 29(4), 54–76.

Are Nurse Stafﬁng Measures Adequate for Health Services Research?

141

Montalvo, I. (2007). The national database of nursing quality indicators TM (NDNQIs). Online Journal of Issues in Nursing, 12(3). Available at http://www.nursingworld. org/MainMenuCategories/ANAMarketplace/ANAPeriodicals/OJIN/TableofContents/ Volume122007/No3Sept07/NursingQualityIndicators.aspx Needleman, J., Buerhaus, P., Mattke, S., Stewart, M., & Zelevinsky, K. (2002). Nurse-stafﬁng levels and the quality of care in hospitals. The New England Journal of Medicine, 346(22), 1715–1722. NQF. (nd). Nursing care quality at NQF. Available at http://www.qualityforum.org/nursing/ O’Brien-Pallas, L., Cockerill, R., & Leatt, P. (1992). Different systems, different costs? An examination of the comparability of workload measurement systems. Journal of Nursing Administration, 22(12), 17–22. Person, S. D., Allison, J. J., Kiefe, C. I., Weaver, M. T., Williams, O. D., Centor, R. M., & Weissman, N. W. (2004). Nurse stafﬁng and mortality for Medicare patients with acute myocardial infarction. Medical Care, 42(1), 4–12. Reinier, K., Palumbo, M. V., McIntosh, B., Rambur, B., Kolodinsky, J., Hurowitz, L., & Ashikaga, T. (2005). Measuring the nursing workforce: Clarifying the deﬁnitions. Medical Care Research and Review, 62(6), 741–755. Shullanberger, G. (2000). Nurse stafﬁng decisions: An integrative review of the literature. Nursing Economics, 18(3), 124–132146–8. Spetz, J. (2000). Hospital use of nursing personnel: Holding steady through the 1990s. Journal of Nursing Administration, 30(7–8), 344–346. Spetz, J. (2004). Hospital nurse wages and stafﬁng, 1977 to 2002: Cycles of shortage and surplus. Journal of Nursing Administration, 34(9), 415–422. Unruh, L. (2001). Licensed nursing staff reductions and substitutions in Pennsylvania hospitals, 1991–1997. Journal of Public Health Policy, 22(3), 286–310. Unruh, L. (2002). Nursing staff reductions in Pennsylvania hospitals: Exploring the discrepancy between perceptions and data. Medical Care Research and Review, 59(2), 197–214, discussion 215–22. Unruh, L. (2003). Licensed nurse stafﬁng and adverse events in hospitals. Medical Care, 41(1), 142–152. Unruh, L. (2008). Nurse stafﬁng and patient, nurse, and ﬁnancial outcomes. American Journal of Nursing, 108(1), 62–71. Unruh, L. Y., & Fottler, M. D. (2006). Patient turnover and nursing staff adequacy. Health Services Research, 41(2), 599–612. Unruh, L. Y., Fottler, M. D., & Talbott, L. L. (2003). Improving nurse stafﬁng measures: Discharge day measurement in adjusted patient days of care. Inquiry, 40(3), 295–304.

Author(s)

1

Year

1996 1996a

3

Anderson, G. F., & Kohn, L. T.

1996b

4

Aydin, C. E., et al.

2004

5

Berney, B., & Needleman, J.

2005

6

Bolton, L. B., et al.

2001

7

Bond, C. A., & Raehl, C. L.

2000

8

Coffman, J. M., Seago, J. A., & Spetz, J.

2002

9

Donaldson, J. B., et al.

2001

Journal, Volume (Issue), Page Number(s)

Downsizing the hospital nursing workforce Employment trends in hospitals, 1981–1993 Hospital employment trends in California, 1982–1994 Creating and analyzing a statewide nursing quality measurement database Trends in nurse overtime, 1995–2002 A response to California’s mandated nursing ratios Changes in pharmacy, nursing, and total personnel stafﬁng in U.S. hospitals, 1989–1998 Minimum nurse-to-patient ratios in acute care hospitals in California Nurse stafﬁng in California hospitals 1998–2000: Findings from the California nursing outcomes coalition database project

Health Affairs, 15(4), 88–92 Inquiry, 33(1), 79–84 Health Affairs, 15(1), 152–158 Journal of Nursing Scholarship, 36(4), 371–378 Policy, Politics & Nursing Practice, 6(3), 183–190 Journal of Nursing Scholarship 33(2), 179–184 American Journal of HealthSystem Pharmacy, 57(10), 970–974 Health Affairs, 21(5), 53–64

Policy, Politics & Nursing Practice, 2(1), 19–28

LYNN UNRUH ET AL.

2

Aiken, L. H., Sochalski, J. & Anderson, G.F. Anderson, G. F., & Kohn, L. T.

Article Title

142

APPENDIX. LIST OF FULLY REVIEWED ARTICLES INCLUDED IN THE LITERATURE REVIEW

2005

11

Hodge, M. B., et al.

2004

12

2000

13

Kovner, C., Jones, C., & Gergen, P. J. Spetz, J.

1998

14

Spetz, J.

2000

15

Spetz, J.

2001

16

Spetz, J.

2004

17

Unruh, L.

2001

18

Unruh, L.

2002

19

Unruh, L., Fottler, M. D., & Talbott, L. L.

2003

Impact of California’s licensed nurse-patient ratios on unitlevel nurse stafﬁng and patient outcomes Licensed caregiver characteristics and stafﬁng in California acute care hospital units Nurse stafﬁng in acute care hospitals, 1990–1996 Hospital employment of nursing personnel. Has there really been a decline? Hospital use of nursing personnel: Holding steady through the 1990s What should we expect from California’s minimum nurse stafﬁng legislation? Hospital nurse wages and stafﬁng, 1977 to 2002: Cycles of shortage and surplus Licensed nursing staff reductions and substitutions in Pennsylvania hospitals, 1991–1997 Nursing staff reductions in Pennsylvania hospitals: Exploring the discrepancy between perceptions and data Improving nurse stafﬁng measures: Discharge day measurement in ‘‘adjusted patient days of care’’

Policy, Politics & Nursing Practice, 6(3), 198–210

Journal of Nursing Administration, 34(3), 125–133 Policy, Politics & Nursing Practice, 1(3), 194–204 Journal of Nursing Administration, 28(3), 20–27 Journal of Nursing Administration, 30(7–8), 344–346 Journal of Nursing Administration, 31(3), 132–140 Journal of Nursing Administration, 34(9), 415–422 Journal of Public Health Policy, 22(3), 286–310

Medical Care Research and Review, 59(2), 197–214

Inquiry, 40(3), 295–304

143

Donaldson, N., et al.

Are Nurse Stafﬁng Measures Adequate for Health Services Research?

10

144

APPENDIX. (Continued ) Author(s)

Year

Article Title

Unruh, L.

2003a

The effect of LPN reductions on RN patient load

21

Unruh, L.

2003b

22

Unruh, L., & Fottler, M. D.

2006

23

Upenieks, V. V., et al.

2007

Licensed nurse stafﬁng and adverse events in hospitals Patient turnover and nursing staff adequacy Value-added care.

24

Upenieks, V. V., et al.

2007

25

Welton, J. M., et al.

2006

26

Aiken, L. H., Smith, H. L., & Lake, E. T.

1994

27

Aiken, L. H., Clarke, S. P., & Sloane, D. M.

2000

28

Aiken, L. H., et al.

2002

Assessing nursing stafﬁng ratios: Variability in workload intensity Hospital nursing cost, billing, and reimbursement Lower Medicare mortality among a set of hospitals known for good nursing care Hospital restructuring: does it adversely affect care and outcomes? Hospital nurse stafﬁng and patient mortality, nurse burnout, and job dissatisfaction

Journal of Nursing Administration, 33(3), 201–208 Medical Care, 41(1), 142–152 Health Services Research, 41(2), 599–612 The Journal of Nursing Administration, 37(5), 243–252 Policy, Politics, & Nursing Practice, 8(1), 7–19 Nursing Economics, 24(5), 239–262 Medical Care, 32(8), 771–787 Journal of Nursing Administration, 30(10), 457–465 Journal of the American Medical Association, 288(16), 1987–1993

LYNN UNRUH ET AL.

20

Journal, Volume (Issue), Page Number(s)

2003

30

Amaravadi, R. K., et al.

2000

31

Archibald, L. K., et al.

1997

32

Barkell, N. P., Killinger, K. A., & Schultz, S. D.

2002

33

Berney, B., & Needleman, J.

2006

34

Blegen, M. A., et al.

1998a

35

Blegen, M. A., & Vaughn, T.

1998 b

36

Bolton, L. B., et al.

2003

37

Bond, C. A., et al.

1999

38

Bond, C. A., Raehl, C. L., & Franke, T.

2001

Educational levels of hospital nurses and surgical patient mortality ICU nurse-to-patient ratio is associated with complications and resource use after esophagectomy Patient density, nurse-to-patient ratio and nosocomial infection risk in a pediatric cardiac intensive care unit The relationship between nurse stafﬁng models and patient outcomes: A descriptive study Impact of nursing overtime on nurse-sensitive patient outcomes in New York hospitals, 1995–2000 Nurse stafﬁng and patient outcomes A multisite study of nurse stafﬁng and patient occurrences Nurse stafﬁng and patient perceptions of nursing care Health care professional stafﬁng, hospital characteristics, and hospital mortality rates Medication errors in United States hospitals

Journal of the American Medical Association, 290(12), 1617–1623 Intensive Care Medicine, 26(12), 1857–1862

Pediatric Infectious Disease Journal, 16(11), 1045–1048 Outcomes Management, 6, 27–33 Policy, Politics, & Nursing Practice, 7(2), 87–100

Nursing Research, 47(1), 43–50 Nursing Economics, 16(4), 196–203 Journal of Nursing Administration, 33(11), 607–614 Pharmacotherapy, 19(2), 130–138 Pharmacotherapy, 21(9), 1023–1036

145

Aiken, L. H., et al.

Are Nurse Stafﬁng Measures Adequate for Health Services Research?

29

146

APPENDIX. (Continued ) Author(s)

Year

Cho, S. H., et al.

2003

40

Czaplinski, C., & Diers, D.

1998

41

Dang, D., et al.

2002

42

Dimick, J. B., et al.

2001

43

Dunton, N., B., et al.

2004

44

Elting, L. S., et al.

2005

45

Fridkin, S. K., et al.

1996

The effects of nurse stafﬁng on adverse events, morbidity, mortality, and medical costs The effect of staff nursing on length of stay and mortality Postoperative complications: Does intensive care unit staff nursing make a difference? Effect of nurse-to-patient ratio in the intensive care unit on pulmonary complications and resource use after hepatectomy Nurse stafﬁng and patient falls on acute care hospital units Correlation between annual volume of cystectomy, professional stafﬁng, and outcomes: A statewide, population-based study The role of understafﬁng in central venous catheterassociated bloodstream infections

Journal, Volume (Issue), Page Number(s) Nursing Research, 52(2), 71–79 Medical Care, 36(12), 1626–1638 Heart and Lung, 31(3), 219–228 American Journal of Critical Care, 10(6), 376–382

Nursing Outlook, 52(1), 53–59 Cancer, 104(5), 975–984

Infection Control and Hospital Epidemiology, 17(3), 150–158

LYNN UNRUH ET AL.

39

Article Title

Haley, R. W., & Bregman, D. A.

1982

47

Haley, R. W., et al.

1995

48

Halm, M., et al.

2005

49

Hitcho, E. B., et al.

2004

50

Kovner, C., & Gergen, P. J.

1998

51

Kovner, C., et al.

2002

52

Krakauer, H., et al.

1992

53

Krauss, M. J., et al.

2005

Journal of Infectious Diseases, 145(6), 875–885

Journal of Infectious Diseases, 171(3), 614–624

Clinical Nurse Specialist, 19(5), 241–251

Journal of General Internal Medicine, 19(7), 732–739

Image–The Journal of Nursing Scholarship, 38(2), 187–193 Health Services Research, 37(3), 611–629

Health Services Research, 27(3), 317–335 Journal of General Internal Medicine, 20(2), 116–122

147

The role of understafﬁng and overcrowding in recurrent outbreaks of staphylococcal infection in a neonatal specialcare unit Eradication of endemic methicillin-resistant Staphylococcus aureus infections from a neonatal intensive care unit Hospital nurse stafﬁng and patient mortality, emotional exhaustion, and job dissatisfaction Characteristics and circumstances of falls in a hospital setting: A prospective analysis Nurse stafﬁng levels and adverse events following surgery in U.S. hospitals Nurse stafﬁng and postsurgical adverse events: an analysis of administrative data from a sample of U.S. hospitals, 1990–1996 Evaluation of the HCFA model for the analysis of mortality following hospitalization A case-control study of patient, medication, and care-related risk factors for inpatient falls

Are Nurse Stafﬁng Measures Adequate for Health Services Research?

46

148

APPENDIX. (Continued ) Author(s)

Year

Kuhn, E. M., et al.

1991

55

Landon, B. E., et al.

2006

56

Lichtig, L. K., Knauf, R. A., & Milholland, D. K.

1999

57

Manheim, L. M., et al.

1992

58

Mark, B. A., et al.

2004

59

Mark, B.A., & Harless, D.W.

2007

60

Mark, B.A., Harless, D. W., & Berman, W. F.

2007

61

McKay, N. L., & Deily, M. E.

2005

The relationship of hospital characteristics and the results of peer review in six large states Quality of care for the treatment of acute medical conditions in US hospitals Some impacts of nursing on acute care hospital outcomes Regional variation in Medicare hospital mortality A longitudinal examination of hospital registered nurse stafﬁng and quality of care Nurse stafﬁng, mortality, and length of stay in for-proﬁt and not-for-proﬁt hospitals Nurse stafﬁng and adverse events in hospitalized children. Comparing high- and lowperforming hospitals using risk-adjusted excess mortality and cost inefﬁciency

Journal, Volume (Issue), Page Number(s) Medical Care, 29(10), 1028–1038

Archives of Internal Medicine, 166, 2511–2571 Journal of Nursing Administration, 29(20), 25–33 Inquiry, 29(1), 55–66 Health Services Research, 39(2), 279–300 Inquiry, 44, 167–186

Policy, Politics, & Nursing Practice, 8(2), 83–92 Health Care Management Review, 30(4), 347–360

LYNN UNRUH ET AL.

54

Article Title

Minnick, A. F., & Pabst, M. K.

1998

63

Morales, I. J., et al.

2003

64

Needleman, J., et al.

2002

65

Newhouse, R. P., et al.

2005

66

Person, S. D., et al.

2004

67

Potter, P. N., et al.

2003

68

Pronovost, P. J., et al.

1999

69

Ritter-Teitel, J.

2004

70

Robertson, R. H., & Hassan, M.

1999

Improving the ability to detect the impact of labor on patient outcomes Hospital mortality rate and length of stay in patients admitted at night to the intensive care unit Nurse-stafﬁng levels and the quality of care in hospitals

New England Journal of Medicine, 346(22), 1715–1722 AORN Journal, 81(3), 508–509, 513–522, 525–528 Medical Care, 42(1), 4–12

Nursing Economics, 21(4), 158–166

Journal of the American Medical Association, 281(14), 1310–1317 Journal of Nursing Administration, 34(4), 167–169 Health Services Management Research, 12(4), 258–268

149

Perioperative nurses and patient outcomes-mortality, complications, and length of stay Nurse stafﬁng and mortality for Medicare patients with acute myocardial infarction Identifying nurse stafﬁng and patient outcome relationships: a guide for change in care delivery Organizational characteristics of intensive care units related to outcomes of abdominal aortic surgery Registered nurse hours worked per patient day: The key to assessing stafﬁng effectiveness and ensuring patient safety Stafﬁng intensity, skill mix and mortality outcomes: The case of chronic obstructive lung disease

Journal of Nursing Administration, 28(12), 17–21 Critical Care Medicine, 31(3), 858–863

Are Nurse Stafﬁng Measures Adequate for Health Services Research?

62

150

APPENDIX. (Continued ) Author(s)

Year

Article Title

Seago, J. A., & Ash, M.

2002

Registered nurse unions and patient outcomes

72

Seago, J. A., Williamson, A., & Atwood, C.

2006

73

Shultz, M. A., et al.

1998

74

Sochalski, J.

2004

75

Stone, P. W., et al.

2007a

76

Unruh, L., Joseph, L., & Strickland, M.

2007

77

Whitman, G. A., et al.

2002

Longitudinal analyses of nurse stafﬁng and patient outcomes: More about failure to rescue The relationship of hospital structural and ﬁnancial characteristics to mortality and length of state in acute myocardial infarction patients Is more better? The relationship between nurse stafﬁng and the quality of nursing care in hospitals Nurse working conditions and patient safety outcomes Nurse absenteeism and workload: negative effect on restraint use, incident reports and mortality The impact of stafﬁng on patient outcomes across specialty units

Journal of Nursing Administration, 32(3), 143–151 Journal of Nursing Administration, 34(5), 228–237 Outcomes Management Nursing Practice, 2, 130–136

Medical Care, 42(2 Suppl), II67–73

Medical Care, 45(6), 517–578 Journal of Advanced Nursing, 60(6), 673–681

Journal of Nursing Administration, 32(12), 633–639

LYNN UNRUH ET AL.

71

Journal, Volume (Issue), Page Number(s)

Clarke, S. P., et al.

2002

79

Clarke, S. P., Sloane, D. M., & Aiken, L. H.

2002

80

Lanza, M. L., et al.

1997

81

Mark, B. A., et al.

2007

82

Stone, P. W., et al.

2007b

83

Stone, P. W., & Gershon, R. R. M.

2006

84

Behner, K. G., et al.

1990

85

Bloom, J. R., Alexander, J. A., & Nuchols, B. A.

1997

Organizational climate, stafﬁng, and safety equipment as predictors of needlestick injuries and near-misses in hospital nurses Effects of hospital stafﬁng and organizational climate on needlestick injuries to nurses Stafﬁng of inpatient psychiatric units and assault by patients

American Journal of Public Health, 92(7), 1115–1119 Journal of American Psychiatric Nurse’s Association, 3, 42–48 Journal of Safety Research, 38, 431–446

Health Services Research, 42(3), 1085–1104

Policy, Politics, & Nursing Practice, 7(4), 240–247 Health Care Management Review, 15(4), 63–71

Social Science & Medicine, 44(2), 147–155

151

Does safety climate moderate the inﬂuence of stafﬁng adequacy and work conditions on nurse injuries Nurse working conditions, organizational climate, and intent to leave in ICUs: An instrumental variable approach Nurse work environments and occupational safety in intensive care units Nursing resource management: analyzing the relationship between costs and quality in stafﬁng decisions Nurse stafﬁng patterns and hospital efﬁciency in the United States

American Journal of Infection Control, 30(4), 207–216

Are Nurse Stafﬁng Measures Adequate for Health Services Research?

78

152

APPENDIX. (Continued ) Author(s)

Year

Brown, M. P., Sturmna, M. C., & Simmering, M. J.

2002

87

McCue, M., Mark, B. A., & Harless, D. W. Needleman, J., et al.

2003

1997

90

Robertson, R. H., Dowd,S. B., & Hassan, M. Robertson, R. H., & Hassan, M.

1999

91

Rothberg, M. B., et al.

2005

92

Seago, J. A., Spetz, J., & Mitchell, S.

2004

93

Shamian, J., B., et al.

1994

88

89

2006

The beneﬁts of stafﬁng and paying more: The effects of stafﬁng levels and wage practices for registered nurses on hospitals’ average lengths of stay Nurse stafﬁng, quality, and ﬁnancial performance Nurse stafﬁng in hospitals: Is there a business case for quality? Skill-speciﬁc stafﬁng intensity and the cost of hospital care Stafﬁng intensity, skill mix and mortality outcomes: The case of chronic obstructive lung disease Improving nurse-to-patient stafﬁng ratios as a costeffective safety intervention. Nurse stafﬁng and hospital ownership in California The relationship between length of stay and required nursing care hours

Journal, Volume (Issue), Page Number(s) Advances in Health Care Management, 3, 45–57

Journal of Health Care Finance, 29(4), 54–76 Health Affairs, 25(1), 204–211 Health Care Management Review, 22(4), 61–71 Health Services Management Research, 12(4), 258–268 Medical care, 43(8), 785–791 Journal of Nursing Administration, 34(5), 228–237 Journal of Nursing Administration, 24(7–8), 52–58

LYNN UNRUH ET AL.

86

Article Title

2001

95

Titler, M., et al.

2005

96

Titler, M., et al.

2007

97

Arndt, M., & Crane, S.

1998

98

Berney, B., Needleman, J., & Kovner, C.

2005

99

Blegen, M. A., Vaughn, T., & Vojir, C. P.

2007

100

Brewer, C. S., & Frazier, P.

1998

101

Lake, E. T., & Friese, C. R.

2006

102

Lindrooth, R. C., et al.

2006

103

Mark, B. A., Salyer, J., & Wan, T. T.

2000

What should we expect from California’s minimum nurse stafﬁng legislation? Cost of hospital care for elderly at risk of falling. Cost of care for seniors hospitalized for hip fracture and related procedures. Inﬂuences on nursing care volume Factors inﬂuencing the use of registered nurse overtime in hospitals, 1995–2000 Nurse stafﬁng levels: Impact of organizational characteristics and registered nurse supply The inﬂuence of structure, staff type, and managed-care indicators on registered nurse stafﬁng Variations in nursing practice environments: Relation to stafﬁng and hospital characteristics The effects of changes in hospital reimbursement on nurse stafﬁng decisions at safety net and nonsafety net hospitals Market, hospital, and nursing unit characteristics as predictors of nursing unit skill mix: A contextual analysis

Journal of Nursing Administration, 31(3), 132–140 Nursing Economics, 23(6), 290–306 Nursing Outlook, 55(1), 5–14 Journal of the Society for Health Systems, 5(4), 38–49 Journal of Nursing Scholarship, 37(2), 165–172 Health Services Research, 43(1), 154–173 Journal of Nursing Administration, 28(9), 28–36 Nursing Research, 55(10), 1–9

Health Services Research, 41(3), 701–711

Journal of Nursing Administration, 30(11), 552–560

153

Spetz, J.

Are Nurse Stafﬁng Measures Adequate for Health Services Research?

94

154

APPENDIX. (Continued ) Author(s)

Year

Mark, B. A., Salyer, J., & Harless, D. W.

2002

105

Seago, J. A., Spetz, J., & Mitchell, S.

2004

106

Shamian, J., Hagen, B., Hu, T. W., & Fogarty, T. E.

1994

107

Zhang, N., et al.

1999

108

Harless, D. W., & Mark, B. A.

2006

109

Jiang, H. J., Stocks, C., & Wong, C. J.

2006

110

Mark, B. A.

2006

111

Reinier, K., et al.

