Physical Therapy- Journal of the APTA- Feburary 2009, Volume 89, Issue 2

N O R T H C O A S T M E D I C A Put the Pain on Ice. PhysioIce. i Enhance modality treatments with Ph...

Author: APTA

45 downloads 797 Views 3MB 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

N

O

R

T

H

C

O

A

S

T

M

E

D

I

C

A

Put the Pain on Ice.

PhysioIce. i

Enhance modality treatments with PhysioICEi analgesic gel. Cool, soothing PhysioICEipenetrates deeply to help relax muscles and increase blood flow for long-lasting relief from aches and pain associated with arthritis, tendinitis, strains and sprains. Non-greasy and fast-drying, PhysioICEi will not stain skin or clothes, and can be applied before, during and after treatment. For quick, effective relief that lasts, Put the Pain on Ice. PhysioICEi. TM

For a free sample of Physio ICEi, call toll-free, 800-821-9319 and ask for NC70080.

North Coast

North Coast Medical • 18305 Sutter Boulevard • Morgan Hill, CA 95037 • Toll-Free: 800-821-9319 Toll-Free Fax: 877-213-9300 • Local/Int’l: 408-776-5000 • www.ncmedical.com See Reader Service No 3 at www.apta.org/adinfo

L

0WFS:FBST5SVTUFE2VBMJUZ&YQFSJFODF

1"-µT%TDBOOFSTPGUXBSFJODSFBTFT ZPVSSFGFSSBMTBOEQSBDUJDFSFWFOVF It’s your prescription foot orthotic solution 1SBDUJUJPOFSTIBWFDPNNFOUFE ²1BUJFOUTBSFTPBNB[FEBUUIFTPGUXBSF UIFZJNNFEJBUFMZ UFMMUIFJSGSJFOET³ ²#FDBVTFJUµTTPFBTZ XFIBWFQSFTDSJCFENPSFPSUIPUJDTJO UXPNPOUITUIBOXFOPSNBMMZEJEJOBOFOUJSFZFBS³ ²1BUJFOUTBSFFYDJUFEBCPVUUIFUFDIOPMPHZBOEMPWFUIF UVSOBSPVOEUJNF³

See more patients... ...see more proﬁts! Call immediately to for your Xtremityscan! Visit our website at www.PALHealth.com for more information. See Reader Service No 5 at www.apta.org/adinfo

Physical Therapy Journal of the American Physical Therapy Association

Editorial Office

Editor in Chief

Managing Editor / Associate Director of Publications: Jan P Reynolds, [email protected]

Rebecca L Craik, PT, PhD, FAPTA, Philadelphia, PA [email protected]

PTJ Online Editor / Assistant Managing Editor: Steven Glaros

Deputy Editor in Chief

Associate Editor: Stephen Brooks, ELS Production Manager: Liz Haberkorn Manuscripts Coordinator: Karen Darley Permissions / Reprint Coordinator: Michele Tillson Advertising Manager: Julie Hilgenberg Director of Publications: Lois Douthitt

APTA Executive Staff Senior Vice President for Communications: Felicity Feather Clancy Chief Financial Officer: Rob Batarla Chief Executive Officer: John D Barnes

Advertising Sales Ad Marketing Group, Inc 2200 Wilson Blvd, Suite 102-333 Arlington, VA 22201 703/243-9046, ext 102 President / Advertising Account Manager: Jane Dees Richardson

Board of Directors President: R Scott Ward, PT, PhD Vice President: Randy Roesch, PT, MBA, DPT Secretary: Babette S Sanders, PT, MS Treasurer: Connie D Hauser, PT, DPT, ATC Speaker of the House: Shawne E Soper, PT, DPT, MBA Vice Speaker of the House: Laurita M Hack, PT, DPT, MBA, PhD, FAPTA Directors: William D Bandy, PT, PhD, SCS, ATC; Sharon L Dunn, PT, PhD, OCS; Kevin L Hulsey, PT, DPT, MA; Dianne V Jewell, PT, DPT, PhD, CCS, FAACVPR; Aimee B Klein, PT, DPT, MS, OCS; Stephen CF McDavitt, PT, DPT, MS, FAAOMPT; Paul A Rockar Jr, PT, DPT, MS; Lisa K Saladin, PT, PhD; John G Wallace Jr, PT, MS, OCS

Daniel L Riddle, PT, PhD, FAPTA, Richmond, VA

Editor in Chief Emeritus Jules M Rothstein, PT, PhD, FAPTA (1947–2005)

Steering Committee Anthony Delitto, PT, PhD, FAPTA (Chair), Pittsburgh, PA; J Haxby Abbott, PhD, MScPT, DipGrad, FNZCP, Dunedin, New Zealand; Joanell Bohmert, PT, MS, Mahtomedi, MN; Alan M Jette, PT, PhD, FAPTA, Boston, MA; Charles Magistro, PT, FAPTA, Claremont, CA; Ruth B Purtilo, PT, PhD, FAPTA, Boston, MA; Julie Whitman, PT, DSc, OCS, Westminster, CO

Editorial Board Andrea Behrman, PT, PhD, Melrose, FL; Rachelle Buchbinder, MBBS(Hons), MSc, PhD, FRACP, Malvern, Victoria, Australia; W Todd Cade, PT, PhD, St Louis, MO; John Childs, PT, PhD, Schertz, TX; Charles Ciccone, PT, PhD, FAPTA (Continuing Education), Ithaca, NY; Joshua Cleland, PT, DPT, PhD, OCS, FAAOMPT (The Bottom Line), Concord, NH; Janice J Eng, PT/OT, PhD, Vancouver, BC, Canada; G Kelley Fitzgerald, PT, PhD, OCS, Pittsburgh, PA; James C (Cole) Galloway, PT, PhD, Newark, DE; Kathleen Gill-Body, PT, DPT, NCS, Boston, MA; Paul JM Helders, PT, PhD, PCS, Utrecht, The Netherlands; Maura D Iversen, PT, MPH, ScD, Boston, MA; Diane U Jette, PT, DSc, Burlington, VT; Gregory Karst, PT, PhD, Omaha, NE; Christopher Maher, PT, PhD, Lidcombe, NSW, Australia; Christopher J Main, PhD, FBPsS, Keele, United Kingdom; Kathleen Kline Mangione, PT, PhD, GCS, Philadelphia, PA; Patricia Ohtake, PT, PhD, Buffalo, NY; Carolynn Patten, PT, PhD, Gainesville, FL; Christopher Powers, PT, PhD, Los Angeles, CA; Linda Resnik, PT, PhD, OCS, Providence, RI; Val Robertson, PT, PhD, Copacabana, NSW, Australia; Patty Solomon, PT, PhD, Hamilton, Ont, Canada

Statistical Consultants Steven E Hanna, PhD, Hamilton, Ont, Canada; John E Hewett, PhD, Columbia, MO; Hang Lee, PhD, Boston, MA; Samuel Wu, PhD, Gainesville, FL

The Bottom Line Committee Joanell Bohmert, PT, MS; Lara Boyd, PT, PhD; James Cavanaugh IV, PT, PhD, NCS; Todd Davenport, PT, DPT, OCS; Ann Dennison, PT, DPT, OCS; William Egan, PT, DPT, OCS; Helen Host, PT, PhD; Evan Johnson, PT, DPT, MS, OCS, MTC; M Kathleen Kelly, PT, PhD; Catherine Lang, PT, PhD; Tara Jo Manal, PT, MPT, OCS, SCS; Kristin Parlman, PT, DPT, NCS; Susan Perry, PT, DPT, NCS; Maj Nicole H Raney, PT, DSc, OCS, FAAOMPT; Rick Ritter, PT; Eric Robertson, PT, DPT; Kathleen Rockefeller, PT, MPH, ScD; Michael Ross, PT, DHS, OCS; Patty Scheets, PT, DPT, NCS; Katherine Sullivan, PT, PhD; Mary Thigpen, PT, PhD; Jamie Tomlinson, PT, MS; Brian Tovin, DPT, MMSc, SCS, ATC, FAAOMPT; Nancy White, PT, MS, OCS; Julie Whitman, PT, DSc, OCS

112 ■ Physical Therapy Volume 89 Number 2

February 2009

Subscriptions

Physical Therapy (PTJ) (ISSN 00319023) is published monthly by the American Physical Therapy Association (APTA), 1111 North Fairfax Street, Alexandria, VA 22314-1488, at an annual subscription rate of $15 for members, included in dues. Nonmember rates are as follows: Individual (inside USA)— $99; individual (outside USA)—$119 surface mail, $179 air mail. Institutional (inside USA)—$129; institutional (outside USA)—$149 surface mail, $209 air mail. Periodical postage is paid at Alexandria, VA, and at additional mailing offices. Postmaster: Send address changes to Physical Therapy, 1111 North Fairfax Street, Alexandria, VA 22314-1488. Single copies: $15 USA, $15 outside USA; with the exception of January 2001: $50 USA, $70 outside USA. All orders payable in US currency. No replacements for nonreceipt after a 3-month period has elapsed. Canada Post International Publications Mail Product Sales Agreement No. 0055832.

Members and Subscribers Send changes of address to: APTA, Attn: Membership Dept, 1111 North Fairfax St, Alexandria, VA 22314-1488. Subscription inquiries: 703/684-2782, ext 3124. PTJ is available in a special format for readers who are visually impaired. For information, contact APTA’s Membership Department at 703/684-2782, ext 3124.

Mission Statement

Founded in 1921, Physical Therapy (PTJ) is the official publication of the American Physical Therapy Association (APTA) and is an international, scholarly, peer-reviewed journal. PTJ serves APTA members, other health care professionals, and patients/clients by (1) documenting basic and applied knowledge related to physical therapy, (2) providing evidence to guide clinical decision making, and (3) publishing a variety of research that is relevant to the field and diverse opinions that are based in scholarly arguments. PTJ, like the profession it serves, strives to enhance the health and well-being of all members of society.

February 2009

Readers are invited to submit manuscripts to PTJ. PTJ’s content—including editorials, commentaries, letters, and book reviews—represents the opinions of the authors and should not be attributed to PTJ or its Editorial Board. Content does not reflect the official policy of APTA or the institution with which the author is affiliated, unless expressly stated.

Technologies. Ingenta provides online document delivery for articles published since September 1988. Full-text articles are available for free at www.ptjournal.org 12 months after the publication date. Full text also is provided through DataStar, Dialog, EBSCOHost Academic Search, Factiva, InfoTrac, ProFound, and ProQuest.

Reprints

PTJ Online at www.ptjournal.org PTJ’s Web site posts articles ahead of print and rapid reader responses to articles. Articles, letters to the editor, and editorials are available in full text starting with Volume 79 (1999) and in searchable PDF format starting with Volume 60 (1980). Entire issues are available online beginning with Volume 86 (2006) and include additional data, video clips, and podcasts.

Indexing and Document Delivery

Readers should direct requests for reprints to the corresponding author of the article. Students and other academic customers may receive permission to reprint copyrighted material from this publication by contacting the Copyright Clearance Center Inc, 222 Rosewood Dr, Danvers, MA 01923. Authors who want reprints should contact Nancy Boyles, Cadmus Reprints, at 800/4079190, or [email protected]. Nonacademic institutions needing reprint permission information should go to www.ptjournal.org/misc/terms.dtl.

Advertising

PTJ is indexed and/or abstracted by Abridged Index Medicus, Abstracts of Health Care Management Studies, AgeLine, AMED (Allied and Complementary Medicine Database), Bibliography of Developmental Medicine and Child Neurology, Current Contents, Cumulative Index to Nursing and Allied Health Literature (CINAHL), EMBASE/Exerpta Medica, Exceptional Child Education Resources, Focus on: Sports Science and Medicine, General Science Index (GSI), Health Index, Hospital and Health Administration Index, Index Medicus, Inpharma Weekly, International Nursing Index, ISR, Medical & Surgical Dermatology, MEDLINE, Neuroscience Citation Index, Personal Alert: Automatic Subject Citation Alert (ASCA), Pharmacoeconomics and Outcomes News, Physical Education Index, Reactions Weekly, RECAL Bibliographic Database, Science Citation Index (SCI), Social Sciences Citation Index (SSCI), and SportsS. Article abstracts are available online at www.ptjournal.org (1980 through present) and via DataStar, Dialog, FirstSearch, Information Access, Ovid

Advertisements are accepted by PTJ when they conform to the ethical standards of the American Physical Therapy Association. PTJ does not verify the accuracy of claims made in advertisements, and acceptance does not imply endorsement by PTJ or the Association. Acceptance of advertisements for professional development courses addressing advanced-level competencies in clinical specialty areas does not imply review or endorsement by the American Board of Physical Therapy Specialties.

Statement of Nondiscrimination APTA prohibits preferential or adverse discrimination on the basis of race, creed, color, gender, age, national or ethnic origin, sexual orientation, disability, or health status in all areas including, but not limited to, its qualifications for membership, rights of members, policies, programs, activities, and employment practices. APTA is committed to promoting cultural diversity throughout the profession.