2005

What explains nurses’ perceptions of stafﬁng adequacy? Nurse stafﬁng and hospital ownership in California The relationship between length of stay and required nursing care hours The effect of managed care on hospital stafﬁng and technological diffusion Addressing measurement error bias in nurse stafﬁng research Disparities between two common data sources on hospital nurse stafﬁng Methodological issues in nurse stafﬁng research Measuring the nursing workforce: Clarifying the deﬁnitions

Journal, Volume (Issue), Page Number(s) Journal of Nursing Administration, 32(5), 234–242 Journal of Nursing Administration, 34(5), 228–237 Journal of Nursing Administration, 24(7–8), 52–58 Health Policy, 48(3), 189–205 Health Services Research, 41(5), 2006–2024 Journal of Nursing Scholarship, 38(2), 187–193 Western Journal of Nursing Research, 28(6), 694–709 Medical Care Research and Review, 62(6), 741–755

LYNN UNRUH ET AL.

104

Article Title

MATRIX MENTORSHIP IN ACADEMIC MEDICINE: SUSTAINABILITY OF COMPETITIVE ADVANTAGE$ Jay A. Fishman ABSTRACT The healthcare system is undergoing rapid change as medical centers are confronted with constricted reimbursements for healthcare services while adapting to growth in medical knowledge, major technological advances in medical practice, and a changing regulatory environment. Academic medical centers thought themselves immune to the forces that shape most service enterprises but are forced to compete based on customer service and the efﬁciency, quality, and safety of medical care, while continuing to compete in the academic world. These challenges are not unique to academic medicine, but these institutions are, perhaps, least suited to the leadership challenges posed by this environment.

$

Dr. Fishman reports no conﬂicts relevant to this manuscript. Portions of this discussion were presented by Dr. Fishman in the President’s address for the American Society of Transplantation, Seattle Washington, 2005.

Biennial Review of Health Care Management: Meso Perspectives Advances in Health Care Management, Volume 8, 155–170 Copyright r 2009 by Emerald Group Publishing Limited All rights of reproduction in any form reserved ISSN: 1474-8231/doi:10.1108/S1474-8231(2009)0000008010

155

156

JAY A. FISHMAN

Certain attributes of these centers raise barriers to successful adaptation to the changing healthcare environment. The need for systemic change in academic medicine requires commitment to programs that create change agents willing to assume leadership roles and to guide institutional evolution. In academic medicine, traditional one-on-one relationships between mentors and trainees do not provide the breadth of guidance needed in the complex environment of research, medical practice, and teaching. A structured system of ‘‘matrix mentorship’’ and structured evaluation will advance institutional values, provide leaders with an essential set of skills and values consistent with institutional goals, and provide competitive advantage for medical centers in academic healthcare.

INTRODUCTION Academic medical centers provide state-of-the-art medical care, train young physicians and medical sub-specialists, and perform basic and clinical research. These centers are increasingly buffeted by external forces including an explosive growth in medical knowledge, technological advances in medical practice, the changing regulatory environment, and constricted reimbursements for healthcare. Institutions are increasingly forced to focus on customer service and the efﬁciency, quality, and safety of medical care, while competing for patients, staff, and ﬁnancial support. These challenges span many service enterprises. However, certain attributes of academic medical centers, including tenure systems and antipathy for leadership roles, raise barriers to adaptation to the changing healthcare environment (Lorsch, 1986). Adaptation to change requires the development of leadership throughout each organization to support institutional values and priorities and to maintain technological excellence and innovation. Investment in formal programs that provide mentorship for trainees and junior level faculty are essential to create alignment with institutional strategies. The demands of academic medicine require leaders who have received interdisciplinary training that exceeds the skill set of most individual mentors. Such integrated development can be achieved through a program of ‘‘matrix mentorship’’ that provides support, evaluation, and opportunities for junior staff (Fig. 1). Such programs may provide a sustainable competitive advantage for the institution.

157

Matrix Mentorship in Academic Medicine

Clinical, Educational Mentors

Institutional Values, Beliefs, Mission, Boundaries

Family, Friends

Peers, Colleagues

Grant Reviewers Protegé

Professional Societies

Research Mentors

“Life” Mentors

Patients

Education

Clinical Expertise

Research

Administration

Metrics Publications Evaluations Lectures Promotion

Patient Outcomes Technical Skills New Techniques

Publications Grant Support Patents issued Employment

Evaluations Leadership and Managerial Skills

Fig. 1. Matrix Mentorship. Multiple Mentors are Needed for Academic Trainees to Guide All Aspects of Professional and Psychosocial Development. A Formal Program in Mentorship Must Include Measurable Parameters to Assess the Success of Mentorship and the Development of Skills by Individuals Participating in Such Programs.

CHALLENGES OF THE CHANGING HEALTHCARE ENVIRONMENT Major cultural changes in the healthcare system have occurred relatively suddenly. Cost containment and improvements in the quality of patient care have been emphasized by the Institute of Medicine as part of the process of ‘‘creating value’’ in healthcare. The rising cost of healthcare has stimulated demands for improved efﬁciency in the delivery of healthcare. These costs are greater in academic medical centers than in community hospitals due to investments in advanced technologies, specialty care, training programs, basic research laboratories, clinical and research information systems, and associated administrative personnel, all of which increase institutional overhead costs. Medical centers must also respond to increasing public and

158

JAY A. FISHMAN

governmental scrutiny of the quality and safety of medical care. Reimbursement is linked to quality measures and scarce resources are diverted to quality assurance and compliance programs. Physicians are increasingly burdened with administrative responsibilities. None of these efforts are adequately reimbursed. Technological advances have also altered clinical practice. For example, the human genome project and advances in molecular biology allow analysis of disease pathways and exploration of individual’s susceptibility to disease. However, these advances require considerable investment in clinical laboratories and in personnel speciﬁcally trained in genetic technologies and in the use of information technology to decipher genetic data. This shifting healthcare landscape requires that academic medical centers be ﬂexible – agile enough to anticipate or respond to changes in regulations or technology. The keys to making institutional change achievable have been widely dissected in the realm of business (Kotter, 1996). The alignment of staff with institutional attitudes, values, and technologies requires leadership distributed throughout the organization to champion change and to support innovation and learning (Argyris, 1991, 1999). ‘‘Depth’’ of institutional leadership is the core ‘‘asset’’ required for competitive advantage. At present, the competitive strength of the academic institution depends largely on the degree to which the interests of the organization and those of its ‘‘stars’’ are aligned. If the stars and leaders perform in ways that advance organizational goals, it is likely that more junior staff will emulate these behaviors to the beneﬁt of the center (Lorsch & Tierney, 2002). However, these groups are often misaligned (discussed below) necessitating alternative strategies to advance programmatic goals. A formalized program of mentorship serves these purposes. It requires the articulation of institutional values and goals and serves to align participants, both stars and other employees, with those goals. The ability to sustain competitive advantage relies on an investment in the development of committed leaders and innovators throughout the organization.

BARRIERS TO INSTITUTIONAL ALIGNMENT AND THE NURTURING AND RETENTION OF ‘‘STARS’’ Modern academic medical institutions recruit faculty to direct externally funded biomedical research programs or to provide expertise in highly complex, technical procedures (e.g., stem cell transplantation). Such offerings provide a basis for differentiation of centers and, thus, create one source of

Matrix Mentorship in Academic Medicine

159

sustainable competitive advantage (O’Reilley & Pfeffer, 2000). Furthermore, advanced medical procedures are often reimbursed by insurance companies at higher rates (Porter, 1996). As a result, academic medical centers must compete to attract increased numbers of patients to develop the technical proﬁciency in advanced techniques and to compensate the centers for the upfront investment in high-tech equipment and for the recruitment and retention of expert faculty who direct these innovative programs. The satisfaction and retention of medical stars at academic medical institutions are increased by ‘‘tasks’’ that are inherently challenging and well deﬁned. Critically ill patients, research publications, grant applications, and academic promotions provide clear parameters for achievement. Alignment around ‘‘providing superior clinical care,’’ ‘‘life-long learning,’’ or integrity in science is relatively straightforward (Hill, 2003). In these areas, the goals of the individual and the academic medical center coincide. Unfortunately, academic stars are often ambivalent, or worse, about administrative leadership roles. One component of this ambivalence may be that when thrust into leadership roles, few know what to expect from the new role or have been adequately trained for their new organizational responsibilities. Becoming a manager also implies responsibility for the productivity and performance of other professionals and staff that may detract from the specialties for which they were recruited (Hill, 2003). Furthermore, recruited faculty members are generally highly competitive professionals who have been successful in achieving individual recognition for their research, publication records, external grant support, or as teachers. Institutional programs emphasizing interdependency and shared responsibility may be shunned. Delong et al. address the problem of alignment with corporate goals and strategies by ‘‘high-performance individuals’’ or ‘‘high need for achievement personalities’’ (Delong, Gabarro, & Lees, 2007). Such individuals are less likely to ‘‘buy-in’’ to institutional needs such as marketing to referring physicians, attention to the psychosocial needs of trainees, use of electronic medical records, or routine hand cleansing. They are generally disinterested in corporate programs aimed at non-academic issues such as cost containment, customer satisfaction, interdisciplinary clinical care, teaching, administration, or business or professional development. Their commitment to institutional strategic goals may be suspect. Unfortunately, these stars also need a great deal of nurturing. Despite admirable records of achievement and recognition in a competitive environment, they require constant reassurance and feedback. They require autonomy, loathe bureaucracy, and are highly skilled, but need a sense of inclusion and importance in the organization (Delong et al., 2007). Younger stars have additional needs for training,

160

JAY A. FISHMAN

feedback, coaching, career guidance, and to be challenged by increasingly difﬁcult (‘‘stretch’’) assignments that provide the greatest opportunities for learning, growth, and visibility (Hill, 2003). The potential for competition between younger and older physicians may detract from the collegial environment needed to develop young leaders. Coalescence around corporate goals is complicated by the multiple generations of stars that coexist at academic medical centers, particularly as retirement occurs at later ages. Highly skilled, young physician-scientists (in their 20s and 30s) have beneﬁted from rapid technological advances in science, particularly in information technology, molecular biology and genetics, immunology, and related disciplines. They are often highly specialized and, in comparison with their elders, highly focused – which provides a competitive advantage in gaining research grant support and professional recognition. Physicians in their 40s and 50s are ready to accept leadership responsibilities and may have attained signiﬁcant professional recognition. However, their advancement is constrained on one side by older leaders who are protecting their ‘‘turf ’’ and on the other by younger professionals whose career advancement is often astronomical. Given limited resources (space, money, time), support of younger staff (‘‘new stars’’) occurs at the expense of other priorities including more senior physicians. Thus, senior academic physicians may see their younger colleagues as competitors. This competition may be augmented in academia where a tenure system may protect the employment of senior staff but not their research space or ﬁnancial support. Administrative structures of academic medical center are not generally conducive to the adoption of new initiatives that require active participation by diverse faculties. Academic hierarchies are generally complex matrix structures with clinical activities dependent on both departmentalization by specialty (e.g., Medicine, Surgery) and also by area of research (e.g., immunology). Some clinical areas are also centered on a service line (e.g., organ transplantation), crossing multiple medical and surgical subspecialties while professional accountability may remain divided between their departmental leadership and clinical or research ‘‘centers.’’ Thus, physicians generally report to multiple individuals. As a result, academic medical centers need interdepartmental mechanisms designed to create acceptance of institutional values and cultural change while supporting innovation (research) and training of medical and scientiﬁc experts (Table 1). Given the demands of academic life – clinical, research, teaching, and administration – senior staff lack the time to devote to mentoring of younger physicians. Furthermore, achievements in teaching, training, and

Matrix Mentorship in Academic Medicine

Table 1.

161

Institutional Goals: Sustainable Competitive Advantage.

Leadership throughout the organization Alignment of employees with corporate values and goals Reduced inertia in senior personnel Succession planning Achievement of diversity Technical skill maintenance Constant learning Strategic management of talent – supervision of junior staff, optimization of employees lacking in technical skills or corporate values Weed out failures (programs or individuals) Control system for empowerment of employees – innovation within boundaries

mentorship for junior staff and contributions to institutional goals are difﬁcult to assess and may detract from professional promotion. In the absence of the time or the inclination on the part of senior staff to develop nurturing relationships on a personal level, the institution must provide an alternative support structure to develop and retain talent (DeWitt, Curtis, & Burke, 1998; Epstein, 1999; Ognibene, 2007; Ross, 1984). For this support structure to be effective, the institution must also endorse a clear set of values, skills, and behaviors of proven leaders that can be used to guide the mentoring process (Ready & Conger, 2003).

MENTORSHIP AND ORGANIZATIONAL PERFORMANCE In a changing and highly competitive environment, it is increasingly important that management be able to translate strategy into operational behavior. For any professional service provider to be competitive, they must provide an increasingly diverse set of unique and customer-speciﬁc services while simultaneously reducing costs. The effort to increase billable service time is often at the expense of investment in employee development – much to the detriment of the long-term competitive advantage of the corporation. By contrast, programs that reinforce corporate culture (values, philosophy, ethics) and demonstrate a corporate commitment to the development of leaders will beneﬁt from a climate of creativity in which innovation, new technologies, and risk taking ﬂourish (McCauley, 2007; Pfeffer, 1994, 2005). These advantages provide the ﬂexibility to change with the business environment and to create unique value for customers (Lorsch, 1986; Pfeffer,

162

JAY A. FISHMAN

1994). Such a corporate environment cannot be easily duplicated elsewhere and results in retention of both skilled employees and valuable clients. The institutionalization of mentorship for healthcare professionals is one component of a program that can address the growing need for change agents within academic medical centers and achieve alignment of staff with organizational (rather than individual) goals (McCauley, 2007; Wilson & Elman, 1990). The highest performing individuals generally have the ability to attract mentors to support their own development. They are nurtured by the institution to be productive in their areas of expertise. Given their academic ‘‘star’’ value, deﬁciencies in commitment to the broader agenda are often overlooked. Thus, the maintenance of the clinical and research enterprise falls on the majority of institutional employees who are more ‘‘average’’ – highly functioning and generally highly dedicated, but without star billing. These individuals, by default, must be targeted as the carriers of institutional values. Unfortunately, the average employee often fails to attract mentors who will nurture their development, drive alignment with institutional goals, or aid efforts to become stars in their own right (Delong et al., 2007; Hill, 2003). Mentorship is a valuable tool in professional development, diffusion of leadership skills, and alignment with corporate values (Chao, Walz, & Gardner, 1992; Wilson & Elman, 1990). However, mentorship is not a goal in itself – rather a component of the ‘‘learning environment’’ in which employees are nurtured and trained (Argyris, 1991, 1999). The optimal structuring of mentorship programs is poorly deﬁned. Different employees have different developmental needs. Programs that address these differences enhance the value of each employee. As for mentorship, the best learning and growth experiences (‘‘stretch assignments’’) generally fall to the star employees (‘‘A players’’ per DeLong et al., 2007) who easily attract resources and organizational support (Conger, 2004; Conger & Benjamin, 1999; Delong et al., 2007; DeLong, Gabarro, & Lees, 2008; Hill, 2003; McCall, 1997). As was noted, these individuals are often excused from participation in corporate alignment programs and may not conform to corporate philosophies. In the cultural environment of academic medicine as in other ﬁelds, ‘‘star worship’’ is often difﬁcult to overcome, but is to the long term detriment of the corporation both in terms of alignment of values and the beneﬁts of evaluative interactions that serve to ‘‘consolidate the lessons’’ of stretch experiences (Argyris, 1991; Hill, 2003; Kram, 1985, 1988). By contrast, development programs for the average employee increases the likelihood that ‘‘B’’ players with ‘‘A’’ (leadership) potential will get recognized and supported and that ‘‘C players’’ who detract from

Matrix Mentorship in Academic Medicine

163

corporate performance can be identiﬁed (Delong et al., 2007; DeLong et al., 2008). Mentorship programs demonstrate the company’s commitment to individual development and create an environment that increases loyalty and decreases turnover. If mentorship is an ‘‘accidental’’ pairing of teachers and prote´ge´s, and only the top employees attract suitable mentors, how does a company proceed? Hill, Higgins, and Kram, based on social networking theory, advocate the cultivation of multiple and diverse developmental networks rather than awaiting the ‘‘perfect’’ mentor (Higgins & Kram, 2001; Hill, 2003; Kram, 1988; Kram & Isabella, 1985). These multiple developmental relationships are often more transient than classical mentorships. The creation of sequential developmental relationships is driven by an individual’s changing needs and professional contacts over time and their personal ability to attract mentors. Thus, this approach is disadvantageous to employees less skilled in attracting mentors and/or optimal assignments – the average employee. It may also be less efﬁcient in achieving alignment with corporate culture (Hill, 2003). In contrast with individual relationships, group mentorship may more rapidly identify and rectify gaps in an individual’s skill set. Once a prote´ge´ incorporates a mentor’s offerings, others in the developmental ‘‘team’’ may identify attributes needed to move to the next level. Multiple relationships may provide a comfort level with advice – greater comfort receiving critique from peers than from superiors (Kram & Isabella, 1985). As a result, networks of relationships enhance communication and collaboration and provide oversight of the effectiveness and ethics of individuals comprising a group (Cohen & Bradford, 1989, 2005; Hill, 2003; Kotter, 1985, 1995; O’Reilly & Pfeffer, 2000). Some such networks have been internet-based, allowing continuous interactions between mentor and prote´ge´ over time (Whiting & de Janasz, 2004). The recognition of interdependency between employees also aligns divergent interests of the members of such groups. Thus, it serves the needs of a corporation to encourage the development of supportive relationships at all levels. In academic medical centers, clear lines of authority and guidelines for training are often absent. Thus, formalized programs for evaluation and training of faculty members are needed to assess adherence to departmental and corporate goals. A formalized program of mentorship will increase alignment around institutional values and provide the advantages of mentorship to a broader group of individuals (Chao et al., 1992). Participation will require institutional commitment to these programs and incentives (ﬁnancial, promotion, protected time) to compensate for time ‘‘lost’’ in organizational initiatives (Ragins & Scandura, 1999).

164

JAY A. FISHMAN

MATRIX MENTORSHIP AND ACADEMIC MEDICINE Mentors come in various forms such as Yoda in Star Wars or Gandalf in Tolkein’s Lord of the Rings trilogy. Many are not fortunate enough to have one of these godly and magical ﬁgures (apparently a prerequisite) as mentors. The mentor is less a teacher than a guide for a journey toward attainment of knowledge and a productive and satisfying career. Relationships formed between mentors and their prote´ge´s are largely based on luck and personal compatibility. In academia, ﬁnding a mentor is often more easily accomplished in the research environment where close proximity and frequent contacts between senior scientists and post-doctoral fellows in their laboratories support the development of scientiﬁc protocols and preparation of manuscripts (Souba, 2000). The process is more difﬁcult, but equally beneﬁcial, for physicians in clinical, educational, and administrative careers. In the absence of self-selected mentors, multiple studies suggest that assigned mentors in carefully structured programs can provide many of the same beneﬁts in the academic environment (Allen, Eby, & Lentz, 2006; Allen, Eby, Poteet, Lentz, & Lima, 2004; Curtis, Adam, & Shelov, 1995; Morzinski, 2005; Morzinski, Diehr, Bower, & Simpson, 1996; Morzinski, Simpson, Bower, & Diehr, 1994). These beneﬁts may be observed in both clinical and basic research careers (Blixen, Papp, Hull, Rudick, & Bramstedt, 2007; Marx, 2006). In addition, mentorship enhances job satisfaction and thus retention of valuable health professionals (Gilster & Accorinti, 1999; Ramanan, Phillips, Davis, Silen, & Reede, 2002). Mentoring relationships are generally seen as enhancing the productivity of the mentor as well as their prote´ge´s. Mentorship has been shown to have a positive impact on career development in academic medicine (Blackburn & Fox, 1976; Bland & Schmitz, 1986; Chesney et al., 2001; Souba, 2000; Whitworth, 2007). Mentors serve as role models in clinical skills and compassion, while teaching clinical skills and providing inspiration, psychosocial support, career guidance, and self-conﬁdence (Bhagia & Tinsley, 2000; Bland & Schmitz, 1986; Wright, 1996). Successful mentors provide frequent communication with their prote´ge´s to follow career progress and offer research and career advice (Wright, 1996). They identify opportunities for mentees to demonstrate independence and presentation skills and to develop professional networks. As a result, mentees have better research careers, more publications, and tend to have greater self-conﬁdence (Bland & Schmitz, 1986; Hollingsworth, 2002; Palepu et al., 1998; Ramanan, Taylor, Davis, & Phillips, 2006; Riechelmann, Townsley, Pond, & Siu, 2007). This applies both to research and to

Matrix Mentorship in Academic Medicine

165

clinical careers (DeWitt et al., 1998; Frohlich, 2000; Rock, 1999; Rogers, Holloway, & Miller, 1990). Selwa suggests that mentorship is an ‘‘essential catalyst for a successful medical career’’ (Selwa, 2003). In many ways, academic medicine – clinical, research, and education – has become too complex for a traditional ‘‘one mentor per trainee’’ approach (Barondess, 1997). The ‘‘triple threat physicians’’ (clinician, researcher, teacher) of the past have been increasingly replaced by technical specialists. Solo mentorship, while personally advantageous, risks that trainees become too narrow and unsuited for the interdisciplinary nature of clinical practice, biomedical research, and administration. Single mentorship limits the trainee to knowledge and experiences carried by the mentor. Alternatively and commonly, over time, the prote´ge´ ‘‘outgrows’’ the mentor and moves on to other relationships. The goal of interdisciplinary training is not easily reached – for example, the best person to teach technical skills may not be the best for behavioral growth. Such multiple roles often require multiple mentors – a ‘‘matrix mentorship’’ (see Fig. 1). Having each trainee linked to multiple mentors provides support, continuous assessment, and opportunity for development. The institution can select physicians willing to work in interdisciplinary groups that could serve as mentors for each trainee. These mentors would share institutional goals and values and a ‘‘team mentality.’’ The institution must ‘‘value’’ mentorship and provide clear guidelines for these relationships. Even so, mentorship is time-consuming and not always rewarding (i.e., not promotion-friendly for the mentor). As noted in Fig. 1, incentives need to be provided to the mentors in terms of time and/or money and support for academic promotion (Connor, Bynoe, Redfern, Pokora, & Clarke, 2000). The goal of the system is to utilize mentorship-type relationships to optimize training and education, support institutional values, priorities and initiatives, and to identify those with advanced leadership potential. The complexity of the interdependent relationship between mentor and prote´ge´ cannot be created by ﬁat. Not all relationships will function equally well. To assess such a program requires the development of metrics – a scorecard for each mentor and their trainees integrated into routine evaluations to measure the effectiveness of the program and the mentor (Djerassi, 1999; Maudsley, 2001; McCauley, 2007; Motta, 2002; Schindler, Winchester, & Sherman, 2002). The scorecard outlined in Fig. 1 includes measurements of commitment to institutional values as well as professional skills, commitment, process excellence, and productivity. Such metrics must undergo rigorous reevaluation and revision to assure continued relevance and adherence (Kram & Isabella, 1985; Ready & Conger, 2003). The scorecard must link leadership development to business outcomes.