Volume 89 Number 2 Physical Therapy ■ 113

Muscle and Tendon Changes in Infants Born Preterm

Invited Commentary Grant-Beuttler et al1 present an interesting report describing the measurement of gastrocnemius-soleus muscle tendon unit (MTU) lengths from term to 12 weeks of age in infants born preterm and infants born full term. This work joins that of other authors2 in suggesting that infants born preterm have a different developmental trajectory than infants born full term and they have identifiable and measurable impairments during early infancy. This study examined the differences in muscle-tendon tautness, length, and stretch over time. The authors note that a limitation of their study was the lack of measurements of knee flexion angle of newborn infants, both full term and preterm, to demonstrate extension limitations of the knee. Because the gastrocnemius is a 2-joint muscle, any degree of knee flexion may be important. It is very difficult to stretch the gastrocnemius muscle at the ankle with any degree of knee flexion. In this population, it is unlikely that any measurements were taken with the knee in full extension. As such, the measurements taken in this study were largely of the soleus muscle. It is intriguing that infants born preterm show biases toward plantar flexion during passive range of motion, within kicking patterns, and during early gait. The direction of abnormal movement patterns at the ankle during kicking and stepping is constant with the direction of musculoskeletal abnormalities at the ankle in this article. However, the gastrocnemius muscle is a large contributor to plantar flexion. Toe walking may be habitual, neurological, or orthopedic, as the authors suggest, but it seems unlikely that the soleus muscle is a primary contributor.

February 2009

Jill C Heathcock

Another interesting question is how to design the most effective intervention given the MTU differences at the ankle of preterm infants. The authors suggest splinting the infant’s ankle as early as possible and provide 2 reasons that splinting may be effective in re-creating the MTU length in full-term infants: (1) studies in animal models have demonstrated a change in the number of sarcomeres with immobilization, and (2) infants born preterm are missing the constrained uterine space during a full-term pregnancy that promotes prolonged dorsiflexion. First, splinting and positioning of the ankle are primarily passive interventions. As with any kind of passive immobilization, the result will likely be a weaker muscle with a possible change in length of the MTU that may (or may not) result in improvements in functional skills. In addition, splinting and positioning of the ankle may affect the soleus muscle, but infants— especially preterm infants— kick a lot. As a knee flexor, the gastrocnemius muscle will likely play a large role in leg movements. Therefore, it is unlikely that there will be any morphological changes of the gastrocnemius muscle during immobilization of the ankle.

variety of sensory stimulation from the uterus that is dynamic and adaptive. What’s more, half of the infants born preterm who participated in this study were from multiple births. Twin and triplet pregnancies result in confined quarters earlier in gestation. It is not clear whether these infants from multiple births received more dorsiflexion stretch earlier in gestation and how that might affect their MTU length from birth to 12 weeks of age. Confinement and prolonged stretch are likely not the end of the story. Because the developmental trajectory of infants born preterm is likely different from that of full-term infants, re-creating the MTU length in full-term infants may not be the right goal. Our challenge may be to find ways to constrain the movement without taking away the ability to learn how to move. JC Heathcock, PT, PhD, is Assistant Professor, Division of Physical Therapy, School of Allied Medical Professions, Ohio State University, Atwell Hall, 453 W 10th Ave, Columbus, OH 43210 (USA). Address all correspondence to Dr Heathcock at: [email protected]. Dr Heathcock acknowledges Lynn SnyderMackler, PT, ScD, SCS, FAPTA, for her comments on the commentary. DOI: 10.2522/ptj.20070306.ic

Second, in the last 4 weeks of gestation, full-term infants may be confined, but their ankles are in no way immobilized. In addition to a prolonged stretch, the uterine environment provides a wealth of sensory information to the infant and to the legs. The experience these infants born preterm are “missing” is more active and reciprocal than just confinement. The infants’ own body size changes rapidly (also causing a stretch to musculature), and all movements (big or not) result in a

References 1 Grant-Beuttler M, Palisano RJ, Miller DP, et al. Gastrocnemius-soleus muscle tendon unit changes in infants born preterm over the first 12 weeks of adjusted age. Phys Ther. 2009;89:136 –148. 2 van Haastert IC, de Vries LS, Helders PJ, Jongmans MJ. Early gross motor development of preterm infants according to the Alberta Infant Motor Scale. J Pediatr. 2006; 149:617– 622

Volume 89

Number 2

Physical Therapy f

e1

Muscle and Tendon Changes in Infants Born Preterm Marybeth Grant-Beuttler, Robert J Palisano, Debra P Miller, Barbara Reddien Wagner, Carolyn B Heriza, Patricia A Shewokis

Author Response We thank Heathcock for taking the time to write her commentary1 regarding our study.2 Her comments are interesting, and our subsequent investigation into these ideas has led us to uncover some enlightening evidence in relation to the topics raised in the commentary. We will attempt to review each of the topics raised and the evidence associated with these topics.

Measurement of Gastrocnemius-Soleus Muscle Tendon Unit Heathcock1 suggests that “the measurements taken in this study were largely of the soleus muscle” secondary to knee contractures. There are 3 important issues to consider when contemplating this perspective. First, we measured infants born full term and infants born preterm over 3 different ages. In the infant born full term at the newborn measurement, knee extension is limited. During data collection, we were careful to ensure that we extended the knee as much as possible, but in the newborn infant born full term, the knee has a small flexion contracture that limits full extension. Although this muscle may not have been fully lengthened, it was lengthened as much as possible. This limitation to extension was not an issue at 6 or 12 weeks of age in the infants born full term or in the infants born preterm at any age. Second, we do not know which structures are limiting knee extension. If the gastrocnemius muscle was the structure limiting knee extension, then by extending the knee fully, we lengthened the muscle. Based on research by Brown and Swenson,3 the difference in ankle dorsiflexion in infants born full term at newborn age with the knee flexed e2

f

Physical Therapy

Volume 89

to 90 degrees versus fully extended was 3 degrees in boys and 2 degrees in girls. This difference in ankle dorsiflexion with full extension and 90 degrees of knee flexion is small, and changes observed at the ankle with the small knee flexion we had present in our participants should have been even smaller than that found with 90 degrees of knee flexion and full extension, possibly too small to reliably detect via a goniometric measurement. Third, we were measuring the gastrocnemius-soleus muscle tendon unit. During measurement, palpation occurred at the Achilles tendon, which is composed of the tendinous attachment of both muscles. Distinction between the gastrocnemius and soleus muscles cannot be made in the Achilles tendon, and shortening in either muscle would likely result in changes of the Achilles tendon. Measuring the gastrocnemius-soleus muscle tendon unit gives us information about available ankle motion for function. Clinically, this is a reasonable and accepted measurement. Matjacˇic´ et al4 used a device to artificially create shortening in the gastrocnemius muscle, in the soleus muscle, and in both muscles. The results of their study suggest that during function, plantar flexion at the ankle would be limited by shortening either the gastrocnemius muscle or the soleus muscle or by shortening both muscles, and all 3 of these scenarios would result in toe-walking. Although the proximal insertions of the gastrocnemius and soleus muscles produce different effects on trunk support, acceleration and deceleration of the trunk and leg during gait,5 there is significant evidence to suggest that both muscles

Number 2

are implicated in toe-walking.6 –11 Both muscles have been found via electromyographic (EMG) analysis to activate earlier in the gait cycle and remain active for more of the gait cycle during toe-walking.6 –11 During computer-simulated EMG recreations of 10 adults during toewalking, increased activity of the soleus and gastrocnemius muscles was observed, with the soleus muscle contributing the most to both the vertical and horizontal ground reaction forces in early stance.10,11 The notion that the soleus muscle does not contribute to toe-walking is not supported by the evidence.

Splinting Heathcock1 accurately reports that we suggest splinting. Our suggestion of a splint “that limits plantar flexion and allows active dorsiflexion”2(p146) would allow the infant to kick and move in all directions, with the exception of limiting plantar flexion. The splint we suggested would not limit kicking, cause immobilization, or confine the infant’s kicking in any direction except extreme plantar flexion at the ankle. We would like to restate our ideas on timing of the splinting intervention. The most important reason we suggest early intervention is based on animal studies that demonstrated differences in a muscle’s response to lengthening in young animals.12,13 Muscle tendon unit measurements in the full-term infant mimic the response documented in very young infants.2 If intervention was delayed, the outcome might not be the same, whereas having intervention occur prior to term age matches the timing of the lengthening that occurs in the infant born full term. Physical therapy intervention to lengthen the gastrocnemius-soleus muscle tendon February 2009

Muscle and Tendon Changes in Infants Born Preterm unit immediately following preterm birth also would be more likely to alter movement patterns that may be developing during this period and encourage patterns similar to those of the infant born full term by increasing the similarities of the biomechanical system. Lastly, for practical reasons, this is when the infants may be available for intervention when they are in the neonatal intensive care unit (NICU). Of course, the infant born preterm is, following birth, external to the uterine environment, and we cannot at this point exactly mimic the uterine environment in the NICU. We can try, however, to decrease influences of this external uterine environment by positioning and providing postural support. Many NICU therapists position infants in flexed positions to physically improve stability and promote self-regulation. These interventions have been the topic of many research articles over the years. Clinically, when we find postural changes in the infant born preterm, we traditionally have attempted to decrease the environmental influences that may result in these postural changes. Our interventions might not perfectly recreate the uterine environment, but they are designed to improve NICU outcomes. In addition to positioning changes, Heathcock1 suggests that “infants— especially preterm infants— kick a lot.” There is very limited evidence documenting the frequency of kicking in infants born preterm prior to term age. Droit et al14 examined kicking in infants at low risk for developmental issues who were born preterm prior to term age. They reported (1) a mean of 1.33 bouts of kicking per hour, with an average of 6.5 kicks per bout, at 31 to 35 weeks gestation and (2) a mean of 1.83 bouts of kicking per hour, with an average of 7.4 kicks per bout, at 37 February 2009

to 39 weeks gestation. Kicking frequency documented at and following term age in the infant born preterm increases to a mean of 3.7 to 28.6 kicks per minute (versus hours as measured prior to term) from term to 4 months after term age.15–18 None of the research examined kicking over days or weeks; therefore, we have no evidence concerning what an average kicking rate would be over a day or week. Research suggests that kicking frequency changes with arousal level, and when infants are fussing and crying, the frequency of kicking increases.19 Based on the definition of infant state 1, deep sleep, kicking would not be observed in this state. If the infants born preterm are physiologically stressed, kicking may be less than when the physiological system is stable. Despite the frequency of kicking, the splint we suggested would not limit kicking and would support the infants learning a lowerextremity pattern without the extremes in plantar flexion, similar to the movement of infants born full term. Heathcock1 suggests that splinting the ankle would not affect the length of the gastrocnemius muscle because it is a 2-joint muscle. We would like to raise 3 issues regarding this statement. First, changes in ankle dorsiflexion have been documented in children after short leg casting (below the knee).8,20 –23 Some studies documented functional improvements that accompanied the increases in dorsiflexion.8,22 Second, movement in utero, during the last few weeks of gestation, would most likely not include full knee extension, because the infant would not have enough room to fully extend the knee. Splinting just the ankle would mimic the plantar-flexor lengthening in utero found in infants born full term. Third, infants born preterm are generally more extended. If the infant born preterm

has a splint on, which limits plantar flexion, and the knee is generally in a position of more extension, then both ends of the gastrocnemius muscle would be stretched and lengthening of this muscle would be likely. Based on these points, we hypothesize that limiting plantar flexion at the ankle may very well be an effective intervention.

Multiple Births Versus Singleton Birth We agree with Heathcock1 that there may be a difference in uterine movement between single and multiple gestations. Multiple gestations may decrease available space for infant movement. Eleven of the 22 infants included in our preterm group were from multiple-gestation pregnancies. It is possible that the infants from multiple births may encounter more cramping earlier in gestation and could have experienced gastrocnemius-soleus muscle tendon unit lengthening as a result of crowding prior to birth. This would have resulted in the preterm, multiplegestation infants looking more like the full-term group. Significant differences between our 2 groups were found when the multiple-gestation preterm infants were included in the preterm group. If you examine Figure 3 of our article2, there is a distinct difference between the preterm and full-term groups, which is supported by our statistical analysis. Two additional points regarding inclusion of infants from multiplegestation births in research are worth considering. First, there has been a steady increase in multiple births,24 and the addition of infants from multiple births represented the population of the NICU where we recruited infants. If ankle range and movement is an issue in these children, we as physical therapists should be identifying and addressing it. Second, historical precedence can be found to include multiple-gesta-

Volume 89

Number 2

Physical Therapy f

e3

Muscle and Tendon Changes in Infants Born Preterm tion preterm infants in preterm groups. Heathcock et al18 included 9 multiple-gestation infants out of 10 preterm infants when investigating differences between full-term and preterm groups. In light of some of the concerns Heathcock1 raises, we would like to review briefly a study by Griffin et al,8 who examined children who did and did not idiopathically toe-walk and recorded muscle activations via EMG during gait. The children who toe-walked were placed in short leg casts for a minimum of 6 weeks. Immediately following intervention, these children demonstrated EMG gastrocnemius and soleus muscle activations similar to those of the children who did not toe-walk. The children also demonstrated positive changes in their ability to heel-strike, which continued. It appears, based on this study, that a change in muscle length has potential to alter use of the muscle during function and, in this case, ankle movement during gait. Our results are consistent with the findings of numerous studies on infant kicking and have the potential to shed light on ankle movement during gait and possibly toe-walking. In addition, this study confirms our previous findings. Although there is still much work to do in this area, we hope that our investigation provides a positive step in the process of understanding the biomechanics of the infant’s moving system.