166

JAY A. FISHMAN

A commitment to a mentorship program requires the diversion of resources from other commitments. In academic medicine as in other service industries, measurements of customer satisfaction, quality assurance programs, cost saving programs, and regulatory compliance generally dominate budget discussions. However, these programs will have little traction without employee leadership to achieve institution-wide acceptance. Thus, allocation of resources to mentorship programs is essential to meeting of the challenges posed by the current environment. One approach links the development of a matrix mentorship program with each new institutional initiative. Thus, as resources are dedicated to each, for example, quality assurance or compliance program, a network of junior and senior clinical and administrative staffs aligned with the project is constructed with the simultaneous goals of project completion and group mentorship. Compensation is provided for the time spent on the project and in personnel development. The participating groups would be vetted by human resource personnel before and during the lifespan of each project. As the institution reallocates resources to each group, individuals with leadership capacity are selected for future leadership roles. Such a system is based on a new cultural reality – all professional staff will be expected to participate in the development of institutional programs, in the mentorship of junior staff, and in a continuous evaluation system. This will enhance not only buy-in by disparate staff members but also communication (team building) between administrative and professional staffs.

CONCLUSION Change in academic medicine, as in all professions, requires leadership and commitment throughout each organization. Academics are generally untrained and, by temperament, in many ways unsuited for major leadership roles. Leadership must be emphasized at all professional levels. Leadership development programs require that the institution has a clearly deﬁned mission statement and values and standards for performance. These values include the construction of the next generation of leaders; talented individuals who will sustain a culture of knowledge creation and innovation. A matrix mentorship program may provide many individuals with the guidance and support needed for career development while instilling institutional commitment. Such programs require an institutional commitment, both ﬁnancial and cultural. Gains in adaptability of the staff to shifting requirements, support for institutional values, and continuous gains

Matrix Mentorship in Academic Medicine

167

in efﬁciency, knowledge creation, and the quality of patient care should enhance the institution’s competitive position and compensate for the required initial investments.

ACKNOWLEDGMENTS The contributions of Drs. Linda Hill, Srikant Datar, and Benjamin Esty and the faculty of the Harvard Business School Executive Education Program in General Management (GMP3) to the development of this manuscript are gratefully acknowledged.

REFERENCES Allen, T. D., Eby, L. T., & Lentz, E. (2006). Mentorship behaviors and mentorship quality associated with formal mentoring programs: Closing the gap between research and practice. Journal of Applied Psychology, 91(3), 567–578. Allen, T. D., Eby, L. T., Poteet, M. L., Lentz, E., & Lima, L. (2004). Career beneﬁts associated with mentoring for protegee: A meta-analysis. Journal of Applied Psychology, 89(1), 127–136. Argyris, C. (1991). Teaching smart people how to learn. Harvard Business Review, 69(3), 99. Argyris, C. (1999). On organizational learning (2nd ed.). Oxford: Blackwell. Barondess, J. A. (1997). Mentoring in biomedicine. Journal of Laboratory & Clinical Medicine, 129(5), 487–491. Bhagia, J., & Tinsley, J. A. (2000). The mentoring partnership. Mayo Clinic Proceedings, 75(5), 535–537. Blackburn, R. T., & Fox, T. G. (1976). The socialization of a medical school faculty. Journal of Medical Education, 51(10), 806–817. Bland, C. J., & Schmitz, C. C. (1986). Characteristics of the successful researcher and implications for faculty development. Journal of Medical Education, 61(1), 22–31. Blixen, C. E., Papp, K. K., Hull, A. L., Rudick, R. A., & Bramstedt, K. A. (2007). Developing a mentorship program for clinical researchers. Journal of Continuing Education in the Health Professions, 27(2), 86–93. Chao, G. T., Walz, P. M., & Gardner, P. D. (1992). Formal and informal mentorships: A comparison on mentoring functions and contrast with nonmentored counterparts. Personnel Psychology, 45(3), 619. Chesney, R. W., Dungy, C. I., Gillman, M. W., Rivara, F. P., Schonfeld, D. J., Takayama, J. I., Alexander, D. F., Cairo, M. S., Dreyer, B. P., van Dyck, P., Ferrieri, P., Kohrt, A. E., McAnarney, E. R., Margolis, L. H., Orr, D. P., Rothstein, E., Simpson, L., Weitzman, M., Schonfeld, D. J., Yudkowsky, B. K., & Committee on Pediatric, Research. (2001). Promoting education, mentorship, and support for pediatric research. Pediatrics, 107(6), 1447–1450. Cohen, A. R., & Bradford, D. L. (1989). Inﬂuence without authority: The use of alliances, reciprocity, and exchange to accomplish work. Organizational Dynamics, 17(3), 4.

168

JAY A. FISHMAN

Cohen, A. R., & Bradford, D. L. (2005). Inﬂuence without authority (2nd ed.). Hoboken, NJ: Wiley. Conger, J. A. (2004). Developing leadership capability: What’s inside the black box? The Academy of Management Executive, 18(3), 136. Conger, J. A., & Benjamin, B. (1999). Building leaders: How successful companies develop the next generation. San Francisco, CA: Jossey-Bass. Connor, M. P., Bynoe, A. G., Redfern, N., Pokora, J., & Clarke, J. (2000). Developing senior doctors as mentors: A form of continuing professional development. Report of an initiative to develop a network of senior doctors as mentors: 1994–99. Medical Education, 34(9), 747–753. Curtis, J. A., Adam, H., & Shelov, S. P. (1995). A formal mentoring program in a pediatric residency. Academic Medicine, 70(5), 453–454. Delong, T., Gabarro, J., & Lees, R. (2007). When professional have to lead – A new model for high performance. Boston: Harvard Business School Press. DeLong, T. J., Gabarro, J. J., & Lees, R. J. (2008). Why mentoring matters in a hypercompetitive world. Harvard Business Review, January, 115–121. DeWitt, D. E., Curtis, J. R., & Burke, W. (1998). What inﬂuences career choices among graduates of a primary care training program? Journal of General Internal Medicine, 13(4), 257–261. Djerassi, C. (1999). Who will mentor the mentors? Nature, 397(6717), 291. Epstein, R. M. (1999). Mindful practice. JAMA, 282(9), 833–839. Frohlich, E. D. (2000). A renewed call to mentor. Hypertension, 36(3), 309–311. Gilster, S. D., & Accorinti, K. L. (1999). Mentoring program yields staff satisfaction. Mentoring through the exchange of information across all organizational levels can help administrators retain valuable staff. Provider, 25(10), 99–100. Higgins, M. C., & Kram, K. E. (2001). Reconceptualizing mentoring at work: A developmental network perspective. Academy of Management. The Academy of Management Review, 26(2), 264. Hill, L. (2003). Becoming a manager: How managers master the challenges of leadership (2nd ed.). Boston, MA: Harvard Business School Press. Hollingsworth, J. H. (2002). The difference between a mentor and a teacher. American Journal of Cardiology, 89(8), 1004–1005. Kotter, J. P. (1985). Power and inﬂuence. New York: Free Press. Kotter, J. P. (1995). Leading change: Why transformation efforts fail. Harvard Business Review, 73(2), 59. Kotter, J. P. (1996). Leading change. Boston, MA: Harvard Business School Press. Kram, K. E. (1985). Improving the mentoring process. Training and Development Journal, 39(4), 40. Kram, K. E. (1988). Mentoring at work: Developmental relationships in organizational life. New York: University Press of America. Kram, K. E., & Isabella, L. A. (1985). Mentoring alternatives: The role of peer relationships in career development. Academy of Management Journal, 28(1), 110. Lorsch, J. W. (1986). Managing culture: The invisible barrier to strategic change. California Management Review, 28(2), 95. Lorsch, J. W., & Tierney, T. J. (2002). Aligning the stars: How to succeed when professionals drive results. Boston, MA: Harvard Business School Press.

Matrix Mentorship in Academic Medicine

169

Marx, E. (2006). Mentor programs help create leaders of tomorrow. Health Care Strategic Management, 24(12), 10–11. Maudsley, R. F. (2001). Role models and the learning environment: Essential elements in effective medical education. Academic Medicine, 76(5), 432–434. McCall, M. W. (1997). High ﬂyers: Developing the next generation of leaders. Boston, MA: Harvard Business School Press. McCauley, R. (2007). Building a successful mentoring program. The Journal for Quality and Participation, 30(2), 17. Morzinski, J. A. (2005). Mentors, colleagues, and successful health science faculty: Lessons from the ﬁeld. Journal of Veterinary Medical Education, 32(1), 5–11. Morzinski, J. A., Diehr, S., Bower, D. J., & Simpson, D. E. (1996). A descriptive, crosssectional study of formal mentoring for faculty. Family Medicine, 28(6), 434–438. Morzinski, J. A., Simpson, D. E., Bower, D. J., & Diehr, S. (1994). Faculty development through formal mentoring. Academic Medicine, 69(4), 267–269. Motta, M. M. (2002). Mentoring the mentors: The Yoda factor in promoting scientiﬁc integrity. American Journal of Bioethics, 2(4), W17. O’Reilley, C. A., & Pfeffer, J. (2000). Hidden value: How great companies achieve extraordinary results with ordinary people. Cambridge, MA: Harvard Business School Press. O’Reilly, C. A., & Pfeffer, J. (2000). Hidden value: How great companies achieve extraordinary results with ordinary people. Boston, MA: Harvard Business School Press. Ognibene, F. P. (2007). Mentoring and lifelong learning: A critical part of who we are and what we do. Critical Care Medicine, 35(5), 1227–1229. Palepu, A., Friedman, R. H., Barnett, R. C., Carr, P. L., Ash, A. S., Szalacha, L., & Moskowitz, M. A. (1998). Junior faculty members’ mentoring relationships and their professional development in U.S. medical schools. Academic Medicine, 73(3), 318–323. Pfeffer, J. (1994). Competitive advantage through people: Unleashing the power of the work force. Boston, MA: Harvard Business School. Pfeffer, J. (2005). Producing sustainable competitive advantage through the effective management of people. The Academy of Management Executive, 19(4), 95. Porter, M. E. (1996). On competition. Boston, MA: Harvard Business School Publishing. Ragins, B. R., & Scandura, T. A. (1999). Burden or blessing? Expected costs and beneﬁts of being a mentor. Journal of Organizational Behavior, 20(4), 493. Ramanan, R. A., Phillips, R. S., Davis, R. B., Silen, W., & Reede, J. Y. (2002). Mentoring in medicine: Keys to satisfaction. American Journal of Medicine, 112(4), 336–341. Ramanan, R. A., Taylor, W. C., Davis, R. B., & Phillips, R. S. (2006). Mentoring matters. Mentoring and career preparation in internal medicine residency training. Journal of General Internal Medicine, 21(4), 340–345. Ready, D. A., & Conger, J. A. (2003). Why leadership-development efforts fail. MIT Sloan Management Review, 44(3), 83. Riechelmann, R. P., Townsley, C. A., Pond, G. R., & Siu, L. L. (2007). The inﬂuence of mentorship on research productivity in oncology. American Journal of Clinical Oncology, 30(5), 549–555. Rock, J. A. (1999). Mentoring in gynecology: Presidential address. American Journal of Obstetrics & Gynecology, 181(6), 1293–1295. Rogers, J. C., Holloway, R. L., & Miller, S. M. (1990). Academic mentoring and family medicine’s research productivity. Family Medicine, 22(3), 186–190.

170

JAY A. FISHMAN

Ross, A. (1984). The mentor’s role in developing new leaders. Hospital & Health Services Administration, 29(5), 21–29. Schindler, N., Winchester, D. P., & Sherman, H. (2002). Recognizing clinical faculty’s contributions in education. Academic Medicine, 77(9), 940–941. Selwa, L. M. (2003). Lessons in mentoring. Experimental Neurology, 184(Supplement 1), S42–S47. Souba, W. W. (2000). The essence of mentoring in academic surgery. Journal of Surgical Oncology, 75(2), 75–79. Whiting, V. R., & de Janasz, S. C. (2004). Mentoring in the 21st century: Using the internet to build skills and networks. Journal of Management Education, 28(3), 275. Whitworth, M. (2007). Mentorship in academic medicine. Medical Education, 41(9), 919. Wilson, J. A., & Elman, N. S. (1990). Organizational beneﬁts of mentoring. The Executive, 4(4), 88. Wright, S. (1996). Examining what residents look for in their role models. Academic Medicine, 71(3), 290–292.

THE IMPACT OF HOSPITAL OWNERSHIP CONVERSIONS: REVIEW OF THE LITERATURE AND RESULTS FROM A COMPARATIVE FIELD STUDY Lawton R. Burns, Rajiv J. Shah, Frank A. Sloan and Adam C. Powell ABSTRACT Change in ownership among U.S. community hospitals has been frequent and, not surprisingly, remains an important issue for both researchers and public policy makers. In the past, investor-owned hospitals were long suspected of pursuing ﬁnancial over other goals, culminating in several reviews that found few differences between for-proﬁt and nonproﬁt forms (Gray, 1986; Sloan, 2000; Sloan, Picone, Taylor, & Chou, 2001). Nevertheless, continuing to the present day, several states prohibit investorownership of community hospitals. Conversions to investor-ownership are only one of six types of ownership change, however, with relatively less attention paid to the other types (e.g., for-proﬁt to nonproﬁt, public to nonproﬁt). This study has two parts. We ﬁrst review the literature on the various types of ownership conversion among community hospitals. This Biennial Review of Health Care Management: Meso Perspectives Advances in Health Care Management, Volume 8, 171–229 Copyright r 2009 by Emerald Group Publishing Limited All rights of reproduction in any form reserved ISSN: 1474-8231/doi:10.1108/S1474-8231(2009)0000008011

171

172

LAWTON R. BURNS ET AL.

review includes the rate at which conversions occur over time, the relative frequency in conversions between speciﬁc ownership categories and the observed effects of conversion on hospital operations (e.g., strategic direction and decision-making processes) and performance (e.g., access, quality, and cost). Overall, we ﬁnd that the impact of ownership conversion on the different measures is mixed, with slightly greater evidence for positive effects on hospital efﬁciency. As one explanation for these ﬁndings, we suggest that the impact of ownership conversion on hospital performance may be mediated by changes in the hospital’s strategic content and process. Such a hypothesis has not been proposed or examined in the literature. To address this gap, we next study the role of strategic reorientation following hospital conversion in a ﬁeld study. We conceptualize ownership conversion within a strategic adaptation framework, and then analyze the changes in strategy content and process across sixteen hospitals that have undergone ownership conversions from nonproﬁt to forproﬁt, public to for-proﬁt, public to nonproﬁt, and for-proﬁt to nonproﬁt. The ﬁeld study ﬁndings delineate the strategic paths and processes implemented by new owners post-conversion. We ﬁnd remarkable similarity in the content of strategies undertaken but differences in the process of strategic decision making associated with different types of ownership changes. We also ﬁnd three main performance effects: hospitals change ownership for ﬁnancial reasons, experience increases in revenues and capital investment post-conversion, and pursue labor force reductions postconversion. Membership in a multi-hospital system, however, may be a major determinant of both strategy content and decision-making process that is confounded with ownership change. That is, ownership conversion may mask the impact of system membership on a hospital’s strategic actions. These ﬁndings may explain the pattern of performance effects observed in the literature on ownership conversions.

INTRODUCTION AND OVERVIEW Researchers have long studied the nonproﬁt organizational form among U.S. hospitals, asking why this form dominates the industry and whether it differs in behavior and performance from for-proﬁts. Typically these questions are studied cross-sectionally. Less frequently, they are analyzed dynamically by studying conversions in hospital ownership. This is surprising, given the relative frequency of these changes in ownership form. Moreover, there is little understanding of why so many different types of

The Impact of Hospital Ownership Conversions

173

conversion occur (e.g., nonproﬁt to for-proﬁt, public to for-proﬁt, nonproﬁt to public, for-proﬁt to public, for-proﬁt to nonproﬁt, public to nonproﬁt) and little theoretical work to explain it all (Sloan, 2000). This chapter seeks to improve our understanding of hospital conversions in several ways. First, we review the literature on ownership conversions among U.S. community hospitals to determine the impacts on hospital operations (e.g., competitive behavior, stafﬁng, decision making) and three measures of performance: access to care (e.g., provision of uncompensated care or unproﬁtable services), quality of care (mortality rates), and ﬁnances (efﬁciency, costs, revenues). This review concludes that conversions lead to changes in competitive strategy, headcounts, and locus of decision authority and exert mixed effects on performance, with perhaps some positive impact on hospital ﬁnances. Second, we develop a strategic adaptation framework in which performance effects are mediated by changes in strategic content and decision-making processes, which may help to explain the disparate ﬁndings. Such a perspective has not been utilized in prior hospital conversion research. Third, we summarize results from a ﬁeld investigation of such changes in sixteen hospitals undergoing four of the six different types of ownership conversion identiﬁed earlier. The ﬁeld study investigates whether there are similarities or differences in strategy across different types of conversions. Our research aim here is twofold: to discern any patterned variation in strategic reorientation associated with speciﬁc types of ownership changes and to assess whether such variation might impact subsequent performance. From a theoretical perspective, this chapter seeks to extend the strategic adaptation framework by explicating the ‘‘ﬁt’’ between a ﬁrm and its environment. Ownership conversion represents an adaptation to environmental pressures by changing the ﬁrm’s mission and orientation. However, there are multiple types of ownership change, each of which presents opportunities for executives to initiate changes in the ﬁrm’s strategic behaviors and decisionmaking processes, which may be differentially related to subsequent performance. Thus, we conceptualize the process of ﬁt as a series of changes in the ﬁrm’s strategic content and decision making, which may determine whether the adaptation is successful in improving organizational outcomes. The next two sections outline the importance of ownership conversions for both researchers and policy makers, and the strategic adaptation framework. The following section highlights the different rationales for conversion and how particular types of conversion may reﬂect conscious, strategic choices by organizational decision makers (e.g., administrators, physicians, trustees). We next consider the statistical evidence on conversion rates across types and over time and review the evidence on the determinants

174

LAWTON R. BURNS ET AL.

and performance results of ownership conversions. The latter portion of the chapter describes a comparative ﬁeld study of sixteen hospital conversions and the observed impact of conversion on hospital strategic content and decision-making process. We conclude with the implications of our ﬁndings.

SALIENCE AND SIGNIFICANCE OF OWNERSHIP CONVERSIONS The relationship between ownership and ﬁrm performance remains a particularly important policy issue in the context of U.S. hospitals. Whether accurate or not, there is a widespread perception that for-proﬁt hospitals engage primarily in proﬁt maximization, whereas nonproﬁt hospitals pursue more diverse missions including quality of care and community service (Newhouse, 1970; Weisbrod, 1989). Presumably, a nonproﬁt institution is more willing to trade net revenues and proﬁts for higher quality or improved access for underserved communities (Sloan & Steinwald, 1980; Pauly, 1987). Furthermore, public and nonproﬁt hospitals are the recipients of community investments, in the form of either direct taxpayer support or tax-exemption. As a result, when these institutions convert to for-proﬁt status, government policy makers and community representatives often fear that a community is losing a valuable resource and that sellers receive too little for their assets. The importance of this issue waxes and wanes with the frequency of ownership conversions. The ﬁrst surge in nonproﬁt to for-proﬁt conversions happened in the 1970s and early 1980s when for-proﬁt companies aggressively purchased nonproﬁt facilities. Numerous evaluations of the differences between nonproﬁts and for-proﬁts culminated in an inﬂuential report published by the Institute of Medicine (Gray, 1986). The conversion trend slowed with changes in Medicare reimbursement in the mid-1980s and major divestitures (e.g., HCA and HealthTrust), but picked up again in the mid1990s (Gray, 1991a, 1991b; Hadley, Gray, & Collins, 2001). In response to this later trend, at least nineteen states passed legislation governing nonproﬁt to for-proﬁt conversions. Moreover, the federal government aggressively pursued fraud investigations against three large for-proﬁt hospital chains (National Medical Enterprises, Columbia/HCA, Tenet) and levied hefty ﬁnes on each. These trends combined to increase the salience of hospital conversions for policy makers, and the need for systematic evaluations by researchers. Ownership conversions can be studied as both independent and dependent variables. On one hand, researchers may wish to study what

The Impact of Hospital Ownership Conversions

175

precipitates a conversion in ownership and distinguish the importance of many possible explanations, such as deteriorating ﬁnancial performance, the need to access capital, the desire to improve market share, or the desire to escape regulatory constraints. On the other hand, researchers may wish to determine the impact of conversion (one type of strategic change) on other types of strategic changes (e.g., change in competitive behavior or operations) and hospital performance measures (access, quality, cost). Hospital ownership changes constitute potentially meaningful events given the many known (and the many presumed) differences between for-proﬁt, nonproﬁt, and public forms. Among other things, ownership deﬁnes the distribution of accounting proﬁt (e.g., to equity holders vs. the community vs. government bodies), the clarity of ownership and owner goals, the election or appointment of the governing board, and the means by which capital can be raised (e.g., equity, taxable and tax-exempt debt, philanthropy, government subsidies) (Arrow, 1963). Such ownership changes also generate tremendous controversy regarding the costs, quality, access, and health outcomes of the community institutions involved. Hospitals can convert through multiple mechanisms. They may convert themselves without engaging in any transactions with an external organization, or they may be acquired by or merge with an external organization. Hospitals can also convert for multiple reasons. Although hospitals typically convert out of ﬁnancial desperation, they may also undergo conversions from a position of strength. Each of these situations results in a unique set of beneﬁts and concerns. Different types of concerns normally manifest in different parties involved in the sale of a hospital (Townsend, 1983). Consumers are often aware of the changing status of their hospitals, in part due to the publicity campaigns mounted by the purchasing companies, and are concerned about issues related to indigent care, local input in hospital policies, and the ﬁnancial aspects of the sale. Board members must agree before detailed negotiations can proceed and are often concerned about issues such as quality of care and the new advisory board that will provide input toward the governance of the converted hospital. Administrators, while heavily involved in negotiations, have expressed concerns about both staff job security and beneﬁt packages, as well as about quality of care. Physicians appear to be most concerned about the quality of the plant and equipment that they will have access to under corporate ownership. They also are concerned about whether their voices will be heard by the corporation. Finally, politicians express a similar set of concerns to those of consumers, including how to ﬁnance indigent care after the sale and how money can be saved to potentially buy back the hospital.

176

LAWTON R. BURNS ET AL.