References 1 Heathcock JC. Invited commentary on “Gastrocnemius-soleus muscle tendon unit changes over the first 12 weeks of adjusted age in infants born preterm. Phys Ther. 2009;89:e1. 2 Grant-Beuttler M, Palisano RJ, Miller DP, et al. Gastrocnemius-soleus muscle tendon unit changes over the first 12 weeks of adjusted age in infants born preterm. Phys Ther. 2009;89:136 –148. 3 Brown GA, Swenson DR. A descriptive system for lower extremity evaluation in children: data for the newborn infant. Orthopedics. 2000;23:111–115. 4 Matjacˇic´ Z, Olenˇsek A, Bajd T. Biomechanical characterization and clinical implications of artificially induced toe-walking: differences between pure soleus, pure gastrocnemius, and combinations of soleus and gastrocnemius contractures. J Biomech. 2006;39:255–266. 5 Neptune RR, Kautz SA, Zajac FE. Contributions of the individual ankle plantarflexors to support forward progression and swing initiation during walking. J Biomech. 2001;34:1387–1398. 6 Perry J, Burnfield J, Gronley J, Mulroy S. Toe walking: muscular demands at the ankle and knee. Arch Phys Med Rehabil. 2003;84:7–16. 7 Couillandre A, Maton B, Breniere Y. Voluntary toe-walking gait initiation: electromyographical and biomechanical aspects. Exp Brain Res. 2002;147:313–321. 8 Griffin P, Wheelhouse W, Shiavi R, Bass W. Habitual toe-walkers: a clinical and electromyographic gait analysis. J Bone Joint Surg Am. 1977;59:97–101. 9 Doute DA, Sponseller PD, Tolo VT, et al. Soleus neurectomy for dynamic equinus in children with cerebral palsy. Am J Orthop. 1997;26:613– 616. 10 Neptune RR, Burnfield J, Mulroy S. The neuromuscular demands of toe walking: a forward dynamics simulation analysis. J Biomech. 2007;40:1293–1300. 11 Sasaki K, Neptune RR, Burnfield J, Mulroy S. Muscle compensatory mechanisms during able-bodied toe walking. Gait Posture. 2008;27:440 – 446. 12 Tardieu C, Tabary JC, Tabary C, Heut de la Tour E. Comparison of the sarcomere number and adaptation in young and adult animals. Influence of tendon adaptation. J Physiol (Paris). 1977;73:1045–1055.

13 Williams PE, Goldspink G. Changes in sarcomere length and physiological properties in immobilized muscle. J Anat. 1978;127:459 – 468. 14 Droit S, Boldrini A, Cioni G. Rhythmic leg movements in low-risk and brain-damaged preterm infants. Early Hum Dev. 1996; 44:201–213. 15 Heriza C. Comparison of leg movements in preterm infants at term with healthy full-term infants. Phys Ther. 1988;68: 1687–1693. 16 Van der Heide JC, Paolicelli PB, Bolderini A, Cioni G. Kinematic and qualitative analysis of lower-extremity movements in preterm infants with brain lesions. Phys Ther. 1999;79:546 –557. 17 Jeng S, Chen L, Yau K. Kinematic analysis of kicking movements in preterm infants with very low birth weight and full-term infants. Phys Ther. 2002;82:148 –159. 18 Heathcock JC, Bhat AN, Lobo MA, Galloway JC. The performance of infants born preterm and full-term in the mobile paradym: Learning and memory. Phys Ther. 2004;84:808 – 821. 19 Thelen E, Bradshaw G, Ward JA. Spontaneous kicking in month-old infants: manifestation of a human central locomotor pattern. Behav Neural Biol. 1981;32: 45–53. 20 Katz MM, Mubarak SJ. Hereditary tendo Achillis contractures. J Pediatr Orthop. 1984;4:711–714. 21 Brouwer B, Wheeldon RK, StradiottoParker N, Allum J. Reflex excitability and isometric force production in cerebral palsy: the effect of serial casting. Dev Med Child Neurol. 1998;40:168 –175. 22 Brouwer B, Davidson LK, Olney SJ. Serial casting in idiopathic toe-walkers and children with spastic cerebral palsy. J Pediatr Orthop. 2000;20:221–225. 23 Cottalorda J, Gautheron V, Metton G, et al. Toe-walking in children younger than six years with cerebral palsy: the contribution of serial corrective casts. J Bone Joint Surg Br. 2000;82:541–544. 24 Fiero P. Twin and multiple birth rate: most recent U.S. data. Available at: http://multiples.about.com/od/funfacts/a/twinbirth rate.htm. Accessed January 12, 2009.

DOI: 10.2522/ptj.20070306.ar

e4

f

Physical Therapy

Volume 89

Number 2

February 2009

Research Report

Clinical Prediction Rules for Physical Therapy Interventions: A Systematic Review Jason M Beneciuk, Mark D Bishop, Steven Z George JM Beneciuk, PT, DPT, FAAOMPT, is currently enrolled in the Rehabilitation Sciences Doctoral Program (PhD), Department of Physical Therapy, University of Florida, PO Box 100154, Gainesville, FL 32610-0154 (USA). Address all correspondence to Dr Beneciuk at [email protected]. MD Bishop, PT, PhD, is Assistant Professor, Department of Physical Therapy, University of Florida. SZ George, PT, PhD, is Assistant Professor, Department of Physical Therapy, Brooks Center for Rehabilitation Studies, University of Florida. Mailing address: Department of Physical Therapy, University of Florida, PO Box 100154, Gainesville, FL 32610-0154 (USA). Address all correspondence to Dr George at: [email protected]. [Beneciuk JM, Bishop MD, George SZ. Clinical prediction rules for physical therapy interventions: a systematic review. Phys Ther. 2009;89:114 –124.] © 2009 American Physical Therapy Association

Background and Purpose. Clinical prediction rules (CPRs) involving physical therapy interventions have been published recently. The quality of the studies used to develop the CPRs was not previously considered, a fact that has potential implications for clinical applications and future research. The purpose of this systematic review was to determine the quality of published CPRs developed for physical therapy interventions. Methods. Relevant databases were searched up to June 2008. Studies were included in this review if the explicit purpose was to develop a CPR for conditions commonly treated by physical therapists. Validated CPRs were excluded from this review. Study quality was independently determined by 3 reviewers using standard 18-item criteria for assessing the methodological quality of prognostic studies. Percentage of agreement was calculated for each criterion, and the intraclass correlation coefficient (ICC) was determined for overall quality scores.

Results. Ten studies met the inclusion criteria and were included in this review. Percentage of agreement for individual criteria ranged from 90% to 100%, and the ICC for the overall quality score was .73 (95% confidence interval⫽.27–.92). Criteria commonly not met were adequate description of inclusion or exclusion criteria, inclusion of an inception cohort, adequate follow-up, masked assessments, sufficient sample sizes, and assessments of potential psychosocial factors. Quality scores for individual studies ranged from 48.2% to 74.0%. Discussion and Conclusion. Validation studies are rarely reported in the literature; therefore, CPRs derived from high-quality studies may have the best potential for use in clinical settings. Investigators planning future studies of physical therapy CPRs should consider including inception cohorts, using longer follow-up times, performing masked assessments, recruiting larger sample sizes, and incorporating psychological and psychosocial assessments.

Post a Rapid Response or find The Bottom Line: www.ptjournal.org 114

f

Physical Therapy

Volume 89

Number 2

February 2009

Clinical Prediction Rules for Physical Therapy Interventions

C

linical prediction rules (CPRs) (or clinical decision rules) have become popular in the physical therapy literature. The intent of CPRs is to assist clinicians in making a diagnosis, establishing a prognosis, or implementing an intervention.1 Although it has been suggested that well-constructed CPRs can improve clinical decision making and practice,2 there is a lack of consensus as to what constitutes a methodologically sound CPR, especially in the derivation stage. McGinn and colleagues3 recommended a 3-step process in the development and testing of a CPR. The first step involves the derivation of the CPR (derivation studies), and the second step involves the validation of the CPR (validation studies). The third step involves assessment of the impact of the rule on clinical behavior, also referred to as an “impact analysis.” For CPRs for physical therapy interventions, steps 2 and 3 are not routinely performed. Although it has been suggested that a validated CPR can be applied in various settings with confidence in its accuracy,4 our impression is that most CPRs reported in the physical therapy literature are derivation studies. Furthermore, the lapse in time before validation occurs can be extensive. This situation presents clinicians with the dilemma of whether they should incorporate the results of a derivation study into their clinical practice. Our opinion is that the quality of a derivation study is one factor that should be considered before a CPR is implemented into clinical practice. This interpretation can be difficult, however, because the quality of CPR derivation studies pertaining to interventions has not been reported. Assessing the quality of derivation studies has potential advantages for physical therapy practice and research. First, a quality assessment February 2009

will assist clinicians in deciding whether a given CPR is appropriate for implementation into clinical practice. Second, a quality assessment will assist future researchers in the design of high-quality studies for the development of new CPRs. Therefore, the purpose of this systematic review was to determine the quality of CPRs developed for interventions used by physical therapists. Studies were included in this review if the explicit purpose was to develop a CPR related to a specific intervention approach for conditions commonly treated by physical therapists. Previously validated CPRs were excluded from this review because there is less debate over the clinical application of validated CPRs2 and because methodological concerns about derivation studies are of less concern when a validation study has been reported.

Method Data Sources and Searches A systematic review of relevant databases (PubMed, CINAHL, ProQuest, and Academic Search Premier) from their inception up to June 2008 resulted in the retrieval of 49 potential publications (Figure). The search strategy began with the filter “predict$ OR clinical$ OR outcome$ OR risk$.”5 In comparison with a gold standard,6 this filter has a sensitivity of 98.4% for retrieving CPRs from the literature.5 A second search strategy consisted of the key words “clinical prediction rule.” The first author examined reference lists from all selected publications to verify that no pertinent publications were missed during the above-described searches. Studies were included in this review if the explicit purpose was to develop a CPR related to a specific intervention approach for conditions commonly managed by physical therapists. Previously validated CPRs were excluded from this review.

Data Extraction and Quality Assessment Quality scores were independently assigned to eligible studies by 3 reviewers using a modified version of a list of criteria, reported by Kuijpers et al,7 for assessing the methodological quality of prognostic studies. These criteria were selected because, in our experience, the physical therapy literature has followed a model that uses cohort designs for CPR derivation studies. Therefore, CPR derivation studies involving physical therapy interventions are appropriate for quality assessment tools that are sensitive to methodological issues that affect prognostic studies.8 Another reason why these criteria were appropriate for our purpose was that they were developed by authors aware of issues related to CPRs. Specifically, Kuijpers et al9 later reported a CPR for determining the prognosis for patients with shoulder pain in general practice settings. The original list of criteria7 was altered slightly by removing the criterion related to the rate of response of potential study participants because this item is not commonly reported in the physical therapy literature. Additionally, we added an important criterion by including masking of outcome assessors and treating clinicians.6,8 In our opinion, the resulting criteria are similar to those that have been suggested for evaluating the quality of prognostic studies for patients receiving physical therapy care8 and are consistent with the process of evaluating prognostic variables.10 The 18 criteria used to assess quality in this systematic review represented 8 categories: study population, response information, followup, intervention, outcome, masking, prognostic factors, and data presentation. A description of each criterion is provided in the Appendix.

Volume 89

Number 2

Physical Therapy f

115

Clinical Prediction Rules for Physical Therapy Interventions

Figure. Flow chart depicting search and selection process for clinical prediction rule (CPR).

The criteria could be scored as positive, negative, or unclear. A positive score indicated that the criterion was identified in the study and met specific requirements consistent with a high-quality prognostic study. A negative score indicated that the criterion was identified in the study but did not meet specific requirements. A score of “unclear” meant that the study provided insufficient information regarding that criterion. To obtain a conservative estimate of quality, negative and unclear ratings were collapsed when study quality was rated. A total quality score was 116

f

Physical Therapy

Volume 89

determined by adding positive scores, providing a potential high score of 18 (100%). Ratings of individual studies to determine quality scores were independently assigned by the 3 reviewers before a meeting on the interpretation of the 18-item list of criteria (time 1). The meeting provided an opportunity for reviewers to assess agreement and discuss criteria that resulted in high disagreement. In addition, components for a given criterion that may have been overlooked by reviewers were clarified. As ap-

Number 2

propriate, a quality criterion was updated to reflect an updated interpretation. For example, there were differences in opinion about what constituted a prospective study (criterion E). During the meeting, an operational definition for a prospective study was approved. After the meeting, the 3 reviewers again independently assigned ratings (time 2), this time using the final guidelines provided in the Appendix. Data Synthesis and Analysis Statistical pooling of results was not performed because of the obvious February 2009

Clinical Prediction Rules for Physical Therapy Interventions heterogeneity among studies in populations used, interventions applied, and outcome measures administered. Reliability analyses were performed with SPSS 15.0 for Windows* and Excel.† Percentage of agreement was calculated for individual items. Negative and unclear ratings were collapsed into one variable so that ratings could be dichotomized into 2 categories. Interrater reliability was reported for the total quality score by use of the intraclass correlation coefficient (ICC [2,1]) and respective 95% confidence intervals (CI).11 To determine a single overall quality score for an individual study (Tab. 1), each of the 3 reviewers’ scores for a particular study were averaged to account for the possibility that interrater agreement was less than 100% after time 2. Therefore, the overall quality scores solely reflect the results after time 2 ratings. As in other reviews using this scoring system, high-quality studies were operationally defined as those that had average quality scores of greater than 60%.7

Results The initial search strategy yielded a total of 4,080 publications. A majority of the publications did not meet the criteria for review because they did not pertain to physical therapy interventions described in the Guide to Physical Therapist Practice.12 After titles were reviewed and duplicates were determined, 25 potential publications were further reviewed. Of these, 17 publications were excluded on the basis of the following criteria: not intervention based (n⫽12),9,13–23 nonspecific intervention (n⫽3),24 –26 and eventually validated CPR (n⫽2).27,28 Publications were excluded on the basis of the “nonspecific intervention” criterion * SPSS Inc, 233 S Wacker Dr, Chicago, IL 60606. † Microsoft Corp, One Microsoft Way, Redmond, WA 98052-6399.