One example of the process that a hospital must undergo when it considers conversion is the case of Good Samaritan Health System, which sold its four-hospital integrated healthcare system to Columbia/HCA in 1996 (Coye, 1997). Managed care had driven hospital occupancy below 55% in the county that the hospitals inhabited – a market in which Good Samaritan had the largest share (22% of discharges). In 1994, Good Samaritan performed a study that concluded that half the hospital beds in the county would be unneeded, causing increased competition among the area hospitals. The hospital’s board decided to pursue a strategic alliance with a larger player for three explicit reasons and one implicit reason. As California health plans consolidated, more leverage was needed. Furthermore, outside capital was needed to invest in physical infrastructure, IT, and service consolidation. Finally, a merger might yield economies of scale. Although not said openly, another likely motivation was that the nonproﬁt structure of the ﬁrm was regarded as impeding change, as the board had been unable to reach a consensus in addressing several issues in the past year and was hampered by the various constituencies that the board members served, and the public’s role in the decision-making process. Six potential acquirers (three for-proﬁts, three nonproﬁts) expressed interest in merging with Good Samaritan. Columbia/HCA succeeded in winning the acquisition because it offered the best mixture of cash, community beneﬁt, and reputation for quality care. There was little public opposition to the sale, and the proceeds were used to create a health-related foundation. In making the decision, the board considered ten issues: the tax implications, the provision of charity care, the support of academic residency programs, the quality of care, the degree to which hospital services in the community would continue to be provided, the acquirer’s experience in improving hospital information systems, changes in physician relations and the hospital’s organizational structure, the implications for the cost of care, the acquirer’s promise to endow charitable programs in the community, and the purchase price (Coye, 1997). When Academic Health Centers (AHCs) undergo conversions, an even broader set of concerns may be raised. Blumenthal & Weissman (2000) studied three AHCs that converted into for-proﬁt hospitals. They examined both the motivation for the sales and the effect of the sales on the operations and mission of the hospitals. In all three cases, the sale was prefaced by either actual ﬁnancial hardships or the fear of impending hardships. In two of the three cases, university presidents and administrators, rather than AHC leaders, initiated the sale to protect the university from the ﬁnancial hazards posed by the AHC. For-proﬁt chains were chosen as negotiations with

The Impact of Hospital Ownership Conversions

177

nonproﬁts had been unsuccessful. Nonproﬁts tended to be local competitors who intended to close the AHCs rather than upgrade them. For-proﬁts offered AHCs greater academic control after acquisition. For-proﬁt acquirers considered support of the social mission of AHCs to be part of the cost of owning them and abided by their commitments to continue offering uncompensated care. The acquirers may have been particularly well-behaved because for each of them, the examined AHC acquisition was their ﬁrst, and they may have wished to remain favorable suitors for subsequent AHCs.

STRATEGIC ADAPTATION FRAMEWORK One line of inquiry in strategic management research examines the ﬁrm’s performance as a function of the alignment or ﬁt between its external environment and its internal resources, capabilities, and organizational routines (Schendel & Hofer, 1979; Chakravarthy, 1982; McKee, Varadarajan, & Pride, 1989; Venkatraman & Prescott, 1990; Schindehutt & Morris, 2001; Jennings, 2004). External changes may undermine this alignment and the ﬁrm’s performance, requiring the ﬁrm to reorient its strategy and reconﬁgure its operations to achieve a better ﬁt with its environment. Such adaptations have been examined in the context of ﬁrms coping with industry deregulation (Kumaraswamy, Mudambi, Saranga, & Tripathy, 2008), economic liberalization reforms (Ray, 2003), reductions in federal reimbursement (Trinh & Begun, 1999), and growing complexity of client needs (Zinn, Mor, Feng, & Intrator, 2007). Strategic adaptation is the process by which an organization adopts a new set of strategies to better position itself in a changing market to ensure its viability and performance (Chakravarthy, 1982; Kimberly & Zajac, 1985; Shortell, Morrison, & Friedman, 1990). Changes in ownership constitute an attempt to adapt to external pressures that cause internal problems and misalignment. The adaptation may involve changes in corporate strategy and strategic decision-making processes that help the ﬁrm to survive and succeed. According to this perspective, strategy is the conduit through which changes in ownership affect ﬁrm performance. Ownership changes increase the ability of ﬁrm managers to initiate and implement strategic changes to meet the demands posed by environmental jolts and new internal imperatives. Strategy, deﬁned as ‘‘the plans and activities developed by an organization in pursuit of its goals and objectives, particularly in regard to positioning itself to meet external demands relative to its competition,’’ encompasses both strategic content and strategic processes (Shortell et al., 1990, p. 29). The content of strategy refers to the actual

178

LAWTON R. BURNS ET AL.

strategic direction and strategic behaviors pursued by a ﬁrm – including both intended and unintended-but-realized strategies (Mintzberg, 1978). Strategy process refers to the development and implementation process, which determines strategic behavior. Thus, from a strategic adaptation perspective, ownership changes represent a change in orientation and mission, which then entail new strategies and/or new types of strategic decision making. Strategic adaptation theory also implies that ownership changes occur largely to allow an organization to pursue a new strategic direction, better execute an existing set of strategies, or revise how the organization makes strategic decisions. Strategic adaptation theory has two implications for the time frame of changes in ownership, strategic behavior, and ﬁrm performance. First, the types of strategies pursued may themselves vary with time. For example, ﬁrms may privatize to pursue certain short-term strategies, such as restructuring or downsizing to gain operational efﬁciencies, or to pursue new long-term strategies, such as developing new capabilities or product lines that serve as the basis for future growth and competitive advantage. Second, the strategic adaptation perspective suggests that research time frames – which are usually short term – may be inadequate for observing changes in organizational performance following changes in governance. Instead, researchers may beneﬁt from studying the short-term changes in strategy and strategic decision making that ultimately lead to long-term changes in performance.

RATIONALES FOR CONVERSION AND STRATEGIC CHOICES AMONG TYPES Conversion Rationales Despite all the issues that are brought forth by conversion, and despite the potential of conversion to affect many aspects of their operation, a substantial number of hospitals have chosen to convert. Motivations for conversion have included increased access to capital, increased efﬁciency, necessity for survival, lessened regulatory constraints, and the favorability of conversion for directors and managers. Hospitals have also converted as a means of maintaining their independence, preserving their culture, implementing strategic change (in cases of conversion away from nonproﬁt status, through the formation of a foundation), deal with ﬁnancial or strategic weakness, and respond to competition. Two types of nonproﬁt to for-proﬁt conversions occur: the leveraged buyout of a hospital by internal members or the purchase of a hospital’s

The Impact of Hospital Ownership Conversions

179

assets by an external organization. Although nonproﬁt to for-proﬁt conversions may be done out of a position of strength, these conversions appear to primarily have been done by hospitals with few other options (Gray, 1997). The primary motivation cited for conversion to for-proﬁt status is to gain access to capital, as for-proﬁts can sell equity, which may be cheaper than taking on debt (Claxton, Feder, Schactman, & Altman, 1997; Cutler & Horwitz, 2000). Other reasons that have been suggested are to improve efﬁciency, to ensure the survival of the hospital or the continuity of its mission, to reduce the extent to which the hospital is subject to regulatory constraints (nonproﬁt hospitals are less able to enter into proﬁt-sharing arrangements with their staffs than for-proﬁt hospitals), to ﬁnancially beneﬁt the directors and managers of the hospital (whom may have a ﬁnancial relationship with the buyer), to obtain relief from unmanageable debt, to increase efﬁciency through superior management, to maintain independence, to preserve culture, and to implement strategic change through the formation of a foundation (Cutler & Horwitz, 2000; Cain Brothers, 2002). Conversions may also be driven by the culture of the board members; businessmen may be more amenable to for-proﬁt ownership than people from religious or nonproﬁt careers (Cutler & Horwitz, 2000). According to Gray (1997), there are six advantages and three disadvantages related to the conversion of nonproﬁts to for-proﬁts. Nonproﬁt conversions are beneﬁcial in that they facilitate the adoption of government-based coverage of the uninsured, move organizations into the tax base, put charitable assets to more productive use (the proceeds from the sale of a hospital may be used to establish a foundation), provide hospitals with better access to capital, ease consolidation and capacity reductions, and help eliminate the ﬁctitious belief that nonproﬁts are more socially beneﬁcial than for-proﬁts. However, there are also several disadvantages to conversion. Conversions in which a sale is made to an internal party may result in windfalls for the buyer, if the buyer is able to coordinate the sale at an artiﬁcially low price. After a conversion, it is often unclear who is responsible for protecting the public’s interest, as neither the sellers or purchasers are responsible for considering the impact of a conversion on the community. Finally, there is a potential loss of social beneﬁts arising from the nonproﬁt delivery of care, such as the elimination of a lever that the government has for ensuring the delivery of charity care (government bodies may require that a nonproﬁt hospital deliver a certain amount of charity care in order to receive nonproﬁt status), the loss of trustworthiness perceived to be possessed by nonproﬁt hospitals due to their reduced conﬂicts of interest, and reduction in community beneﬁts delivered by the hospital (indigent care, unproﬁtable programs).

180

LAWTON R. BURNS ET AL.

Changes in ownership status may help hospitals be more nimble in how they can spend their assets. Nonproﬁt hospitals may face legal difﬁculties in repositioning their ﬁnancial assets, as charitable trust laws and the cy pres doctrine (the legal doctrine through which the original intent of a charitable organization is interpreted when it cannot be literally fulﬁlled due to its illegality, impracticality, or impossibility) constrain them. In contrast, forproﬁt companies may be more agile in moving their assets across geographic markets and industry segments, allowing them to be more responsive to changing needs and innovation (Cain Brothers, 2002). In contrast, for-proﬁt to nonproﬁt conversions may result from hospitals not delivering their owners the desired rate of ﬁnancial return. Public hospitals convert to nonproﬁt ownership either because they wish to expand beyond their political jurisdictions, or they wish to increase their ability to compete for managed care contracts, or because the governance and management restrictions resulting from public ownership have made it difﬁcult to compete with other hospitals (Legnini, Rosenberg, Perry, & Robertson, 2000). Finally, governments have been unwilling to subsidize hospitals that exclusively provide care to the indigent; as a result, hospitals had to adapt so that they can also meet the needs of paying clients (Legnini et al., 1999). Thus, it appears that the motivation for conversion is highly dependent on the type of conversion a hospital is experiencing. Conversion to for-proﬁt status, the most heavily researched variety of conversion, appears to be motivated primarily by the desire to increase ﬂexibility in ﬁnancing. Conversion from for-proﬁt to nonproﬁt status also appears to be motivated by negative ﬁnancial performance; however, conversion is used as a means of unloading the hospital rather than increasing ﬂexibility in its ﬁnancing.

Strategic Choice among Conversion Types It is important to distinguish between hospitals that actively seek an ownership change from those that are driven to an ownership change by virtue of consolidation (Claxton et al., 1997). When considering a conversion, hospital executives and physicians must contemplate whether it is best for a hospital to maintain its current ownership structure, and if not, which type of conversion should be pursued. Given that physicians often have a degree of job mobility, hospital executives must consider the perspectives of physicians when making conversion decisions to prevent physician ﬂight.

The Impact of Hospital Ownership Conversions

181

Converting a hospital requires a complete reworking of the dynamics of power within the hospital. When a nonproﬁt or public hospital is bought through acquisition, a foundation must be created. Creating a foundation requires the development of a new corporate structure and mission. An alternative to acquisition is to form a joint venture with a for-proﬁt operator. In joint ventures, the former hospital operators sell a ‘‘share’’ of the hospital to a new joint organization in exchange for cash equal to a fraction of the hospital’s value. By only selling a fraction of the hospital, the nonproﬁt entity can maintain a degree of control.1 For-proﬁts tend to be willing to allow the nonproﬁt’s board to have control over hiring and ﬁring the CEO, taking on debt, changing the hospital’s service offerings, modifying the mission statement, changing the charity care policy, approving the hospital’s budget and strategic plan, and selling the assets. For-proﬁts grant local boards these powers, as they feel that having a hospital be responsive to the needs of the community is vital to its success (Hollis, 1997). Although conversions largely occur without community involvement (GAO, 1997), the boards of directors of nonproﬁt hospitals have a ﬁduciary duty to consider community needs when making decisions on behalf of the hospital. Nonproﬁts and for-proﬁts respond to the competitive need to expand in different ways. Nonproﬁts tend to try to merge with their neighbors to reduce the ability of managed care to play one neighbor off of another. Unfortunately, this is often easier said than done, as there can be substantial resistance to rationalizing clinical services, reducing members of the management team, and choosing a common CEO. When a merger results in an entity with too much power, it may be opposed by the Federal Trade Commission. Another option hospitals have is to merge into a system, often containing a regional management team located at a somewhat distant centralized ofﬁce. Systems tend to be paternalistic and reduce local control. Neither merging with a neighbor nor a system represents an economic transaction. Instead, these options pass the stewardship of the nonproﬁt hospital from the board to a new, larger nonproﬁt entity. It is unlikely for another nonproﬁt to outright buy a hospital, as nonproﬁts lack the capital to make such a purchase, unless ﬁnanced with debt or internally generated cash (Hollis, 1997).

Other Factors Shaping the Conversion Process The relationship that a hospital has with its clinicians can affect the conversion process. Anderson, Allred, and Sloan (2001) sought to determine

182

LAWTON R. BURNS ET AL.

whether the level of decision-making involvement of physicians and nurses differed in hospitals that converted versus those that had not. The authors tested two competing theories of how clinicians respond to conversions. In the ﬁrst, the threat-rigidity perspective, clinicians view conversion as a threat and respond by limiting the ﬂow of information by providing less feedback and by reverting to following standard operating procedures. In the second, the complexity perspective, a discontinuous change (like an ownership conversion) causes a complex system to become more capable of change by bringing the system to the edge of chaos. The researchers found more empirical support for the complexity perspective than the threat-rigidity perspective. Organizations that have been successful in converting hospitals to forproﬁt status have had to demonstrate that they have a good track record in serving communities and are capable stewards of the missions of the hospitals that they are acquiring. They often have to persuade clinician stakeholders that the choice the hospital faces is not a choice between continuing as a public organization and privatizing, but instead between surviving as a private organization and failing as a public one (Legnini et al., 1999). They may also have to assure clinicians that their interests will be served. Research shows that physicians have strongly inﬂuenced the decision of boards to sell to for-proﬁt purchasers instead of to nonproﬁt purchasers that they might have preferred. Although some physicians may have biases against for-proﬁt purchasers, others prefer them, as they feel that under a for-proﬁt culture, they are more likely to be treated as valued customers of the hospital (Collins, Gray, & Hadley, 2001). The conversion process can also be shaped by cultural issues and interpersonal dynamics involving administration and the board. One inhibitor of conversions is a ‘‘culture of conservatism’’ that strongly resists change. This conservatism stems primarily from the executives, although it can originate with the board. Conservative nonproﬁt cultures can lead to ﬁnancial distress, which in turn leads to conversion (Health Care Advisory Board, 2000). From these studies, it appears that both the medical and administrative staff play a vital role in the hospital conversion process. For a conversion to be successful, it must be framed as being of substantial importance to the hospital’s continuation. There is empirical evidence that a major discontinuous change like conversion can make hospital employees and staff more willing to consider other forms of change until a new equilibrium is reached. In conversions to for-proﬁt status, the conversion can often be cast as necessary for the hospital’s survival.

183

The Impact of Hospital Ownership Conversions

HOSPITAL OWNERSHIP CONVERSIONS ACROSS TYPES AND OVER TIME In contrast to most industries, hospital researchers can conduct a diverse experiment in ownership conversions. The hospital market is characterized by (a) the presence of public, private nonproﬁt, and private for-proﬁt ﬁrms and (b) frequent and simultaneous ownership transitions between any pair of these ﬁrms. We thus consider changes in ownership between (not within) the three major classes of hospital ownership: for-proﬁt (investor-owned), nonproﬁt (religious or secular), and public (local/state government). Given three different forms, there are six possible types of ownership transitions that can be studied: public to nonproﬁt, public to for-proﬁt, for-proﬁt to nonproﬁt, nonproﬁt to for-proﬁt, nonproﬁt to public, and for-proﬁt to public. Historically, these types have occurred with differing frequency, affecting a substantial number of hospitals. Between 1970 and 1995, 7% of nonproﬁt hospitals became for-proﬁt (Cutler & Horwitz, 2000). However, of the 87 nonproﬁt hospitals that converted between 1985 and 1994, 44% changed for-proﬁt ownership status more than once, 7% converted back to nonproﬁt status, and 13% ceased operations (Collins et al., 2001). Thus, a hospital’s ownership status should be considered a dynamic, not a static state. According to data supplied by the American Hospital Association2, hospital conversions (regardless of type) occurred with rather consistent frequency between 1991 and 2007 (Table 1). Although a perfect comparison cannot be made due to differences in the lengths of the periods, a roughly comparable number of conversions occurred during seven years in the early 1990s (431) and the ﬁrst eight years of the 2000s (384). The frequency of conversions across types has changed, however. Between 1991 and 2000, the Table 1. The Number of Hospital Conversions, by Type, Over Time.

Nonproﬁt to for-proﬁt For-proﬁt to nonproﬁt Nonproﬁt to public Public to nonproﬁt For-proﬁt to public Public to for-proﬁt Total conversions

1991–1997a

1997–2000

2000–2004

2004–2007

Total

127 58 49 143 19 35 431

45 69 16 81 7 14 232

68 27 28 49 7 17 196

56 30 40 29 13 20 188

296 184 133 302 46 86 1047

Note: Compiled from AHA data by the authors. a From Thorpe et al. (2000), Fig. 1.

184

LAWTON R. BURNS ET AL.

most common type of conversion was from public to nonproﬁt status, and the second most common type of conversion was from nonproﬁt to forproﬁt status. By contrast, between 2000 and 2007 the most common type of conversion was from nonproﬁt to for-proﬁt status. Between 2000 and 2004, the second most common type of conversion was from public to nonproﬁt status, and between 2004 and 2007, the second most common type of conversion was from nonproﬁt to public status. The least common type of conversion during all the years examined was from for-proﬁt to public status, and the second least common was from public to for-proﬁt status. There has been a decline over time in the number of public to nonproﬁt conversions, as well as in the number of nonproﬁt to for-proﬁt conversions. Given that rather similar numbers of nonproﬁt to for-proﬁt (296) and public to nonproﬁt (302) conversions occurred during the period that we examined, it is surprising that the literature has so disproportionately focused on nonproﬁt to for-proﬁt conversions. Fig. 1 examines the change in the distribution of conversions between time periods. Since the time periods are not all of the same length, making absolute comparisons of the number of conversions of each type can be difﬁcult. By using relative comparisons, it is possible to generate conclusions without being hampered by the differences in period lengths. From the donut chart, it appears that there has been an increase in the relative frequency of conversions from nonproﬁt to public status over time, and a decrease in the relative frequency of conversions from public to nonproﬁt status. There has also been a slight increase in the relative frequency of hospitals converting from public to for-proﬁt status. There has been substantial regional variation in the propensity of hospitals to convert, perhaps partially driven by state regulations or prohibitions governing conversion (Table 2). Thorpe, Florence, and Seiber (2000) reported that between 1991 and 1997, the majority of conversions took place in the South, and the fewest took place in the Northeast. Among the 87 for-proﬁt conversions of nonproﬁt hospitals that took place between 1985 and 1994, over a third took place in only three states, and nearly half occurred in the Southeast (Collins et al., 2001). On the basis of more recent data from the American Hospital Association, Fig. 2 classiﬁes the number of hospital conversions that have occurred between 1997 and 2007 by the nine census regions. In all three time periods, the most conversions occurred in the South Atlantic states (DE, DC, FL, GA, MD, NC, SC, VA, and WV), and the second most occurred in the West South Central states (AR, LA, OK, and TX). The fewest number of conversions in all periods occurred in New England (CT, ME, MA, NH,

185

The Impact of Hospital Ownership Conversions

7%

11% 9% 3% 3% 6% 8% 4%

19% 30%

30% 35%

1991-1997 1997-2000

25% 15% 35% 33%

2000-2004 2004-2008 14%30% 11% 7% 14% 21%

14%

16%

Nonprofit to for-profit For-profit to nonprofit Nonprofit to public Public to nonprofit For-profit to public Public to for-profit

Fig. 1.

Change in the Distribution of Conversions over Time (Compiled from AHA data by the authors).

VT, and RI). There appears to be great consistency over time in the relative number of conversions that occur by region. This is likely due to regional variation in the regulation of conversions. Conversions are legally overseen by state attorneys general (AG); little oversight occurs at the Federal level. We also examined how the distribution of conversions across bedsize categories has changed over time. The results are presented in Fig. 3. In all periods, the vast majority of conversions happened in smaller hospitals with fewer than 200 beds. The fewest conversions consistently happened in hospitals with 500þ beds, and the second fewest happened in hospitals with 400–499 beds. The distribution of converted hospitals by bedsize conforms to the expectations that ownership transitions occur primarily among smaller and ﬁnancially distressed hospitals.

186

LAWTON R. BURNS ET AL.

Table 2.

Number and Geographic Distribution of Hospital Conversions, 1991–1997.

Type of Conversion

Region Northeast

Nonproﬁt to for-proﬁt Public to for-proﬁt For-proﬁt to nonproﬁt Public to nonproﬁt Nonproﬁt to public For-proﬁt to public

5 1 2 6 2 0

Total Percentage

South 66 26 32 86 27 15

16 3.71

252 58.47

Total Midwest

West

28 4 8 31 7 1

28 4 16 20 13 3

127 35 58 143 49 19

79 18.33

84 19.49

431 100.00

Source: From Thorpe et al. (2000), Fig. 1.

Fig. 2.

Hospital Conversions by Geographic Region.

Fig. 3.

Percentage

Hospital Conversions by Bedsize.

29.47 8.12 13.46 33.18 11.37 4.41

The Impact of Hospital Ownership Conversions

187

DETERMINANTS AND PERFORMANCE EFFECTS OF CONVERSIONS: REVIEW OF PREVIOUS RESEARCH EVIDENCE Determinants of Ownership Conversion One means of determining the factors that increase a hospital’s propensity to convert is to compare hospitals that chose to and chose not to convert. Sloan, Ostermann, and Conover (2003b) found that hospitals that converted ownership had consistently below-average proﬁt margins, lower occupancy rates, and a higher degree of leverage before conversion than hospitals that chose not to convert. The markets in which converted hospitals were situated also differed from those in which conversions did not occur. Markets in which conversion occurred tended to have lower per capita income, less HMO activity, and higher Herﬁndahl indices of area hospital bed supply. All these factors suggest that these markets were less competitive. Similar ﬁndings have been reported by other researchers. Collins et al. (2001) remarked that hospitals that converted from nonproﬁt to for-proﬁt status tended to be in poor ﬁnancial condition before they were sold. Mark et al. (1997) found that conversions were more likely among hospitals with lower proﬁtability, communities with lower per-capita income and higher unemployment, higher Medicare costs, smaller size, and less involvement in teaching. The percentage of a hospital’s total patient days covered by Medicaid did not appear to impact conversion. As a result, many of them were losing money before conversion (45% of conversions between 1988 and 1995) and were ﬁnancially distressed. From these studies, it appears that conversion is in general driven by below-average performance, in terms of proﬁt, occupancy, and leverage. Conversion also tended to be more prominent in markets with less competition. Unfortunately, all the studies examined in our review of the determinants of conversion either examined conversion in general or examined conversion from nonproﬁt to for-proﬁt status.