February 2009

Table 1. Individual Clinical Prediction Rule (CPR) Studies and Mean Overall Quality Scores % Overall Quality Scorea

CPR (Study)

a

Stabilization exercise for low back pain (LBP) (Hicks et al34)

74.0

Muscle trigger point therapy for chronic tension-type headache (Fernandez-de-las-Penas et al38)

72.2

Inability of people with LBP to show improvement with spinal manipulation (Fritz et al29)

70.4

Thoracic spine manipulation for neck pain (Cleland et al30)

68.5

Multimodal intervention for cervical radiculopathy (Cleland et al37)

61.1

Cervical manipulation for neck pain (Tseng et al31)

59.3

Physical therapist management of cervicogenic headache (Jull and Stanton32)

55.5

Patellar taping for patellofemoral pain syndrome (Lesher et al36)

55.5

Lumbopelvic manipulation for patellofemoral pain syndrome (Iverson et al33)

53.7

Hip mobilization for knee pain and clinical evidence of knee osteoarthritis (Currier et al35)

48.2

Mean of scores from all 3 reviewers after time 2.

if they did not report on the results of a specific physical therapy intervention approach or used multidisciplinary interventions. For example, methods consisting of various combinations of physical therapy interventions that were not specifically described, such as “exercises aimed at restoring neuromuscular control at the shoulder,”26(pp1231–1232) or not described in general, such as “physiotherapy for shoulder pain,”24(p486) were grounds for exclusion. Gross and Battie´25 used a multidisciplinary approach that consisted of physical therapy, occupational therapy, exercise therapy (kinesiology), medicine, and psychology but that was not specific to physical therapy alone; therefore, their study was also excluded. Additionally, studies that potentially included interventions that were not within the scope of physical therapy practice were excluded.26 The remaining 8 studies were included in this review. Two additional publications were included after a review of reference lists and related articles, resulting in the analysis of 10 publications in this review.

Five studies involved CPRs for responses to spinal manipulation.29 –33 The other studies predicted responses to lumbar stabilization,34 hip mobilization,35 patellar taping,36 multimodal interventions for cervical radiculopathy,37 and trigger point therapy for headache38 (Tab. 1). Methodological Criteria Percentage of agreement on ratings of individual items ranged from 70% to 100%; items B, E, H, K, L, and P had the lowest levels of agreement (70%– 86.7%) (Tabs. 2 and 3). After a meeting on the interpretation of the 18-item list of criteria, percentage of agreement on ratings of individual items ranged from 90% to 100% (Tabs. 2 and 3). Individual items commonly rated as low quality (ie, not meeting the criteria in greater than 50% of the studies) were items A, B, F, K, M, and R. Among these items, the inclusion of an inception cohort (item A), description of inclusion and exclusion criteria (item B), and follow-up of ⱖ6 months (item F) were met in ⱕ10% of the studies. The results indicated that for 6 items, all 3 reviewers were in absolute

Volume 89

Number 2

Physical Therapy f

117

Clinical Prediction Rules for Physical Therapy Interventions Table 2. Interrater Reliability

a

Overall Quality Score

Rating Time

Intraclass Correlation Coefficient (95% Confidence Interval)

X (%a)

Range (%a)

1

.67 (.10–.91)

11.47 (63.7)

7–15 (38.8–83.3)

2

.73 (.27–.92)

11.13 (61.8)

7–14 (38.8–77.7)

Overall quality score converted to percentage.

agreement that a given criterion was met across all studies reviewed (eg, all 3 reviewers were in absolute agreement that item B was met in 3% of all studies reviewed) (Tab. 4). Quality Score Absolute agreement among the 3 reviewers on overall quality scores was calculated at time 1 (ICC⫽.67, 95% CI⫽.10 –.91) and after a meeting on

the interpretation of the 18-item list of criteria (time 2) (ICC⫽.73, 95% CI⫽.27–.92) (Tabs. 2 and 3). Next, a mean quality score was calculated for each reviewer (Tab. 1). At time 2, mean quality scores for individual studies (X⫽11.13; 61.8%) ranged from 8.67 to 13.33 (48.2%–74.0%). Five studies29,30,34,37,38 were rated at greater than 60% (range⫽61.1%– 74.0%). Four studies31–33,36 were

Table 3. Individual Item Percentage of Agreement % Agreement Among All 3 Reviewers Time 1 Rating

Time 2 Rating

Inception cohort

90.0

90.0

B

Inclusion/exclusion criteria

70.0

96.7

C

Study population

93.3

93.3

D

Nonresponders vs responders

90.0

90.0

E

Prospective data collection

86.7

96.7

F

Follow-up at ⱖ6 mo

100.0

100.0

G

Dropouts/loss to follow-up of ⬍20%

93.3

93.3

H

Information on subjects completing study vs loss to follow-up/dropouts

83.3

90.0

Item A

118

Description

I

Intervention fully described/standardized

96.7

96.7

J

Standardized assessment of relevant outcome criteria

96.7

96.7

K

Masking of outcome assessor and treating clinician

86.7

93.3

L

Standardized assessment of subject characteristics and potential clinical prognostic factors

83.3

96.7

M

Standardized assessment of potential psychosocial prognostic factors

96.7

96.7

N

Frequencies of most important outcome measures

90.0

90.0

O

Frequencies of most important prognostic factors

96.7

96.7

P

Appropriate analysis techniques

86.7

100.0

Q

Prognostic model presented

93.3

93.3

R

Sufficient numbers of subjects

90.0

90.0

f

Physical Therapy

Volume 89

Number 2

rated at 50% to 60% (range⫽53.7%– 59.3%), and one study35 was rated at less than 50% (48.2%).

Discussion The intent of this review was to rate the methodological quality of CPR derivation studies reported in the physical therapy literature. This is an important issue because the quality of derivation studies for physical therapy interventions has not been systematically considered. The lack of consideration of derivation studies has important clinical and research implications. For example, CPR validation studies often are not performed, a fact that makes the decision to incorporate CPRs into clinical practice a potentially difficult one. One CPR has been validated in clinical settings with samples of patients similar to those included in the derivation study.28 However, this CPR has not been validated in different settings with new groups of patients and different intervention parameters.39 On the basis of the results of this review, several CPRs may be appropriate for clinical applications involving patients and clinical environments similar to those used in the CPR derivation process. We acknowledge that the quality of a derivation study is only one aspect of clinical decision making for the use of a CPR. However, it is information that was not previously reported and therefore might enhance clinical decision making. We used a quality score of greater than 60% as a cutoff score for a high-quality study, because this threshold was used in a previous review of prognostic studies.7 Studies that met this quality index included CPRs for determining factors associated with responses to stabilization exercises,34 responses to muscle trigger point therapy for tension-type headaches,38 the inability of patients with low back pain to show improvement with spinal maFebruary 2009