Effects of Conversion on Hospital Operations Mission, Competitive Behavior, and Repositioning of Assets Little is known about the strategies and decision-making processes employed across the three types of ownership models (let alone how

188

LAWTON R. BURNS ET AL.

changes in ownership inﬂuence changes in strategy). Early thinking tended to develop simpliﬁed objective functions such as the pursuit of access goals in public hospitals, quality goals in nonproﬁt hospitals, and revenue/cost goals in for-proﬁt hospitals (Yoder, 1986). During the period of corporate diversiﬁcation in the mid-1980s, researchers surmised that hospitals with broad missions and goals were better positioned to adapt to environmental shifts by virtue of their diversiﬁed activities (Shortell, Morrison, Hughes, Friedman, & Vitek, 1987). Empirical studies of hospital systems have sought to identify ‘‘strategic orientations’’ theorized to affect performance such as Miles and Snow’s (1978) fourfold typology: prospectors, analyzers, defenders, and reactors (cf. Shortell et al., 1990). There has been a small amount of research on the impact of conversions on the strategic goals of hospitals. Findings relate to how conversions affect hospital goals and missions, values, culture, competitive behavior, relationships with physicians, capital investments, ﬁnancial activities, and operations. The degree of change that must occur during a conversion away from nonproﬁt status may have been mitigated by the fact that nonproﬁt hospitals have had to respond to changes in the payment system that have required them to act more like businesses (Cain Brothers, 2002). Regardless of conversion, nonproﬁts have had to change their missions somewhat, as the rationale for their tax exemptions has shifted from being exclusively linked to ‘‘charity care’’ to being linked to ‘‘the promotion of health for the beneﬁt of the community’’ (Fishman, 1998). Conversion can enable nonproﬁt hospitals to undergo a strategic change in which the acute care operations are separated from the charitable operations. Hospitals have become increasingly complex to manage, and conversion enables the management of operations to be done by a company with expertise in that area, while the residual foundation can go on to oversee the allocation of the hospital’s ﬁnancial assets to serving community needs (Cain Brothers, 2002). However, foundations established after the sale of nonproﬁt hospitals do not always go toward supporting medical care in the community. Of 14 nonproﬁt conversions reviewed by the U.S. General Accounting Ofﬁce, 12 contributed the proceeds of their sales to charitable foundations. One of the foundations examined used the proceeds to support an aerospace program, construction of an art, education, and technology center, and other projects tangential to healthcare. In the two conversions that did not result in establishing a foundation, the proceeds went to a university and a city (GAO, 1997). Thus, it is clear that conversion can transform a charitable medical entity into a charity that has nothing to do with medicine whatsoever.

The Impact of Hospital Ownership Conversions

189

In addition to changing the missions of hospitals, conversions can change the obligations of hospitals. There are several legal and economic differences between for-proﬁt and nonproﬁt hospitals. For-proﬁts are owned by investors, have a legal obligation to promote the wealth of shareholders, may redistribute proﬁts to owners, make management accountable to owners (shareholders), receive capital from investors, have retained earnings and return on equity from payments, derive revenue from the sale of services, and pay income, property, and sales taxes. In contrast, nonproﬁt hospitals either have no owners or are owned by members, have a legal obligation to fulﬁll a mission statement, cannot distribute ﬁnancial surpluses to those who control the organization, make management accountable to voluntary boards, receive capital from contributions, debt, retained earnings, and government grants, derive revenue from the sale of services and from contributions, and are eligible for exemptions from most taxes. Finally, due to their support from state and local taxes, public hospitals have clear obligations as safety-net providers (Gray, 1991a, 1991b; Thorpe et al., 2000) Operations and Stafﬁng Hospital conversions can affect hospital operations through stafﬁng levels and mix. Mark (1999) observed that while nonproﬁt to for-proﬁt conversions were associated with a decline in the hospital’s staff-to-patient ratio, no such decline occurred in for-proﬁt to nonproﬁt conversions. However, nonproﬁt hospitals had higher ratios than the industry average. There was also an increase in the ratio of both registered nurses and administrators to patients in hospitals that converted to nonproﬁt status. In seven of the eight nonproﬁt to for-proﬁt conversions examined by Mark et al. (1997), stafﬁng was reduced after conversion. Although this often happened at the managerial level, it also sometimes occurred with the nursing staff. Furthermore, Shen (2003) found that converting to for-proﬁt status resulted in a greater reduction in stafﬁng than other forms of conversion. On the other hand, there was little change in bed capacity found after conversion to for-proﬁt status, but some reductions found after conversion to government and nonproﬁt status. To improve hospital operations while maintaining cooperation from the staff, for-proﬁt acquirers have had to tread carefully. In three cases studied by Blumenthal and Weissman (2000), AHCs chose for-proﬁt acquirers over nonproﬁt after it became clear that nonproﬁts viewed them as competitors in overbedded markets and saw no advantage in upgrading them. Nonproﬁts sought to close AHCs or demanded greater control than forproﬁt acquirers. Thus, for-proﬁt acquisition was seen as the best means of

190

LAWTON R. BURNS ET AL.

preserving the academic culture of the hospitals. Rather than cutting costs through large lay-offs, acquirers have had to opt for job redesign and labor force reduction through attrition (Legnini et al., 1999). Although labor has often had to agree to a relaxing of some of the beneﬁts and protections provided by the public sector, grandfathering has been used to ease the amount of change that existing employees must face. Cost savings have also arisen through the consolidation of departments and functions across institutions to eliminate redundancies. Operations have been affected by redistributing teaching responsibilities across sets of integrated hospitals. Although operational changes are likely a part of any hospital conversion strategy, they must be approached with care. Decision Making In addition to having ﬁnancial and operational impacts, conversions can affect who makes decisions, how power is balanced, how quickly decisions are reached, and where decisions are made within the greater organization. Using a sample of hospitals that had undergone all six types of conversions, Anderson et al. (2001) found that converted hospitals had greater levels of physician and nurse participation in decision making than nonconverted hospitals. Conversions often required a change in who makes decisions. On the other hand, hospital boards are reluctant to share control. They also have the incentive not to do so, as they are unlikely to be criticized if they procrastinate in taking action, but are quite open to criticism if they choose to participate in a merger or acquisition. After deciding to afﬁliate with another organization, a board must develop a set of criteria for selection. Typically, nonproﬁt boards consider issues such as the mission of the hospital, needs of the community, quality of care provided by the acquirer, access and contract implications, and desires of the physicians. Issues related to ﬁnancial returns and maximizing ﬁnancial value play less of a central role in their decision-making process. Once these criteria have been set, a board may issue a request for proposals (Hollis, 1997). Although some nonproﬁt boards have negotiated contractual commitments to provide charity care into their sales agreements, other boards retained inﬂuence over the operations through the creation of joint ventures (Mark et al., 1997). After a nonproﬁt or public hospital converts to for-proﬁt status, a noncorporate advisory board is often created to help oversee the hospital. Although stakeholders have often felt that the role of newly created hospital advisory boards is ambiguous, they have been supportive when they view the corporate owner as trustworthy. Although a high level of trust often exists, in

The Impact of Hospital Ownership Conversions

191

most cases the original owners of a hospital write an assurance of the continuation of the hospital’s mission into the sale contract (Townsend, 1983). Regardless of assurances, substantial change is inevitable. When a hospital conversion from nonproﬁt or public to for-proﬁt status occurs, power inevitably shifts from the hospital’s board to the for-proﬁt company. Through the use of hybrid organizational forms or advisory committees, it is possible for the community to continue to play a role in the regulation of a converted hospital. In the case of a speciﬁc hospital’s conversion from nonproﬁt to for-proﬁt status, one analyst stated, ‘‘The hospital board must realize that it gives up a certain amount of power in lieu of the corporate board. Long range planning and allocation of capital resources are now up to the corporate board. In addition, the hiring and ﬁring of the CEO is no long in control of the hospital board’’ (Health Care Advisory Board, 1991). Although changes in top level administration are typically considered to be part of the cost of doing business with a corporation, the vast majority of contracts provide some form of assurance of continued employment for hospital staffs, lessening concerns over turnover (Townsend, 1983). State AGs have played inconsistent roles in regulating hospital conversions. Wilkins and Jacobson (1998) argued in favor of the establishment of federal regulation of the conversion process to clarify the AGs’ role, deﬁne how the fair market value of a hospital should be determined, and create guidelines by which foundations should be created. The authors noted that in states in which conversions are not regulated, there has not been widespread establishment of foundations. However, not everyone has been in support of federal regulation, as increased regulation might reduce a nonproﬁt hospital’s maneuverability to save itself from problems related to debt and poor operations. Aside from ﬁnancial concerns, for-proﬁt conversion can also help increase the speed of decision making. By contrast, public institutions face many constraints such as open meeting requirements, limitations on downsizing, an inability to change the mix of employees, difﬁculty in hiring high-quality managers, limits on borrowing, barriers in joining joint ventures, and slower decision-making processes – all of which limit their ability to competitively respond (Legnini et al., 1999). Similarly, changes within nonproﬁts take place more slowly, as boards have political battles related to what the organization should be named and how its assets should be invested. If hospital boards delay conversion, they may lose the opportunity to capture the value of their hospitals before that value erodes (Pallarito, 1996). Thus, if a hospital feels that it is vital that it convert to for-proﬁt status, there are advantages to be gained from doing so rapidly.

192

LAWTON R. BURNS ET AL.

Effects of Conversion on Hospital Performance Efﬁciency/Finances There are good reasons to suspect that the ﬁnancial implications of different types of conversion might have diminished over time as the differences between for-proﬁt and nonproﬁt hospitals have eroded. Both types of hospitals now rely heavily on revenue from the sale of services, depend on their own economic performance to gain access to capital, and have often lost local control through the rise of multi-institutional systems and hybrid for-proﬁt/nonproﬁt organizations. By the 1980s, 95% of nonproﬁt hospital revenues came from patient care services, making them as accountable to the market as for-proﬁts (Gray, 1991a, 1991b). There are also good reasons to suspect the ﬁnancial implications of different types of conversion may have increased over time. Although nonproﬁt hospitals used to rely on charity to fund new construction, they now often rely on debt. However, nonproﬁts issue tax-exempt bonds that pay lower rates of interest than the non-exempt bonds issued by for-proﬁts. As a result, the bonds of for-proﬁt hospitals are more appealing to taxexempt nonproﬁt investors due to their higher returns, while the tax-exempt bonds of nonproﬁt hospitals are more appealing to tax-paying for-proﬁt investors. For-proﬁt and nonproﬁt hospitals differ in the accountability of hospital executives; executives are more centrally accountable in for-proﬁt hospitals and more locally accountable in nonproﬁt organizations. Research evidence from the early 1990s suggests a convergence in the post-conversion performance of for-proﬁts and nonproﬁts. Mark (1999) examined all private acute care hospital conversions that occurred between 1989 and 1992 and found that hospitals which converted either from nonproﬁt to for-proﬁt or from for-proﬁt to nonproﬁt had signiﬁcantly lower proﬁt margins before conversion than hospitals which chose not to convert. This ﬁnding was more strongly true for nonproﬁt hospitals that converted to for-proﬁt status. After conversion in either direction between for-proﬁt and nonproﬁt status, the proﬁtability of the hospitals examined substantially improved. Research evidence from the later 1990s suggests a divergence between for-proﬁts and nonproﬁts. Shen (2003) found that among hospital conversions between 1987 and 1998, conversions to government and forproﬁt ownership resulted in higher proﬁt margins, whereas conversions to nonproﬁt status resulted in an insigniﬁcantly lower proﬁt margin. More recent evidence suggests that different types of conversions have different ﬁnancial impacts, although the ﬁndings are not entirely consistent. Cutler & Horwitz (2000) found efﬁciencies resulting from conversion from

193

The Impact of Hospital Ownership Conversions

nonproﬁt to for-proﬁt form and that for-proﬁts could cut costs in places where nonproﬁts would be constrained from doing so. In contrast, Sloan, Conover, and Ostermann (2003a) found that while all of the conversions they examined resulted in an increase in operating margin, the increase was most dramatic for conversions of for-proﬁts to government-run hospitals and was the least dramatic for conversions of for-proﬁts to nonproﬁts (Table 3). Sloan et al. (2003a, 2003b) also found that conversion to for-proﬁt status lead to high internal rates of return. Even if the internal rate of return was calculated over a period of only ten years, it exceeded the weighted cost of capital, which was assumed to be 5.6%. Thus, the conversion of both government and nonproﬁt hospitals to for-proﬁt status appears to be a proﬁtable endeavor. This implies that hospitals have been underpriced before being sold (Table 4). Collins et al. (2001) found that for-proﬁt owners of formerly nonproﬁt hospitals often were unsuccessful in ending the ﬁnancial troubles of the hospitals that they had acquired. The beneﬁts resulting from superior managerial expertise and enhanced access to capital did not often seem to materialize. Meanwhile, Cain Brothers (2002) noted that although productivity efﬁciencies (economies of scale in management expertise, supply purchasing, and investment in technology) might be anticipated to Table 3. Operating Margins Resulting from Conversions. Before

After

Before Operating Margin

After Operating Margin

Change in Operating Margin

2.18% 2.04% 4.79% 5.86%

0.34% 0.31% 3.53% 0.15%

2.52% 2.35% 1.26% 5.71%

Nonproﬁt For-proﬁt Government For-proﬁt For-proﬁt Nonproﬁt For-proﬁt Government

Source: Sloan et al. (2003a), Table 3.

Table 4. Before

Government Nonproﬁt

Internal Rate of Return Experienced by Hospital Acquirers. After

For-proﬁt For-proﬁt

Source: Sloan et al. (2003a), Table 4.

Internal Rate of Return % Calculated Over: 30 Years

20 Years

10 Years

16.1% 12.7%

15.3% 11.8%

12.7% 6.3%

194

LAWTON R. BURNS ET AL.

result from the purchase of a small hospital by a for-proﬁt hospital management company, these same advantages would likewise also occur if the small hospital had merged or been sold to a larger nonproﬁt system. Thus, conversion is a complicated process that can have implications for both the organizational dynamics of a hospital and its ﬁnancial performance. Different types of conversions can have different effects on a hospital’s operating margin. In the event a hospital converts to for-proﬁt status, its leaders must be careful that it is fairly priced. Under-pricing of hospitals can lead to acquirers experiencing large rates of return and community foundations not receiving as sizable proceeds as they deserve. Prices There is some evidence that conversion to for-proﬁt status may result in an increase in the price of the hospital services. Picone, Chou, and Sloan (2002) found that when hospitals converted to for-proﬁt status from government or nonproﬁt status, payments per Medicare admission increased more than they did when hospitals converted from for-proﬁt status to government or nonproﬁt status. This price increase can either stem from for-proﬁts charging higher prices or being able to invest in the capability to deliver higher-cost services. In contrast, Young and Desai (1999) found no increase in net patient revenue per discharge (their proxy for price) in 43 conversions from nonproﬁt to for-proﬁt status. Access: Community Beneﬁts, Uncompensated Care, Unproﬁtable Services Conversions have caused concerns over reduced access to care in two areas. First and foremost, there is concern that there will be a reduction in uncompensated care (both charity care explicitly provided for the medically indigent and care resulting in uncollectable bad debt). Second, there is a concern that for-proﬁts may be less committed to provide access to unproﬁtable services such as burn treatment, emergency room care, and psychiatric services. Research has tended to focus on the former.3 Moreover, since the missions of nonproﬁt and public hospitals are more likely to include a commitment to the community than the missions of for-proﬁt hospitals (which have a commitment to shareholders), the study of conversion impacts on access have primarily focused on the conversion of nonproﬁt and public facilities to for-proﬁt status. In many conversions, the board of the converting organization creates a contract with the new steward of the hospital which states that a certain level of uncompensated care must continue to be provided post-conversion. For the majority of nonproﬁt hospitals that convert, the provision of

The Impact of Hospital Ownership Conversions

195

indigent care and emergency room visits does not appear to be harmed by the sale to a for-proﬁt; instead the hospitals are better able to fulﬁll their missions after being sold to a stronger organization. The former nonproﬁts’ ability to serve the community, in aggregate, also beneﬁts from the conversion, as the money resulting from the sale can be used to create a foundation to provide beneﬁts to the community (Cain Brothers, 2002). Several research studies have examined the effects of conversion to forproﬁt ownership on access to uncompensated care and found mixed effects. Needleman, Lamphere, and Chollet (1999) found that uncompensated care delivery declined in public hospitals in Florida, which had a substantial commitment to uncompensated care before conversion. However, uncompensated care levels were low in nonproﬁt hospitals before conversion and did not change substantially following conversion. Likewise, using data from California, Florida, and Texas, Desai, Lukas, and Young (2000) found that public hospitals that converted to nonproﬁt status sustained their level of uncompensated care, whereas public hospitals that converted to for-proﬁt status delivered substantially less uncompensated care. Young and Desai (1999) reported no change in the levels of uncompensated care following conversion from nonproﬁt to for-proﬁt ownership. Collins et al. (2001) reported little change in the level of uncompensated care after conversion from nonproﬁt to for-proﬁt status. Most of the hospitals in the sample did not provide high-cost services speciﬁcally for the beneﬁt of the community (burn treatment, trauma care, community outreach, etc.) before the conversion, and as a result, there was little potential for these services to be lost as a result of the conversion from nonproﬁt to for-proﬁt status. Furthermore, the admission of publicly insured patients increased in some of the hospitals. Thorpe et al. (2000) compared the percentage of total expenses devoted to uncompensated care in converted hospitals to that in non-converted hospitals. Their ﬁndings reveal that conversion from nonproﬁt to for-proﬁt ownership reduced expenditures on uncompensated care from 5.3% to 4.7%; conversion from public to for-proﬁt ownership reduced uncompensated care expenditures from 5.2% to 2.5%. By contrast, conversion from for-proﬁt to public ownership increased expenditures from 5.2% to 7.5%. Overall, the research ﬁndings suggest that conversion from public to forproﬁt ownership reduces access to care, whereas conversion from for-proﬁt to public ownership increases access to care. Conversions from nonproﬁt to for-proﬁt status entail little change, primarily due to the low initial levels of uncompensated care provided in nonproﬁts. Sloan, Taylor, and Conover (2000) examined whether conversion from nonproﬁt to for-proﬁt status impacted the provision of several proﬁtable

196

LAWTON R. BURNS ET AL.

services. They found that conversion decreased the likelihood that a hospital performs open-heart surgery and increased the likelihood that it provides MRI and sports medicine. On the other hand, conversion resulted in an increase in emergency room admissions, deliveries, and the fraction of patients covered by Medicaid. Thus, conversion to for-proﬁt status appears to have a mixed effect on the service offerings of a hospital, increasing the availability and utilization of some services while decreasing others. Other researchers have reported no statistically signiﬁcant changes in emergency room visits or births in hospitals before and after conversion, and no signiﬁcant differences in procedures between converted hospitals, controls, their neighbors, or the neighbors of controls (Young & Desai, 1999; Hadley et al., 2001). There is some debate over whether the value of the uncompensated care provided by nonproﬁts is substantial enough to offset the lost tax revenue resulting from nonproﬁt status. Desai, Young, & Lukas (1998) examined ﬁfteen cases of California for-proﬁt hospitals that converted to nonproﬁt status. In eleven of the ﬁfteen conversions, the conversion resulted in an increase in the provision of uncompensated care. However, in only three of the hospitals was the level of uncompensated care valuable enough to offset the lost property taxes that would have went to the community had the hospital remained for-proﬁt. A loss of property taxes can be a substantial problem for a community. Proposition 13 in California and Proposition 2½ in Massachusetts regulate the level of property taxes that communities can charge residents and businesses. Thus, communities in these states may have difﬁculty offsetting the lost property tax revenue that results from the conversion of a for-proﬁt hospital to nonproﬁt status. King and Avery (1999) present opposite results in their case analyses of the Legacy Health System and previously converted hospitals. Their model provided a broad deﬁnition of community beneﬁt, including bad debt, care for the poor, subsidized community programs, education, and research, and positive externalities created by the health system, including employment. The model consisted of an incremental analysis of the community beneﬁts provided by a for-proﬁt hospital paired with a nonproﬁt foundation versus the nonproﬁt hospital in its original state. To conservatively measure the value of beneﬁts produced, the value of community beneﬁt outputs was determined by the cost of the inputs required to produce them. The authors measured the difference between the converted case and the base case in terms of the value of the foundation annuity, employment, corporate taxes generated, impact on market prices, impact on services provided below cost, traditional charity care, unpaid Medicaid care, bad debt, education, research, and community health services. They concluded that the hospital

The Impact of Hospital Ownership Conversions

197

would provide a more substantial level of beneﬁt to the community if it remained as a nonproﬁt organization. The authors noted that covenants in the sale agreement requiring the for-proﬁt to maintain certain community services would likely result in reducing the hospital purchase price, and thus the beneﬁt of such covenants would be offset by the reduced size of the foundation. Claxton et al.’s (1997) review of the literature suggests that while nonproﬁt hospitals provide signiﬁcantly more community beneﬁts than forproﬁt hospitals, the results vary based on the deﬁnition of community beneﬁts provided. When a broad deﬁnition of community beneﬁts is used, extending to charity care, bad debt, losses resulting from public programs, research, and teaching, the amount of community beneﬁts provided by nonproﬁt hospitals outweighs the value of the tax exemption that they receive. However, if the taxes paid by for-proﬁts were counted as community beneﬁts, they would exceed the value of community beneﬁts provided by nonproﬁts. Quality In addition to cost and access, conversions often raise concerns over changes in quality. Across-market comparisons suggest that for-proﬁt hospitals have higher rates of mortality for elderly patients with heart disease than nonproﬁts, whereas within-market comparisons suggest for-proﬁts provide higher quality care than nonproﬁts. Thus, quality may be determined by factors other than ownership status (McClellan & Staiger, 2000). Prior research on the effect of conversion on quality has uncovered mixed results. Farsi (2004) modeled the effect of conversion on acute myocardial infarction and congestive heart failure mortality rates using California hospital data from 1990 to 1998. He found that hospitals that converted to for-proﬁt from nonproﬁt status had signiﬁcantly increased acute myocardial infarction mortality rates post-conversion, while those that converted to nonproﬁt from for-proﬁt status had signiﬁcantly increased congestive heart failure mortality rates post-conversion. Sloan (2002) found that conversion from government to nonproﬁt or for-proﬁt ownership had no effect on hospital mortality, although the rate of complications from pneumonia did increase in conversions to for-proﬁt status. Picone et al. (2002) found that hospitals converting to for-proﬁt from government or nonproﬁt status experienced increased mortality one to two years later, although the decline in quality attenuated over time post-conversion. This phenomenon was not observed in for-proﬁt hospitals that had converted to government or nonproﬁt status.

198

LAWTON R. BURNS ET AL.

Finally, Chen (2002) found that for-proﬁt and government hospitals have a 3–4% greater incidence of adverse outcomes for patients after the treatment of acute myocardial infarction than do nonproﬁt hospitals. Furthermore, the incidence of adverse outcomes increased by 7–9% after a nonproﬁt hospital converted to for-proﬁt status, but did not increase substantially after conversion from government or for-proﬁt status to nonproﬁt status or from nonproﬁt or for-proﬁt status to government status. We should note that making apples-to-apples comparisons of quality between converted and non-converted hospitals may be difﬁcult. Hospitals that convert tend to be ﬁnancially distressed in the ﬁrst place, making them less capable of offering high-quality care. Grabowski and Stevenson (2008) found that there are fundamental differences between nursing homes that chose to convert and those that did not, making it difﬁcult to determine the effect of conversion on the quality of care delivered. However, both before and after conversion, homes that converted from nonproﬁt to for-proﬁt status generally experienced deteriorating quality, whereas homes that converted from for-proﬁt status to nonproﬁt status generally experienced improvements in quality. Thus, although there is evidence that conversion to for-proﬁt status may be correlated with worse quality, it is unclear whether this relationship is causal.

Summary of Prior Research We have sought to summarize the effects of conversion on several measures of hospital operations and performance and, where possible, to distinguish the effects of the six different types of conversion. This review has been only partially successful, since there are few conversion studies and even fewer that examine conversions other than to the for-proﬁt form. What then may we conclude? Most conversions aim to improve negative ﬁnancial performance. Converters to the for-proﬁt form wish to increase their ﬂexibility in ﬁnancing, while converters from the for-proﬁt form seek to unload hospitals that fail to yield acceptable returns. Empirical research supports this conclusion, suggesting that conversions are determined by ﬁnancial distress, low occupancy, and low market leverage, although these ﬁndings are drawn only from studies of conversion to the for-proﬁt form. Case study evidence indicates that conversions lead to increased ﬂexibility by allowing the new for-proﬁt owner to separate the acute care operations of the hospital from its previous charitable activities using a community foundation.

The Impact of Hospital Ownership Conversions

199

With regard to operations, there is some evidence that conversions alter stafﬁng levels, leading to staff decreases (increases) in nonproﬁt conversions to (from) for-proﬁt ownership. All types of conversions seem to entail increases in clinician involvement in hospital decision making, although these ﬁndings derive from only one study. With regard to performance, several types of conversion seem to be associated with increased hospital margins (with the exclusion of for-proﬁt to nonproﬁt conversions); however, many of these types of transitions – especially to the for-proﬁt form – are also associated with an increase in prices, which might help to explain the higher margins observed. With regard to access, public hospital conversions to (from) for-proﬁt status reduce (increase) uncompensated care levels. There is no evidence that hospitals converting from nonproﬁt to for-proﬁt ownership alter their levels of uncompensated care or level of unproﬁtable services. Finally, there are no clear effects of conversion on quality of care. The ﬁndings from the literature on hospital conversions somewhat parallel those of governance changes in industry. Industrial research on the performance outcomes of governance changes typically has focused on the effects of privatization of public ﬁrms on expense, proﬁt, and various other measures of efﬁciency (Bishop & Kay, 1988; Yarrow, 1989; George, 1990). Most of these studies indicate that governance changes have few consistent effects on ﬁrm performance. For example, George, Joll, and Lynk (1992, p. 353) conclude that the evidence ‘‘is by no means clearly in favour of privatisationyAll in all, however, it is difﬁcult to avoid the conclusion that privatisation is not necessary for improving efﬁciency.’’ More recently, Villalonga (2000) summarized the results from both crosssectional and dynamic analyses of privatization and reported that the cumulative evidence was inconclusive. Villalonga’s own longitudinal study of 24 Spanish ﬁrms found that while privatization had an initial negative effect on efﬁciency, in the long-run, privatization increases efﬁciency. Given that most privatization efforts aim to improve ﬁrm performance,4 why do they fail to succeed? Organization theory suggests three plausible explanations. Following population ecology research, market conditions may be a more signiﬁcant determinant of ﬁrm performance than the structure of corporate governance, and ownership changes may occur without concomitant changes in market conditions. Thus, some privatized ﬁrms may have operated in competitive markets while they were publicly owned and thus were already operating efﬁciently. Conversely, public monopolies may become privatized into privately held monopolies where the competitive stimulus for efﬁciency remains weak, causing continued ﬁrm inefﬁciencies.

200

LAWTON R. BURNS ET AL.

Second, following institutional theory, ownership forms may be only weakly linked to differences in ﬁrm behavior and performance. Unless ﬁrms do things differently when under different forms of ownership, the likelihood that a ﬁrm will experience a different performance outcome by changing corporate governance is minimal. For example, if the visible push for privatization during the past two decades has spurred public and nonproﬁt managers to adopt private sector behaviors, corporate governance changes alone would be expected to have minimal effects on ﬁrm performance. Third, following strategic adaptation research, corporate strategy and strategic actions constitute important intervening variables not speciﬁed by researchers that explain variations in organizational performance among ﬁrms experiencing changes in ownership. The change in ownership (one type of strategic change) may foster other types of strategic change that differ by transition type and thus help to explain differences in performance.

Limitations of Current Research Thus far, most of the literature has only focused on one or two of the six varieties of conversions. This is unfortunate, as the conclusions that can be reached from one type of conversion clearly may not be generalizable. A few of the studies that have been presented have shown that the effect of a conversion on an outcome is moderated by the organizational form of the hospital before or after the conversion. Although public (government) and nonproﬁt ownership have at times been grouped, doing so may be unwise, as they may face different operating constraints. There is a need for more research to be conducted on the effect of converting from for-proﬁt status to nonproﬁt or public status. This is surprising, as conversion to nonproﬁt status has been more common than conversion to for-proﬁt status (pre-2000) or nearly as common (2000–2007). There has also been little research on conversion to public status, likely because this occurs relatively infrequently, although it has occurred with greater frequency since 2000. The literature studying hospital governance changes has focused more on the antecedents and outcomes of these conversions than on intervening variables which may link governance changes with outcomes.5 Unfortunately, there is little research on how conversions affect the implementation of corporate strategies. No literature was found on whether conversion had shifted hospitals from bottom-up to top-down change, from a long-term to a

The Impact of Hospital Ownership Conversions

201

short-term horizon, or from decentralized to centralized governance. Some research has shown that conversions improve efﬁciency, but these ﬁndings may be due to higher prices charged.

FIELD STUDY OF SIXTEEN HOSPITAL CONVERSIONS Comparative Case Study Format To address some of the shortcomings in the research literature, we developed a ﬁeld study to evaluate the impact of ownership conversion on changes in hospital strategy process, content, and implementation. The ﬁeld study was part of a wider empirical investigation of the antecedents and impacts of hospital conversions (cf. Sloan et al., 2003a, 2003b; Anderson et al., 2001). The larger study analyzed conversions among the universe of hospital mergers and acquisitions which encompassed ownership changes; the database for this universe was maintained by Irving Levin & Associates. Thus, we study only hospitals that converted ownership as part of a merger or acquisition, not independent hospitals that just changed ownership. Twenty hospitals were selected for case study analysis. Cases where the merger and acquisition involved no change in the hospital’s tax status (e.g., one nonproﬁt hospital acquired by another) were deleted. We sought to include hospitals engaging in multiple types of ownership conversions in different parts of the country. Of the 20 hospital sites identiﬁed, 16 agreed to participate in a series of interviews. The ﬁeld study thus utilized a comparative case study of sixteen hospitals. These hospitals were studied 3–6 years after the conversion occurred. The ﬁnal sample of sixteen provided multiple respondents in three transition types – nonproﬁt to for-proﬁt, for-proﬁt to nonproﬁt, and public to nonproﬁt – and one hospital that transitioned from public to for-proﬁt. Our conclusions are thus limited to these four transition types. Case studies suffer from weaknesses in both internal and external validity. Internally, case studies lack a control group, are unable to identify causation, and are unable to distinguish between multiple treatment variables. For example, since hospital conversions may be associated with joining multi-hospital chains, an observed effect may be attributable to either change. Externally, case studies may not be generalizable to a broad population of organizations. Although comparative case analysis does little

202

LAWTON R. BURNS ET AL.

to improve causal inference or generalizability, it does enhance the researcher’s understanding of the organizational process of change and possible differences in the change process across conversion types. We collected data on the impact of conversion on strategy entirely through structured interviews that asked respondents to identify changes by making pre- vs. post-conversion comparisons. To maximize study resources, the investigators initially conducted on-site visits and face-to-face interviews at ﬁve of the study sites. The site visits were utilized to help train all of the investigators in standardized data collection. Data from informants at the remaining 11 sites were collected through subsequent telephone interviews. The same structured interview protocol was used throughout the study (Appendix A). At each study site, investigators worked with the Chief Executive Ofﬁcer (CEO) to identify 5–8 key stakeholders – both administrators and clinicians – with sufﬁcient information and/or tenure to comment on the conversion’s impact. We interviewed a diverse set of informants: CEO, Chief Operating Ofﬁcer, Vice-President for Strategic Planning, Vice-President for Marketing, Chief Nursing Ofﬁcer, members of the hospital Governing Board, service line administrators, and active physicians on the medical staff. If organizations did not have individuals with these speciﬁc titles, appropriate members of the management team were interviewed in their place. In each case, participants had varying tenure with the institution, positions at the hospital, and membership in key stakeholder groups. Many participants were present at the hospital both pre- and post-conversion. In each case, the mix of stakeholders represented a broad pre- and post-conversion time frame. In this manner, we were able to elicit a range of answers to the study questions and triangulate the results. Given the absence of a literature regarding multiple types of ownership changes, we utilized a partially grounded theory approach (Glaser & Strauss, 1967) to develop the study questions. After reviewing the literature (cf. Gray, 1986) on differences between hospital ownership types (for-proﬁt vs. nonproﬁt vs. public), we convened a preliminary focus group of executives from four East Coast hospitals that had undergone governance transitions (representing nonproﬁt to for-proﬁt, public to for-proﬁt, and public to nonproﬁt) to identify the changes in their strategic decision making and strategic orientation. Using this information, we developed a standardized template of questions for obtaining strategic information from participants in the present study. Interview questions were mostly qualitative and open-ended to obtain information without leading participants or skewing their opinions.

The Impact of Hospital Ownership Conversions

203

Interviews were structured according to the interview guide, which all investigators and research assistants utilized throughout the two-year study. On-site interviewer training was used to standardize data collection efforts and reduce variation; all interviews were tape recorded to fully capture informants’ responses and ensure consistency in methods. We transcribed the interviews and collated answers to the same questions provided by informants at a given hospital. The investigators then distilled the comments from the multiple stakeholders to chronicle the changes in that hospital’s strategic content and process following ownership conversion.

Data and Description of Case Hospitals To best understand the motivations for and effects of hospital conversions on organizational strategy, we analyzed various conversion types. Most of the literature on hospital conversions evaluates hospitals converting from nonproﬁt or public to for-proﬁt ownership. Our study includes these conversions but also evaluates two other, important conversion types – from public to nonproﬁt status and from for-proﬁt to nonproﬁt. Appendix B describes the 16 hospitals studied and how they map across these transition types. We were also able to discuss conversions from stand-alone facilities to hospitals that joined multi-hospital systems, and vice-versa. Although this perspective is not the main focus of this research, the data allow us to evaluate these trends and their effects on hospital strategy.

FIELD STUDY RESULTS Intended Strategy: Motivations for Governance Change In addressing motivations for the conversion, we sought to understand why the hospital changed ownership and whether or not the tax status of the new owner was critical to the decision-making process of the hospital. Congruent with the literature, our ﬁndings indicate that a consistent set of motivations guided hospitals seeking conversion, regardless of conversion type. Nevertheless, there are some important differences across conversion types; to some hospitals, the tax status of the new owner was an important criterion in their decision making.

204

LAWTON R. BURNS ET AL.

Consistent Motivations for Governance Change – Across Transition Type Three primary motivations cut across all of the conversions studied. First, ﬁnancial considerations appear to be a dominant motivation for changes in ownership, although they have many different manifestations depending on the situation. In every case except one, the hospital reported ﬁnancial trouble or anticipated future ﬁnancial difﬁculties. In extreme cases, hospitals faced closure due to a lack of operating revenue. More generally, hospitals felt unable to make necessary capital investments to modernize the facility and service lines offered to patients under current ownership. This inability stemmed from an unwillingness of current owners to make new capital investments – either due to the poor ﬁnancial status of owners or because a hospital chain no longer prioritized the hospital as part of a regional strategy. Other changes beyond ownership were contemplated as ways to address ﬁnancial difﬁculties. Typically, a stand-alone hospital in ﬁnancial difﬁculty tended to look to form an alliance or join a multi-hospital chain as a way of improving likely ﬁnancial inﬂows. However, some ﬁnancially strapped hospitals were already members of multi-hospital systems and reported that the system merely drained ﬁnancial resources from the facility without making appropriate, compensating investments in it. Public hospital managers reported that government ownership precluded access to bond markets and political leaders resisted raising revenues through increased taxation. Second, a rapidly changing health care marketplace and the need to garner managed care contracts served as a powerful motivation for conversion. Nearly all facilities reported an inability to compete for managed care contracts, and this inability contributed to the aforementioned anxiety about future ﬁnancial performance. Freestanding facilities reported that competing health care networks were more successful at obtaining such contracts, and often sought to enter alliances in the hopes of improving their managed care contracting position. Several facilities reported a fear that a dominant ﬁrm like Columbia would purchase and shutter the facility as part of a regional strategy which prioritized its other local hospitals. Third, these ﬁnancial and market forces seemed to overpower decisionmakers’ preference for a certain ownership type. For instance, most public and nonproﬁt hospitals expressed a preference for a nonproﬁt buyer, yet settled for a for-proﬁt buyer based on the perception that they would invest more resources in the facility and be better at obtaining managed care contracts. Signiﬁcantly, a number of hospitals, especially those converting from nonproﬁt to for-proﬁt status, reported that their perceptions of market changes – ranging from the importance of managed care contracting to estimates about Federal reimbursement changes – proved to be exaggerated

The Impact of Hospital Ownership Conversions

205

over time. For-proﬁt entities did not seem to have a preference for buyer type, but in cases where the local hospital was exerting pressure to leave a for-proﬁt hospital chain, the local hospital seemed to hope for increased local control and decision-making upon conversion. Differences in Motivation for Conversion – by Transition Type A myriad of other considerations were also mentioned as important motivating factors, ranging from the geographic position of the hospital to the personal preferences of a dominant hospital leader. Many of these considerations were noted in some transition types more than others. In hospitals transitioning from nonproﬁt to for-proﬁt ownership, a host of speciﬁc ﬁnancial motivations were present. The aforementioned need for capital dominated the motivation for conversion. One hospital board felt that an inappropriate level of cash ﬂow was being siphoned off to support other, less proﬁtable, hospitals within the system. This board believed these resources were being used to prop up poorly managed hospitals in other parts of the country. More generally, hospitals believed they either lacked access to capital needed to grow as a freestanding facility or that the parent system failed to make needed capital investments in the facility. Still other facilities feared closure of the hospital or important service lines due to lack of ﬁscal resources. Nonproﬁt facilities unable to survive on operating proﬁts and debt sought equity capital to stay aﬂoat. Second, an acquirer’s offer often created additional ﬁnancial incentives for managers to support conversion. These included an attractive purchase price by the acquiring for-proﬁt system and the offer to create a community-based, nonproﬁt foundation (using some of the proceeds of the sale) to promote the health and education of the community. Third, aforementioned fears regarding access to managed care contracts, anticipated loss of reimbursement from all payers, and the hospital’s inability to control its costs seemed particularly acute for nonproﬁt owners who felt for-proﬁt managers would excel at these tasks. Beyond ﬁnancial incentives, there were two other motivations for ownership changes relating to leadership and geographic strategy. In one case, a strong physician executive headed the hospital and guided it through sale and conversion. In another case, the hospital was not considered ‘‘strategic enough’’ for further investment by its current parent while the acquiring for-proﬁt system had made other investments in the same geographic region and thus felt that the conversion might leverage those other investments. In such instances, the nonproﬁt hospital preferred to be acquired by another nonproﬁt; however, the need for capital and other resources were more pressing criteria than the ownership of the new parent.

206

LAWTON R. BURNS ET AL.

In hospitals transitioning from for-proﬁt to nonproﬁt ownership, on the other hand, the ﬁnancial motivations were slightly different and often centered around tensions between the hospital’s management team and the national or regional corporate structure of the parent company. One hospital’s management team felt that the fees it had to pay to the investorowned parent company were larger than the value that company was offering in management support. In another instance, the hospital felt that its for-proﬁt parent had failed to make needed capital investments, had reduced stafﬁng to inappropriate levels, and had contributed to lower employee morale. In yet another case, the for-proﬁt had entered a management contract with the hospital, and the hospital board felt that the for-proﬁt managers had failed to appropriately increase cash ﬂow, add new services, and plan for the long term. From the perspective of the forproﬁt hospital chain, managers noted that corporate representatives complained that the hospital was not fulﬁlling its intended strategic role within the system. In one case, the larger system had clearly shifted its strategy and was no longer interested in expanding into the relevant region. Other motivations in the for-proﬁt to nonproﬁt transitions often concerned antipathy to large systems and investor-ownership. In one case, the Federal Trade Commission, citing anti-trust concerns, forced the forproﬁt system to divest the hospital after it had acquired a neighboring hospital. In another case, community activists pushed for the divestiture decrying the outﬂow of hospital and community resources. Here the decision was made to return to freestanding, nonproﬁt status rather than be acquired by another system. In other cases, however, there was less concern regarding the tax status of the new hospital owner. Finally, in hospitals transitioning from public to private (both nonproﬁt and for-proﬁt) ownership, the major ﬁnancial motivations concerned beliefs that future public funding sources – including Medicaid reimbursements and disproportionate share hospital funding – would be consistently reduced. In addition, public managers expressed an inability to obtain accounts receivable in a timely manner and poor accounting practices. Other major motivations concerned the limitations that public ownership placed on competitive strategy and positioning. In one hospital, for example, strategy sessions were by law open to the public and thus to competitors. Strategic decisions were also strongly inﬂuenced by local public bodies to take or not take key actions. In another hospital, public ownership (and political appointments) limited the quality of executives that could be recruited and fostered high administrative turnover. Public facilities also noted that political limitations requiring the hospitals to restrict its patient catchment

The Impact of Hospital Ownership Conversions

207

area to the relevant tax base often precluded effective managed care contracting. Finally, due to public ownership, one hospital switched to nonproﬁt status to form an alliance with a local nonproﬁt hospital and thereby avoid a public bidding process and potential sale to a for-proﬁt system interested in closing the facility to favor its locally owned hospitals.

Realized Strategy: Change in Strategic Content To understand changes in strategic content that may have resulted from ownership changes, we inquired into the major differences in strategy pursued by the hospital before and after the conversion. In many cases, the strategy may have remained the same, but methods of pursuing the organization’s strategic intent may have altered. In contrast to the preceding section, we classify our analysis of changes in strategic content according to various critical, strategic competitive areas (described later). In each area, we note some of the differences between transition types with respect to changes in strategy. To understand changes in overall strategic content, we asked key informants open-ended questions about their hospital’s strategic goals. For instance, we asked about changes in hospital strategy that occurred along with the conversion, i.e., which new strategies were being pursued after the conversion and whether any major previously pursued strategies were being discarded as a result of the conversion. In many cases, certain strategic goals – i.e., improving patient referrals in light of increased market competition – remained the same before and after the conversion. As a result, we placed a special emphasis on understanding whether certain new strategies to achieve such goals could have been enacted effectively under the previous owner. In addition to asking open-ended questions regarding major strategic goals and an institution’s means for achieving them, we explored speciﬁc hospital strategies with respect to important competitive activities. The list of important competitive activities was generated from a review of the health care policy, management, and strategy literatures and supplemented by the focus groups’ comments. We are conﬁdent that these areas represent areas of high strategic interest for hospital managers and policy makers. These important competitive areas include values (hospital mission, philosophy, provision of charity care, and organizational culture), market competition (alliance development, managed care contracting, and pricing), ﬁnance (operational ﬁnances, capital investments), and operations (service and department creations, eliminations, consolidations, continuum-of-care development, cost-containment, and quality improvement).

208

LAWTON R. BURNS ET AL.

Strategic Content: Values As previously noted, motivations for conversion focused around ﬁnancial viability and management efﬁciency. Only rarely did institutional values – i.e. a dedication to provide the best quality care to a broad range of patients including the underserved – explicitly drive a hospital’s motivation to convert. In fact, most managers denied that the conversion led to any signiﬁcant change in their organization’s explicit values. Nevertheless, issues regarding changes in the hospital’s mission and the provision of charity care have been important parts of the public policy debate and the research literature on conversions reviewed earlier. Among transitions from nonproﬁt and public ownership to for-proﬁt ownership, a number of institutions had speciﬁed in the contract that the new, for-proﬁt hospital must fulﬁll a certain amount of charity and community care. According to managers surveyed, this requirement was largely fulﬁlled. However, there seems to be little explicit oversight to validate this claim. Although many institutions reported changes in organization culture upon conversion, few cases noted any change in the hospital’s explicit mission or philosophy. In fact, in most cases, informants used the same language to describe the hospital’s mission before and after the conversion. This phenomenon remained true across conversion type. The exceptions were the few cases in which a public or for-proﬁt hospital converted to nonproﬁt ownership under a Catholic health system. In these cases, the mission and philosophy of the hospital seemed to reﬂect the Catholic health system’s values (e.g., policy dealing with abortion services, etc.).6 Changes in organizational culture usually were more closely related to issues of labor and stafﬁng, cost-containment, and the ability of the hospital to retain localized decision making than to issues relating to the hospital’s mission, philosophy, or ability to provide charity care. Changes in organizational culture consistently involved an increased emphasis on ﬁnancial and marketplace performance, accountability and entrepreneurship for managers and department heads, and operational efﬁciency. This remained true for every conversion type, including nonproﬁt to for-proﬁt and vice-versa. In certain nonproﬁt to for-proﬁt transitions, especially when a freestanding facility was acquired by a larger health system, the organization often had a more difﬁcult transition period and suffered from a negative or cynical organizational culture. Strategic Content: Market Competition In almost every case, market competition issues played a role in motivating the change in ownership and subsequent strategic changes. The most

The Impact of Hospital Ownership Conversions

209

common and highly prioritized shift in strategy concerned the formation of alliances to improve an institution’s bargaining position with managed care payers. This usually took the form of the hospital joining a larger, more dominant hospital player in the same market. This strategic change was closely linked to whether or not a hospital was converting from a freestanding facility to being a part of a larger system and was consistently mentioned across nonproﬁt to for-proﬁt and public to private (nonproﬁt and for-proﬁt) conversions. Key informants in the few hospitals transitioning from for-proﬁt to nonproﬁt status did not emphasize this point, however. Signiﬁcantly, despite the lapse of several years time since the change in governance, informants were uncertain whether such alliances had improved the hospital’s patient referral base or managed care market position in many of the sites. Although many informants at post-conversion hospitals were eager to discuss their efforts to secure managed care contracts, interview comments question the impact of conversion on contracting. Informants at many hospitals – across transition types but primarily nonproﬁt to for-proﬁt and public to private – voiced concern that the new ownership did ‘‘less than expected’’ to win managed care contracts. Speciﬁcally, many informants felt that the hospital’s overall approach to managed care contracting took a backseat to efforts to reduce internal labor costs. As a result, despite joining multi-hospital health systems or partnering with larger, regional facilities, many hospitals continued to struggle to win managed care contracts. This trend is likely due to high pre-conversion expectations (and in some cases misleading pre-conversion promises from potential suitors) as well as the new owner’s general inability to affect the local managed care marketplace, at least in the short term. Nevertheless, most key informants continued to believe that alliances with other hospitals were important to remain competitive. They speciﬁcally reported that alliances that increase market share would provide them with the market power needed to improve their managed care contracting efforts. These ﬁndings support our earlier remarks that strategic changes are more likely observed in the short-term period following conversion, whereas shifts in competitive advantage may take much longer to occur. In this case, a lack of short-term changes in managed care contracting strategies accounts for the longer-term failure to improve managed care contracts which was observed three to six years post-conversion. Finally, these alliances often had important implications for service availability (e.g., investments in technology and high-margin services) and pricing (e.g., leverage managed care for higher reimbursement).

210

LAWTON R. BURNS ET AL.

In almost every case, key informants also claimed that the hospital had tried to improve alliances with physicians after the conversion. This was true for changes from nonproﬁt to for-proﬁt, for-proﬁt to nonproﬁt, and public to private (nonproﬁt and for-proﬁt) ownership and is consistent with prior research. Reported success in implementing this strategy varied and did not appear related to conversion type. Nevertheless, physicians did seem to form closer alliances with the new ownership of many of the hospitals, especially institutions in which the pre-conversion hospital was facing severe ﬁnancial difﬁculty or closure. This suggests that a shared sense of ﬁscal distress facilitates such alliances. Two issues related to physician alliances are important to note, however. First, in almost every case, the pre-conversion leadership of the hospital was trying, often unsuccessfully, to improve physician relations and alliances. Although new ownership strategies may have been more successful than preconversion strategies, strategic intent mostly remained consistent throughout the conversion. Many physician alliances were developed and enacted under the pre-conversion ownership regime, and the post-conversion leadership merely continued to follow long-standing physician–hospital organization strategies. In cases where the conversion resulted in keeping open a hospital slated for closure, physicians seemed to grant the facility’s new ownership a certain ‘‘honeymoon’’ in which relations seemed markedly improved. In these cases, even if relations soured after this initial honeymoon, physicians did acknowledge that while less than ideal, the ownership change allowed the institution to continue to exist. Second, physician relationships seemed to be related to the level of capital investment and operational revenue infusion by the new owner. For example, if the new ownership made important capital investments (e.g., improving the physical plant and/or expanding service lines) physicians seemed more likely to follow-through on stated desires to work with the hospital, and vice-versa. Strategic Content: Finance Financial issues were major motivations in every conversion but one in this study, and changes in ﬁnance were a major, recurring theme in discussions regarding strategic content. First, in each case, revenue from the sale or lease of the facility improved the ﬁnancial position of the hospital (e.g., through debt reduction). Second, with respect to operating revenues, strategies and practices seemed more closely related to whether or not a freestanding hospital joined a larger multi-hospital system than to conversion type. Almost all freestanding facilities that joined a multi-hospital system reported an infusion of working capital or new operating revenues (e.g.,

The Impact of Hospital Ownership Conversions

211

through the sale of new services) that, in some cases, prevented closure of the hospital. Facilities that switched from one multi-hospital system to another also reported an infusion of revenue, presumably because the new system saw greater strategic value in the institution than the previous owner. Facilities that left a multi-hospital system to become a freestanding hospital reported less ‘‘revenue leaving the local facility for corporate headquarters’’ and an improved balance sheet. Facilities that remained freestanding were less likely to report signiﬁcant improvements in operating revenues. These trends held true for nonproﬁt to for-proﬁt, for-proﬁt to nonproﬁt, and public to private conversions. Capital investment strategies – which were often thought to represent the new owner’s long-term commitment to the facility – reinforced these trends. Capital investments improved in most of the hospital sites as part of the conversion process along one of two avenues. First, hospital owners sometimes made important capital investments before the sale of the facility. For example, hospitals would improve their emergency rooms, purchase new equipment, and generally improve the physical plant before the conversion. In certain cases, these investments were made to attract buyers, while in others they were stipulated in the agreement of sale. Although occurring across transition type, these practices seemed more common among nonproﬁt hospitals converting to for-proﬁt ownership than in the other types. Second, as with operating revenues, most hospitals reported increased capital investments for facility improvements to expand service lines. Speciﬁcally, investments seemed to target improvements in certain typically proﬁtable services such as radiology (many facilities obtained new radiographic equipment such as CT scanners) and cardiology (many hospitals reported creating or expanding catheterization laboratories). Again, this trend seemed particularly signiﬁcant among transitions from nonproﬁt to for-proﬁt, but was clearly not limited to that transition type. Finally, although most hospitals did report improved operating revenues and a certain amount of capital investment resulting from the conversion, many facilities felt that the amount of ﬁnancial investment in the hospital was less than promised as part of the sale or conversion. In some cases, these promises were optimistic assumptions made by managers of the previous regime. However, in some cases, these were explicit promises made by representatives of the new owner – although rarely written into the contract of sale. Many facilities, especially those that underwent a public to private or nonproﬁt to for-proﬁt conversion, reported that they had been led to believe that levels of capital investment would be much higher and much faster than what actually happened. This sentiment was emphasized among

212

LAWTON R. BURNS ET AL.

the subset of cases in which a freestanding facility joined a larger multihospital chain. As a result, many hospitals felt they had been misled about the level of investment a new owner would bring to the facility. Strategic Content: Operations In each case, the most quickly enacted and aggressive changes occurred in the area of hospital operations. In particular, three operational areas experienced signiﬁcant changes following conversion. First, hospitals reduced labor and stafﬁng to cut costs. Second, hospitals sought to centralize and streamline the purchasing of supplies to lower costs. Third, hospitals consolidated and regionalized their service lines (where appropriate) to create a network approach to the provision of health care. These operational changes were pursued much more prominently than other strategic changes (e.g., managed care contracting or quality improvement). Service line consolidation was less pervasive than the other two operational changes, however. Almost every hospital in the study reported that one of the most signiﬁcant, early steps taken by the new ownership regime was stafﬁng reductions. In cases where a freestanding facility joined a multi-hospital chain as part of the conversion, the system quickly implemented centralized ﬁnancial management and employed operational benchmarks (i.e., system stafﬁng ratios) to justify these reductions. Across the board, nursing and ancillary staff were hardest hit. In some cases, perhaps reﬂecting the new owner’s management style than tax status, signiﬁcant labor cuts were made in a very short amount of time. In other cases labor cuts were phased in. However, across almost all cases, labor cuts were signiﬁcant and often led to a decrease in morale. In some cases, labor cuts reportedly reduced the quality and safety of patient care: respondents reported that nurse stafﬁng levels were ‘‘dangerously low’’ and that care providers often were concerned about patient safety. In the most egregious cases, hospitals tended to reverse course after initial cutbacks, but the threat to quality of care remained a real concern for multiple respondents at these institutions. In most cases, the severity and speed of the labor cuts took the acquired hospital ‘‘by surprise,’’ although such cuts ‘‘were expected’’ in cases where a long-standing management company bought the hospital as part of the conversion. Other new operational strategies also received high priority. Centralized purchasing was quickly implemented along with stafﬁng cuts. New owners, especially multi-hospital systems, quickly implement centralized purchasing by standardizing equipment and supplies across facilities. Another more time-consuming strategy involved service line expansions, reductions,

The Impact of Hospital Ownership Conversions

213

eliminations, and regionalization. Expanded services tended to include proﬁtable specialty services such as radiology and cardiology, as well as targeted patient markets such as obstetrics and neonatal pediatrics. New hospital owners seemed more likely than previous managers to drop poorly performing services, although this practice was limited to a few cases and not related to conversion type.7 Efforts to create a continuum-of-care and promote quality improvement programs were not prioritized but seemed to continue throughout the conversion.

Changes in the Strategy Process As a ﬁnal perspective on the impact of ownership changes, we investigated changes in the strategic planning and decision-making processes that resulted from the conversion. For example, how were important strategic decisions made, agreed upon, and implemented? What were the implications of changes in these processes? When understood in the context of the intended strategy behind the hospital’s ownership change, answers to such questions further elucidated the impact of ownership changes on strategy. The main area of inquiry concerned how strategic decisions were made. Who was making the decisions (board members, corporate ofﬁces, local management, physicians)? What was the balance of power between strategic decision makers? What role did external and internal consultants play? This line of inquiry also explored the time cycle by which decisions were made, the role of strategic planning, and role of information management. Finally, we also considered questions about the budgeting process and resource allocation (short and long term). Overall, we found that changes in strategy development and decision making varied tremendously across hospital sites. Although each type of organization seemed to have its own strategic decision-making style, changes in strategy process were more closely related to whether an organization belonged to a multi-hospital system or was a freestanding facility. Regardless of conversion type, many freestanding facilities joining multi-hospital chains reported that the locus of decision making almost always shifted from the local facility and its management team and board to a corporate ofﬁce – either headquarters or the ofﬁce of a regional director. This pattern includes several similar characteristics. 1. Nearly all signiﬁcant strategic planning and budgetary decisions required corporate approval, with local managers asked to submit requests and

214

2.

3.

4.

5.

LAWTON R. BURNS ET AL.

ideas. In return, corporate ofﬁces would approve, disapprove, or request more information. Local boards would almost always be retained, but their power diminished from a decision-making body to an advisory one. Speciﬁcally, these local boards were platforms for physician and community involvement, but budget authority was entirely relegated to the corporate board or corporate decision-making body. In cases where a for-proﬁt company purchased a nonproﬁt hospital, board members sometimes left, choosing instead to join the board of the associated nonproﬁt foundation that was created during the sale. More likely in cases where a public or nonproﬁt converted to a for-proﬁt, prior community needs assessments, if conducted, did not drive strategic decision making as much as uniform plans from national headquarters. In many cases, the terms of the sale stipulated speciﬁc or general capital investments that overshadowed the organization’s capital allocation process in the short term. In cases where a strong local leader continued to play a major role in hospital decision making, the process of decision making often relied upon loose, social relationships. In these situations, the basis for determining several strategic and resource allocation decisions switched from interpersonal relationships and networks to more formalized, bureaucratic criteria. Decision-making cycle time – or the length of time taken by corporate headquarters to approve or disapprove of requests – varied with the priority level corporate leaders placed on the issue in question. For many things that were a priority to the corporate leadership, decisions were made quickly and implementation of these decisions improved. However, for requests that were not directly related to explicit corporate goals, respondents noted surprising delays in corporate decision making. Indeed, part of the surprise was attributed to high pre-conversion expectations that private or for-proﬁt owners make decisions more quickly and implement them more efﬁciently.

All of these changes reﬂect the role of a large, bureaucratic, and centralized hospital system driving strategic decision making and capital allocation at a local hospital facility. However, even within multi-hospital systems, there was tremendous variation in how corporate leadership conducted the strategic decisionmaking process for each hospital. For instance, Columbia, Quorum, and NetCare – all for-proﬁt multi-hospital systems with often very similar

The Impact of Hospital Ownership Conversions

215

approaches to strategic content – had very different approaches to strategic process. Columbia relied more on national guidelines and rules to drive strategic decision making, while NetCare seemed to be more locally oriented. NetCare engaged in community-based needs assessments and offered more local control to the hospital’s management team. Even within a single hospital corporation, different fully owned and operated hospitals were treated very differently based on the corporation’s level of perceived importance and involvement in the hospital (as viewed by the hospital). For instance, across the various Columbia hospitals in the study, strategic process development differed greatly. Some facilities that were apparently more integral to Columbia’s growth strategy reported great responsiveness, consulting support, guidance, and investment. Other Columbia facilities reported a hands-off approach to strategic decision making and an unwillingness to make capital investments unless absolutely necessary. Finally, even within a single corporately owned hospital, the degree of system responsiveness, guidance, and successful coordination could vary. In one case, a national ﬁrm was seen as unresponsive, unwilling to meet commitments (causing resentment among local hospital stakeholders), and causing severe delays to and stagnation of the strategic decision-making process immediately following a public to for-proﬁt conversion. However, as the local management learned how to effectively work within the culture of the new corporate ownership (a process assisted greatly by a headquarters-trained CFO joining the local management team), they reported improvements in the structure, process, and speed of strategic decision making. Although an analytical framework relying upon the multi-hospital system versus a freestanding facility does explain much of the observed variation in decision-making changes, there are also minor differences in the pattern of strategic decision making by transition type. In governance transitions from nonproﬁt to for-proﬁt, most strategic decisions were submitted by the hospital to the corporate ofﬁce for approval. By contrast, nonproﬁt chains were oftentimes more likely to be oriented to local decision making than forproﬁt hospital chains. Likewise, among transitions from public to private governance (both nonproﬁt and for-proﬁt), some decisions were now made centrally that were previously handled at the local hospital level. In contrast to the nonproﬁt to for-proﬁt transition, however, the shift in decision-making locus was frequently applauded. In hospitals formerly under public ownership, management often cited political restrictions to engaging in strategies that were viewed as necessary for achieving marketplace success. In one case, a

216

LAWTON R. BURNS ET AL.

hospital could not advertise, engage in network alliances, or otherwise seek to treat patients that lived outside of a narrowly deﬁned service area because it was owned and operated by the county government. In other situations, funding for capital purchases was unavailable as the county’s political leadership fought any effort to increase the hospital tax to raise funds for modernization. In another case, a public hospital sought to improve its payer mix to improve its funding base without being politically pressured into offering certain services. In the most egregious example, a county hospital was denied permits, funds, and the approval to expand into proﬁtable service lines in part because the Mayor had a close relationship with the leadership at the competing hospital. Finally, the change in governance shifted the criteria for strategic decision making away from community-based health assessments and political concerns toward the hospital’s ability to effectively compete in the marketplace.

SUMMARY Strategic adaptation is a process of repositioning the ﬁrm in the face of environmental pressures to achieve better ﬁt and improved performance. Ownership conversions represent an ongoing adaptation made by U.S. hospitals. However, due to the prevalence of three different ownership models (nonproﬁt, for-proﬁt, and public), hospitals can undertake different types of ownership transitions which can be motivated by different rationales and external pressures. Moreover, the different transition types may yield different performance outcomes. This chapter has sought to extend strategic adaptation research by specifying important organizational changes in strategy and process that executives initiate following conversion to improve ﬁt and performance. This study has investigated the impact of changes in ownership on the content and process of hospital strategy and certain areas of hospital operations. The investigation was prompted by the continuing trend toward ownership conversions and the inconsistent ﬁndings in the literature that such changes improve organizational performance. We felt that prior research failed to include rigorous analysis of theoretically important intervening variables such as strategic content and process which might help to explain these earlier results and may offer novel insights about hospital conversions and related policy issues. In this light, we investigated whether different types of ownership transitions display different patterns of strategy and decision making that might plausibly be related to performance differences.

The Impact of Hospital Ownership Conversions

217

Our analysis suggests there are more similarities than dissimilarities in strategic content and process across conversion types. That is, regardless of the type of ownership transition undertaken, one observes similar intended strategies, realized strategies, and processes of strategic decision making. With regard to intended strategies, nearly all ownership transitions are motivated by resolving ﬁnancial difﬁculties and accessing capital, competing for managed care contracts, and surviving in more competitive markets. Thus, ownership conversions (of all types) are motivated primarily by real and perceived resource deﬁciencies. With regard to realized strategies, nearly all ownership transitions involve greater emphasis on ﬁnancial and marketplace performance, greater operating efﬁciency, greater accountability among hospital managers, the formation of hospital systems and networks to leverage managed care, the use of strategic alliances with physicians, access to new sources of operating revenues, decreases in stafﬁng, the use of centralized purchasing, and the consolidation of service lines. With regard to the strategy process, nearly all transitions (especially those in which a free standing institution joins a hospital chain) involved centralizing decision making at regional or corporate levels. At the same time, we do observe variations in strategic content and process across transition types, which may have important implications for policy discussions. With regard to intended strategy, transitions to for-proﬁt governance were typically motivated by both ‘‘push’’ and ‘‘pull’’ factors. The former include dissatisfaction with the siphoning off of hospital cash ﬂow and the lack of strategic consideration accorded the hospital by the former parent; the latter include the ﬁnancially attractive offer by the prospective for-proﬁt parent and the disposition of the proceeds from selling the hospital’s charitable assets. Transitions to nonproﬁt and public governance, on the other hand, were typically motivated only by push factors: the failure of the previous for-proﬁt owner to fulﬁll its contractual obligations and earn its management fee, the previous owner’s excessive stafﬁng cuts or poor management, government antitrust challenges, and anticipated shortfalls in public ﬁnancing. We have also uncovered variations in realized strategy among transition types. Ownership transitions away from public and toward for-proﬁt status typically involved joining or forming hospital systems to improve bargaining power over managed care payers. By contrast, governance transitions from for-proﬁt to nonproﬁt status never involved this strategy. This may be explained by the fact that most investor-owned hospitals already enjoy the beneﬁts of (for-proﬁt) system membership and seek other advantages in switching to nonproﬁt ownership. Transitions to for-proﬁt

218

LAWTON R. BURNS ET AL.

ownership were also more likely to involve capital improvements both before and after the sale to attract buyers and bolster proﬁtable service lines. Overall, our ﬁndings suggest that the different transition types exhibit strategies and decision-making patterns that are more similar than dissimilar. As we have suggested earlier, the major source of variation in observed hospital strategy was whether or not the hospital was free-standing or a multi-hospital system member. For example, freestanding hospitals that changed ownership but remained freestanding were the only respondents who failed to report revenue improvements from the transition. Freestanding hospitals that joined systems typically over-estimated the level of capital investment made by their new corporate parents. Similarly, freestanding hospitals were least likely to report the centralization of decision-making regarding strategy since there was no new corporate parent or regional ofﬁce. We admittedly do not have enough data to fully support this argument, however. Moreover, our sample of hospitals (which includes hospitals undergoing both acquisitions and ownership changes at the same time) does not allow us to disentangle these effects.

DISCUSSION Our ﬁndings indicate that a series of important changes in strategy content and process occur with nearly all hospital conversions. These changes occur across the types of conversion studied and the underlying rationales for conversion. As such, they represent a generic ‘‘conversion effect’’ and have important implications for policy making around hospital ownership changes. Our ﬁndings further suggest that strategy content and process are important variables mediating the relationship between hospital conversions and performance outcomes. For example, the positive impact of conversion on hospital ﬁnances may be due to two important changes in hospital strategy and operations: leveraging managed care for higher prices using system/network alliance strategies and downsizing hospital staff. These effects are suggested in the review of related studies of ownership conversions (Anderson et al., 2001; Picone et al., 2002). Those studies found that work restructuring and price increases occur in all types of ownership transitions, consistent with the existence of a generic conversion effect. We further suggest that the conversion effect may consist of two parts: the generic change in ownership (regardless of type) and the change from freestanding status to multi-hospital system membership. We noted earlier that these two transitions are frequently combined, posing a threat to

The Impact of Hospital Ownership Conversions

219

internal validity. In interpreting our case ﬁndings, we have sought to carefully separate the two effects. Future research on ownership conversions should use multivariate analytic methods to separate these two effects empirically. We also surmise (but cannot deﬁnitively prove) that, based on our interviews, the speciﬁc changes in ownership studied here (e.g., from nonproﬁt to for-proﬁt versus for-proﬁt to nonproﬁt) are less important than the shift between freestanding and hospital system membership. Respondents mentioned that the intended strategies of resolving ﬁnancial difﬁculties and preparing for more competitive managed care markets were stronger motivations than the hospital’s preference for any speciﬁc ownership type. Similarly, observed variations in the strategy process were more strongly related to system membership than to transition type. If true, our ﬁndings supply another possible explanation for why previous conversion studies have found mixed or no evidence for performance improvements. Reviews of the impact of multi-hospital systems have found no improvements in efﬁciency or patient care (Shortell, 1988; Burns & Pauly, 2001; Dranove & Lindrooth, 2003). To the extent that ownership changes entail changes in system membership and that system membership effects dominate, we might expect few resulting differences in performance. We also advance two additional explanations for the lack of signiﬁcant differences across conversion types in performance outcomes. First, our ﬁndings suggest that despite the change in ownership, there is no real change in the organization’s course. That is, despite the variations in ownership change (the exogenous variable), there is little variation in the theorized intervening variables of strategy formation and implementation. Among our sample of hospitals, the change in ownership was accompanied by changes in realized strategy in several areas (market competition, ﬁnance, and operations) but not in overall values or mission. In several ownership transitions, the new owner kept intact strategies enacted before the conversion. Indeed, the hospital’s general strategic mission remained constant while the speciﬁc objectives to pursue the mission changed. In this regard, our work sides with prior literature suggesting only small differences in outcomes (e.g., uncompensated care) across conversion types. To be sure, conversions are associated with speciﬁc strategic objectives. The speciﬁc objectives identiﬁed by our respondents included the formation of horizontal hospital systems and networks, the formation of vertical trading alliances with community-based physicians (e.g., to try to improve patient referrals), work restructuring and downsizing, centralized group purchasing of supplies, and consolidation of service lines.

220

LAWTON R. BURNS ET AL.

Second, our ﬁndings suggest that these strategic objectives either fail to be implemented aggressively or fail to improve organizational performance (i.e., the strategies selected are unsuccessful). Generic strategies such as horizontal and vertical integration have been pursued in healthcare over the past two decades with at best mixed success (cf. Burns & Pauly, 2001). If everyone is pursuing the same strategic objectives – as appears to be the case not only in our sample but also in the broader hospital population – then there is no basis for ﬁrms to develop competitive advantage based on the strategy alone. To achieve any competitive advantage, there must be different levels of successful implementation of the strategy. The responses from our informants suggest that most hospitals encountered problems in leveraging their system membership or improving their physician relationships. Across the board, expectations about a hospital’s ability to implement strategic objectives seemed to exceed actual, post-conversion realized strategies. In some cases the new owner may be unable, at least in the short term, to alter local market forces (e.g., through improved bargaining leverage) to the hospital’s advantage. Respondents frequently indicated that the degree of change expected or promised from the conversion exceeded what actually occurred, and they consistently remarked that the new owners did ‘‘less than expected.’’ This suggests that the hype of changed objectives may exceed their reality and that policy makers ought to base regulatory decisions on actual changes in behavior as opposed to a priori fears. Speciﬁcally, while many policy observers are concerned about changes in charity care and indigents’ access to care, our ﬁndings indicate that sharp post-conversion stafﬁng reductions may weaken a hospital’s patient safety mechanisms and represent a more serious problem. Finally, we consider how these ﬁndings can contribute to a theory on ownership transitions. Two theoretical approaches seem to have low utility here for explaining the observed similarity in performance across transition types. Agency theory implies stronger differences across ownership types than found here. Agency theory argues that the transition from public to private ownership should be accompanied by shifts from social to economic goals and vice-versa. In fact, we observed consistent changes across all transition types – a generic conversion effect. Agency theory also fails to specify the important intervening variables between ownership and organizational performance discussed in the latter half of this chapter. In a similar vein, the common pursuit of improved physician–hospital relationships across transition types suggests that political bargaining theory may be of limited utility. All hospitals face the political problems of relating

The Impact of Hospital Ownership Conversions

221

with their medical staffs. Transitions may involve improved staff relations – especially initially or in situations where a new owner has averted a closure. However, speciﬁc ownership changes do not appear to improve or exacerbate deeper relationship issues. Respondents indicated that the quality of these relationships was not affected by the speciﬁc owner but rather by extreme hospital ﬁscal distress, access to new capital, and the new owner’s ability to keep the hospital open. By contrast, the apparent similarity in performance outcomes across transition types suggests that strategic adaptation represents an important theoretical perspective. The hospital cases illustrate that similar strategic adaptations are undertaken across all transition types. They further suggest that intervening variables such as strategy content and process represent an important unit of observation for theorists, researchers, and policy makers. The cases illustrate how short-term strategies to improve hospital efﬁciency, such as downsizing, are pursued aggressively and may lead to serious policy concerns with respect to quality of care and patient safety. However, despite high expectations that a new owner will bring or develop important long-term competitive capabilities – such as an ability to win more managed care contracts under better terms or infuse large amounts of capital to upgrade facilities – few apparent (or apparently successful) efforts to do so were observed. Furthermore, aside from these strategic content issues, actually strategic processes changed dramatically, especially in situations where a free-standing hospital joined a multi-system chain, and vice-versa. The strategic adaptation perspective suggests that both strategy content and process may be important for improved performance. More importantly, since the strategies appear to be generic, transition efforts and efforts to understand strategic implementation may offer new insights regarding how these institutions change when they experience ownership conversions. To observe true, long-term performance effects, ownership changes may need to be followed by improved efforts to implement strategic objectives and build competitive advantage based on distinctive competences.

NOTES 1. Miller (1997) noted that while federal and most state laws require the proceeds from the sale of nonproﬁts to remain ‘‘in the charitable stream’’ after sale, there is little oversight of these sales. Policing of these laws resides with states’ ofﬁce of the attorney general. Joint ventures are particularly problematic, as they enable forproﬁt buyers to pay less than the fair market value of a hospital to gain control of it. These transactions are troublesome, as they commingle for-proﬁt and nonproﬁt

222

LAWTON R. BURNS ET AL.

assets. Furthermore, the arrangements are often unfavorable to the nonproﬁt hospital. The sales are often shrouded in a degree of secrecy that would not be possible in a transaction involving a publicly traded company. This secrecy reduces the degree to which the bidding process is competitive. Matters are made worse by the manner in which hospitals are valued; they are traditionally assessed using a trailing analysis of cash ﬂows. However, this analysis considers the hospital’s cash ﬂows while it is providing free care and other services that it is not obligated to provide after the conversion. Thus, the valuation does not reﬂect the performance expectations of the buyer. Furthermore, the economic value of control is often not factored into the sale price. Finally, in joint ventures, the cost of paying off debts typically comes out of the hospital’s portion of the proceeds and is not split evenly between both parties in the venture. This beneﬁts the for-proﬁt buyer at the expense of the community. Although covenants can be used to ensure that a hospital maintains certain services after conversion, they are not foolproof. It is impossible for a covenant to include provisions for a hospital to meet unknown future charitable needs, such as treatment for new diseases such as AIDS, or treatment using new, unproﬁtable technologies, such as many related to perinatal care. 2. The authors thank Peter Kralovec for providing these data. 3. In addition to the provision of uncompensated care, another access-related issue is the richness of services offered post-conversion. According to Wilkins and Jacobson (1998), when a board decides whether or not to sell a hospital to a for-proﬁt company, it must consider whether local residents will continue to have access to medical care in the event that the company decides to close the hospital. Furthermore, if the for-proﬁt is not building a local, vertically integrated system offering treatment for a wide range of conditions, the board must consider whether the local population will be forced to navigate a fragmented system. The board must also consider whether the for-proﬁt will be able to provide important but unproﬁtable services and high value services that offer little proﬁt. Finally, the board must consider whether the hospital will consider the health needs of the entire community, and not just the needs of the ill community members that pass through the hospital. 4. To be more precise, new owners seek improvement while previous owners have given up. We thank Brad Gray for his insight on this point. 5. No literature mentioned how conversions furthered innovation. 6. Although the explicit missions of these institutions were more likely to change, conversion contracts often mandated that the health system continue to provide access to certain ‘‘non-Catholic’’ services such as abortions or contraceptive care. 7. Home health services were the exception to this trend as many hospitals scaled back and eliminated such services due to changes in Medicare’s home health reimbursements.

ACKNOWLEDGMENT The authors thank Steve Shortell, Brad Gray, an anonymous reviewer, and the editors for their comments on prior drafts of this chapter. They also thank Jonathan Bluth, Robert DeGraaff, Katherine Greene, and Dawn

The Impact of Hospital Ownership Conversions

223

Levy, all of whom conducted the on-site and telephone interviews and drafted the individual case studies that served as the primary data sources for the latter part of this chapter.

REFERENCES Anderson, R. A., Allred, C., & Sloan, F. (2001). Effect of hospital conversion on organizational decision making and service coordination. Health Care Management Review, 28(2), 141–154. Arrow, K. (1963). Uncertainty and the welfare economics of medical care. American Economic Review, 53(5), 941–973. Bishop, M., & Kay, J. A. (1988). Does privatisation work? Lessons from the UK. Mimeo: London Business School. Blumenthal, D., & Weissman, J. S. (2000). Selling teaching hospitals to investor-owned hospital chains: Three case studies. Health Affairs, 19(2), 158–166. Burns, L. R., & Pauly, M. V. (2001). Integrated delivery networks (IDNs): A detour on the road to integrated healthcare? Health Affairs, 21(4), 128–143. Cain Brothers. (2002). What does the research say about hospital tax status conversions? A guide for health care executives and trustees. Strategies in Capital Finance, 38(7), 1–28. Chakravarthy, B. S. (1982). Adaptation: A promising metaphor for strategic management. Academy of Management Review, 7, 35–44. Chen, Y. (2002). The effect of hospital ownership choice on patient outcomes after treatment for acute myocardial infarction. Journal of Health Economics, 21, 901–922. Claxton, G., Feder, J., Schactman, D., & Altman, S. (1997). Public policy issues in nonproﬁt conversions: An overview. Health Affairs, 16(2), 9–27. Collins, S. R., Gray, B., & Hadley, J. (2001). The for-proﬁt conversion of nonproﬁt hospitals in the U.S. Health care system: Eight case studies. New York: The Commonwealth Fund. Coye, M. (1997). The sale of good Samaritan: A view from the trenches. Health Affairs, 16(2), 102–107. Cutler, D., & Horwitz, J. (2000). Converting hospitals from not-for-proﬁt to for-proﬁt status: Why and what effects? In: D. M. Cutler (Ed.), The changing hospital industry: Comparing not-for-proﬁt and for-proﬁt institutions (pp. 45–89). Chicago: The University of Chicago Press. Desai, K., Young, G., & Lukas, C. (1998). Hospital conversions from for-proﬁt to nonproﬁt status: The other side of the story. Medical Care Research and Review, 55(3), 298–308. Desai, K., Lukas, C. V., & Young, G. (2000). Public hospitals: Privatization and uncompensated care. Health Affairs, 19(2), 167–172. Dranove, D., & Lindrooth, R. C. (2003). Hospital consolidation and efﬁciency: Another look at the evidence. Journal of Health Economics, 22(6), 983–997. Farsi, M. (2004). Changes in hospital quality after conversion in ownership status. International Journal of Health Care Finance and Economics, 4(3), 211–230. Fishman, J. (1998). Checkpoints on the conversion highway: Some trouble spots in the conversion of nonproﬁt health care organizations to for-proﬁt status. Journal of Corporation Law, 23(4), 702–740.

224

LAWTON R. BURNS ET AL.

General Accounting Ofﬁce. (1997). Not-for-proﬁt hospitals: Conversion issues prompt increased state oversight. Washington, D.C.: GAO/HEHS-98-24. George, K. D. (1990). Public ownership versus privatisation. In: P. de Wolf (Ed.), Competition in Europe: Essays in honour of professor Henk W. de Jung. Amsterdam: Kluwer Academic Publishers. George, K. D., Joll, C., & Lynk, E. L. (1992). Industrial organization: Competition, growth and structural change (4th ed.). London: Routledge. Glaser, B. G., & Strauss, A. L. (1967). The discovery of grounded theory: Strategies for qualitative research. Chicago: Aldine Publishing Company. Grabowski, D., & Stevenson, D. (2008). Ownership conversions and nursing home performance. Health Services Research, 43(4), 1184–1202. Gray, B. (1991a). Accountability in nonproﬁt hospitals: How distinctive from investor-owned? In: B. H. Gray (Ed.), The proﬁt motive and patient care (pp. 61–89). Cambridge, MA: Harvard University Press. Gray, B. (1986). For-proﬁt enterprise in health care. Washington, DC: National Academy Press. Gray, B. H. (1997). Conversion of HMOs and hospitals: What’s at stake? Health Affairs, 16(2), 29–47. Gray, B. H. (1991b). The proﬁt motive and patient care. Cambridge: Harvard University Press. Hadley, J., Gray, B., & Collins, S. (2001). A statistical analysis of the impact of nonproﬁt hospital conversions on hospitals and communities, 1985–1996. May 2001 Report of The Commonwealth Fund, pp. 8–12. Health Care Advisory Board. (1991). Not-for-proﬁt hospitals that have been acquired by forproﬁt hospital systems. Document 212-018: Mergers and Acquisitions. Washington, DC: Advisory Board. Health Care Advisory Board. (2000). Avoiding ﬁnancial ﬂashpoints: Forseeing (and preventing) dramatic decline in hospital and health system fortunes. Washington, DC: Advisory Board. Hollis, S. (1997). Strategic and economic factors in the hospital conversion process. Health Affairs, 16(2), 131–143. Jennings, P. (2004). Strategic adaptation: A uni or multi dimensional concept? Strategic Change, 13, 1–10. Kimberly, J. R., & Zajac, E. J. (1985). Strategic adaptation in health care organizations: Implications for theory and research. Medical Care Review, 42(2), 267–302. King, J., & Avery, J. (1999). Evaluating the sale of a nonproﬁt health system to a for-proﬁt hospital management company: The legacy experience. Health Services Research, 34(1), 103–121. Kumaraswamy, A., Mudambi, R., Saranga, H., & Tripathy, A. (2008). Strategic adaptation to deregulation in the Indian auto components industry. Working Paper. Available at http:// www.industry.sloan.org/industrystudies/workingpapers. Legnini, M. W., Anthony, S., Wicks, E., Meyer, J., Rybowski, L., & Stepnick, L. (1999). Privatization of public hospitals. Report by the Kaiser Family Foundation Economic and Social Research Institute. Legnini, M., Rosenberg, L., Perry, M., & Robertson, N. (2000). Where does performance measurement go from here? Health Affairs, 19(3), 173–177. Mark, T., Paramore, C., Rodriguez, E., Good, C., Schur, C., Swalenstocker, E., Neuman, P., & Dorosh, E. (1997). The community impact of for-proﬁt hospital conversions: Results from ten case studies. The Project HOPE Center for Health Affairs.

The Impact of Hospital Ownership Conversions

225

Mark, T. (1999). Analysis of the rationale for, and consequences of, nonproﬁt and for-proﬁt ownership conversions. Health Services Research, 34(1), 83–101. McClellan, M., & Staiger, D. (2000). Comparing hospital quality at for-proﬁt and not-for-proﬁt hospitals. In: D. M. Cutler (Ed.), The changing hospital industry: Comparing not-forproﬁt and for-proﬁt institutions (pp. 93–112). Chicago: The University of Chicago Press. McKee, D. O., Varadarajan, P. R., & Pride, W. M. (1989). Strategic adaptability and ﬁrm performance: A market contingent perspective. Journal of Marketing, 53, 21–35. Miles, R., & Snow, C. (1978). Organizational strategy, structure, and process. New York: McGraw-Hill. Miller, L. (1997). The conversion game: High stakes, few rules. Health Affairs, 16(2), 112–117. Mintzberg, H. (1978). Patterns in strategy formation. Management Science, 24, 934–948. Needleman, J., Lamphere, J., & Chollet, D. (1999). Uncompensated care and hospital conversions in Florida. Health Affairs, 18(4), 125–133. Newhouse, J. P. (1970). Toward a theory of nonproﬁt institutions: An economic model of the hospital. American Economic Review, 60, 64–74. Pallarito, K. (1996). Hospital conversions raise thorny issues. Modern Healthcare (June 17), 104–106. Picone, G., Chou, S., & Sloan, F. (2002). Are for-proﬁt hospital conversions harmful to patients and to Medicare? The RAND Journal of Economics, 33(3), 507–523. Pauly, M. V. (1987). Lessons from health economics: Nonproﬁt ﬁrms in medical markets. American Economic Review, 77(2), 257–262. Ray, S. (2003). Strategic adaptation of ﬁrms during economic liberalisation: Emerging issues and a research agenda. International Journal of Management, 20(3), 271–281. Schendel, D., & Hofer, C. (1979). Strategic management: A new view of business policy and planning. Boston, MA: Little, Brown and Co. Schindehutt, M., & Morris, M. (2001). Understanding strategic adaptation in small ﬁrms. International Journal of Entrepreneurial Behavior and Research, 7(3), 84–107. Shen, Y. (2003). Changes in hospital performance after ownership conversions. Inquiry, 40(3), 217–234. Shortell, S. M. (1988). The evolution of hospital systems: Unfulﬁlled promises and self-fulﬁlling prophesies. Medical Care Review, 45(2), 177–214. Shortell, S. M., Morrison, E., Hughes, S., Friedman, B., & Vitek, J. (1987). Diversiﬁcation of health care services: The effects of ownership, environment, and strategy. In: R. Schefﬂer & L. Rossiter (Eds), Advances in Health Economics and Health Services Research (7, pp. 3–40). Greenwich, CT: JAI Press. Shortell, S. M., Morrison, E. M., & Friedman, B. (1990). Strategic choices for America’s hospitals: Managing change in turbulent times. San Francisco: Jossey-Bass. Sloan, F. (2002). Hospital ownership conversions: Deﬁning the appropriate public oversight role. Frontiers in Health Policy Research, 5(1), 123–166. Sloan, F., Conover, C., & Ostermann, J. (2003a). Rates of return from hospital conversions. Health Care Management Review, 28(2), 107–118. Sloan, F., Ostermann, J., & Conover, C. (2003b). Antecedents of hospital ownership conversions, mergers, and closures. Inquiry, 40(1), 39–56. Sloan, F. A. (2000). Not-for-proﬁt ownership and hospital behavior. In: A. J. Cuyler & J. P. Newhouse (Eds), Handbook of Health Economics (1, pp. 1141–1174). Elsevier Science.

226

LAWTON R. BURNS ET AL.

Sloan, F. A., Taylor, D., & Conover, C. (2000). Hospital conversions: Is the purchase price too low? In: D. M. Cutler (Ed.), The changing hospital industry: Comparing not-for-proﬁt and for-proﬁt institutions (pp. 13–44). Chicago: University of Chicago Press. Sloan, F. A., Picone, G., Taylor, D., Jr., & Chou, S. (2001). Hospital ownership and cost and quality of care: Is there a dime’s worth of difference? Journal of Health Economics, 20(1), 1–21. Sloan, F. A., & Steinwald, B. (1980). Insurance, regulation, and hospital costs. Lexington: Lexington Books. Thorpe, K. E., Florence, C., & Seiber, E. (2000). Hospital conversions, margins, and the provision of uncompensated care. Health Affairs, 19(6), 187–194. Townsend, J. (1983). When investor-owned corporations buy hospitals. In: B. H. Gray (Ed.), The new health care for proﬁt: Doctors in a competitive environment (pp. 51–72). Washington, D.C.: National Academy Press. Trinh, H. Q., & Begun, J. W. (1999). Strategic adaptation of U.S. rural hospitals during an era of limited ﬁnancial resources: A longitudinal study, 1983 to 1993. Health Care Management Science, 2, 43–52. Venkatraman, N., & Prescott, J. E. (1990). Environment-strategy coalignment: An empirical test of its performance implications. Strategic Management Journal, 11, 1–23. Villalonga, B. (2000). Privatization and efﬁciency: Differentiating ownership effects from political, organizational, and dynamic effects. Journal of Economic Behavior & Organization, 42, 43–74. Weisbrod, B. A. (1989). Rewarding performance that is hard to measure: The role of nonproﬁt organizations. Science, 244(4904), 541–546. Wilkins, A., & Jacobson, P. (1998). Fiduciary responsibilities in nonproﬁt health care conversions. Health Care Management Review, 23(1), 77–90. Yarrow, G. (1989). Does ownership matter? In: C. Veljanovski (Ed.), Privatisation and competition: A market prospectus. London: Institute of Economic Affairs. Yoder, S. G. (1986). Economic theories of for-proﬁt and not-for-proﬁt organizations. In: B. Gray (Ed.), For-proﬁt enterprise in health care (pp. 19–25). Washington, DC: National Academy Press. Young, G. J., & Desai, K. (1999). Nonproﬁt hospital conversions and community beneﬁts: New evidence from three states. Health Affairs, 18(5), 146–155. Zinn, J. S., Mor, V., Feng, Z., & Intrator, O. (2007). Doing better to do good: The impact of strategic adaptation on nursing home performance. Health Services Research, 42(3), 1200–1218.

The Impact of Hospital Ownership Conversions

227

APPENDIX A. CASE STUDY INTERVIEW PROTOCOL Intended Strategy: Motivations for Conversion Why did your hospital change ownership? Why was (new ownership status) selected? Realized Changes in Hospital Strategy What changes in hospital strategy resulted from the ownership conversion? What new strategies are being pursued after the conversion? Were any old strategies abandoned after the conversion? Could the strategy change have been pursued without conversion? (Open-ended strategy content questions) 1. Did the conversion result in: (Probe for other changes in hospital strategy) A. Changes

in mission, goals, or other broad philosophical shifts? reductions, eliminations, or regionalizations of hospital departments/services? C. Stafﬁng and labor, downskilling D. Changes in continuum-of-care development efforts? E. Changes in strategic alliance formation? (e.g., with physicians? with other hospitals?) F. Major new cost-containment or total quality management initiatives? G. Changes in contracting or pricing? H. Changes in provision of charity/uncompensated care? Other community services? I. Changes in physical plant or capital budgeting? (has there been an infusion of capital from the new owner?) J. Changes in organizational culture? K. Changes in short-term versus long-term time horizon for achieving results? L. Changes in reliance on non-operating revenue (e.g., philanthropy)? M. Changes in how the hospital competes or efforts to be different from competitors? B. Expansions,

Changes in Strategy Process What changes in strategic planning and decision-making processes resulted from the conversion? Have there been changes in who makes decisions and how the decisions are made?

228

LAWTON R. BURNS ET AL.

Did the conversion result in: (Probe for other changes in strategy process) A. Changes

in planning and decision-making regarding managed care contracting? relationships with physicians? B. Changes in planning, implementing, and controlling operating budgets? C. Changes in capital planning, decision-making, and allocation? D. Any change in use of consultants? E. Any change in strategic decision-making cycle time? F. Changes in the balance of power among the hospital CEO, medical staff, local board, and corporate senior management? Any change in how decisions are made? Any change in level of decision making (e.g., corporate versus local board)? G. Any change in board composition? Any change in community representation or community input? H. Changes in planning and decision-making for contracting/purchasing/ standardizing medical supplies and pharmaceuticals? I. Changes in the role of information management technology? Conversion Problems What types of post-conversion integration problems did you experience (e.g., operational or cultural integration, community reaction, media attention)? Which of these were unanticipated or under-estimated? How did you address them? Have they been resolved? Were there any signiﬁcant state regulatory/legislative issues surrounding the conversion (e.g., impact of state attorney general review)? How did the market react to the conversion? Wrap-Up Questions If you had to do it over again, would you do the same thing? What would you do differently? What do you miss from pre-conversion times? Any regrets concerning the conversion? What did the conversion allow you to do that you would not have otherwise been able to accomplish?

229

The Impact of Hospital Ownership Conversions

APPENDIX B. HOSPITAL CASES BY GOVERNANCE TRANSITION Governance Transition Nonproﬁt to for-proﬁt

Case

A private, nonproﬁt hospital (150þ beds) in Tennessee was sold to a major hospital chain and became a for-proﬁt entity

A private, nonproﬁt hospital (450þ beds) in Florida owned and operated by a nonproﬁt health system was sold to a for-proﬁt health system

A nonproﬁt health system that owned a private nonproﬁt hospital (50þ beds) in

California hired a major for-proﬁt hospital chain as a contract manager. When the contract lease expired, the nonproﬁt system sold the hospital to a private, for-proﬁt hospital system A private free standing, nonproﬁt hospital (200þ beds) in South Carolina was acquired by a for-proﬁt hospital chain and converted to a for-proﬁt hospital A private, nonproﬁt hospital (100þ beds) in South Carolina was sold to a private, for-proﬁt hospital chain A private, stand-alone, nonproﬁt institution (200 beds) in North Carolina joined a for-proﬁt hospital system A private, stand-alone, nonproﬁt hospital (450þ beds) in Tennessee was sold to a for-proﬁt health care company which then merged with a larger for-proﬁt hospital chain A nonproﬁt, hospital (100 beds) in South Carolina was sold to a for-proﬁt hospital management company

Public to for-proﬁt

A public, city-owned and operated facility (300þ beds) in Texas was sold to a

Public to nonproﬁt

A city-owned hospital (250þ beds) in Texas was transferred to a nonproﬁt

for-proﬁt health system hospital chain by entering into a long-term lease agreement

A public county hospital (100þ beds) in North Carolina was sold to a regional nonproﬁt hospital system For-proﬁt to nonproﬁt

A private, for-proﬁt hospital (50þ beds) in Louisiana was owned and operated

by a national hospital chain. This chain was forced to divest the hospital when it obtained a competitor. The divested hospital became a private, stand-alone, nonproﬁt facility A private, for-proﬁt hospital (50 beds) in Tennessee was owned by a number of for-proﬁt companies prior to its sale and conversion to a private, nonproﬁt hospital A county-owned hospital (50þ beds) in North Carolina managed by a for-proﬁt hospital system was transferred to a nonproﬁt management company that was allowed to purchase ownership and manage the hospital A hospital (100þ beds) in Florida owned and operated by a for-proﬁt system was sold to a nonproﬁt hospital system A hospital (50þ beds) in Tennessee that was built, owned and operated by a forproﬁt hospital company was divested by the corporation as part of an anti-trust ruling. The divested facility became part of a private, nonproﬁt hospital system