Clinical Prediction Rules for Physical Therapy Interventions nipulation,29 manipulation of the thoracic spine in patients diagnosed with mechanical neck pain,30 and a multimodal intervention approach for cervical radiculopathy.37 The lower-quality studies included CPRs for predicting favorable responses to cervical manipulation in patients with neck pain,31 the management of cervicogenic headache,32 patellar taping,36 lumbopelvic manipulation in patients with patellofemoral pain syndrome,33 and hip mobilization for knee pain indicative of osteoarthritis.35 Quality scores can assist clinicians in deciding whether to use these nonvalidated CPRs. However, quality scores are not a substitute for CPR validation studies. Validation studies provide more-definitive information for clinical applications because they are independent studies of new subjects and involve a variety of clinicians and patients.1,3,6,40 An important factor to consider for a methodologically sound CPR derivation is the risk-to-benefit ratio associated with its application. Risk was not empirically assessed in the CPRs considered in this review, a fact that is not surprising given the status of the physical therapy literature.41 Clinical prediction rules were originally developed in the medical profession for decisions involving higher associated risks, such as those associated with traumatic injuries.42– 45 The risk associated with the interventions used in the CPRs considered in this review is believed to be minimal in comparison with the risk associated with emergency medicine. For example, consider the risk associated with the use of stabilization exercises for low back pain34 in comparison with the risk of not ordering radiographs for traumatic injuries.42– 44 Failure to detect a fracture is associated with a risk higher than that associated with the use of stabilization exercises when those are not indicated. Unfortunately, the current physical therapy literature February 2009

Table 4. Methodological Criteria Commonly Receiving Low Ratings Item

% Positive Ratinga

Description

B

Inclusion/exclusion criteria

A

Inception cohort

10

3

F

Follow-up at ⱖ6 mo

10

K

Masking

27

R

Sufficient numbers of subjects

40

M

Standardized assessment of potential psychosocial prognostic factors

47

a Items for which all 3 reviewers agreed (absolute agreement) that a given criterion was met across the 10 studies reviewed (eg, all 3 reviewers were in absolute agreement that item B was met in 3% of all studies reviewed).

does not allow a quantitative consideration of the risk-to-benefit ratio, so clinical decisions must be based on qualitative factors. Individual items that received lowquality ratings were similar to previously suggested areas of methodological concern for CPR studies.2,6,40 Specifically, masking of outcome assessors and treating clinicians (item K) did not occur in a majority of the studies reviewed. Masking of outcome assessors and treating clinicians is important for limiting the measurement bias of potential predictor variables.1,3,8 Additional areas of concern identified in this review included the use of an inception cohort and definition of the duration of symptoms in eligibility criteria. To limit potential error in establishing a prognosis, subjects should be enrolled in a common time frame with regard to their current condition.8 This criterion was not a component in a majority of the studies used to develop CPRs. Therefore, samples used to develop CPRs may lack homogeneity, thereby increasing the potential for bias in predictor variables and outcomes. Another area lacking in CPR derivation studies published to date was the follow-up period, which was suggested to be at least 6 months. Immediate effects were commonly reported; such immediate effects

might be beneficial only in demonstrating evidence of responsiveness.8 Longer follow-up times are needed to demonstrate valid clinical implications for the use of a given intervention. Finally, an assessment of potential psychosocial prognostic factors (item M) was commonly not included. Psychological factors, such as kinesophobia, catastrophizing, anxiety, and depression might have strong influences on outcomes related to musculoskeletal conditions.46 – 49 Incorporating these factors into the development process has important clinical implications for future CPRs. Limited sample sizes have been reported to be common methodological flaws in CPR studies.2 It has been suggested that 10 to 15 subjects are required for each prospective predictor variable in CPR studies.50 Not meeting this requirement may lead to inaccurate statistical results because of overfitting of regression models.50 It is important that our sample size determination was based on the final CPR model and not on initial prospective variables. The result was that only 40% of studies had an adequate sample size, and this was a liberal estimate. If we had elected to use initial prospective variables, then no studies would have met the criterion for sample size, a result suggesting that previ-

Volume 89

Number 2

Physical Therapy f

119

Clinical Prediction Rules for Physical Therapy Interventions ously noted concerns about small sample sizes used in CPR derivation studies are legitimate. We suggest that future studies include larger sample sizes to account for derivation regression models in addition to the final, more parsimonious models. Several limitations of this systematic review should be considered in the interpretation of the results. Although the results may be relevant to the decision-making process for implementing a CPR in practice, our findings should not be viewed as definitive. Our data provide complementary information on which CPRs to use in clinical practice, but the ultimate decision must be made in the context of a clinician’s experience and factors specific to the encounter with a patient. These factors include, but are not limited to, whether patients seen in clinical practice are similar to those enrolled in the respective CPR study and whether a quantitative assessment of the risk-to-benefit ratio is available. Another limitation is that the quality criteria used in this review were developed for prognostic studies, not specifically for CPR derivation studies involving interventions. Although these study designs are similar, there may be subtle differences with regard to quality determinations. The quality criteria used in this review did not include certain statistical elements that may have important implications for CPRs. For example, the criteria did not include the consideration of a quantitative risk-to-benefit analysis,41 reporting of potential predictor variable reliability,2,6 or reporting of CIs2,6 and effect sizes.51 Furthermore, the quality scores were equally weighted so that the “methodological importance” of a category was equally distributed among all of the criteria. This decision was made because we did not have clear evidence to follow for weighting decisions. Another relevant issue is that 120

f

Physical Therapy

Volume 89

randomized designs have been suggested to be appropriate for CPR studies involving intervention selection.6,40 Although this may be true, it appears that the use of cohort studies is much more common in the physical therapy literature, because only one study included in this review used a randomized design.32 Therefore, future assessments of the quality of CPR derivation studies should include the development of a standardized rating system with a more-specific statistical criterion, consideration of weighting of quality scores on the basis of the methodological importance of particular categories, and the development of a criterion that is sensitive enough to determine the overall quality of a study design (such as distinguishing between cohort and randomized designs). There was substantial agreement among the raters on individual items and overall quality scores; however, it is clear that agreement can be improved. Improvement can be accomplished by providing quality criteria more explicit than previously reported criteria, especially with regard to inception cohort, responders versus nonresponders, frequency of outcome measures, and sample size determination. The reliability estimates were also imprecise (large 95% CIs); we believe that this result may have been attributable to the relatively small number of studies included in this review. Additionally, we opted to collapse negative and unclear ratings, a strategy that may have influenced the percentage of agreement among the reviewers. However, this decision to collapse the data was made a priori and, even if the data had not been collapsed, the overall quality scores would not have been affected (Tab. 1). These scores considered only positive ratings because negative and unclear ratings were equally weighted as “0”

Number 2

when the overall quality scores were determined.7

Conclusion This systematic review provides impetus for future research. First, it reinforces the importance and priority of performing validation studies. There are currently 10 CPR derivation studies in the physical therapy literature that have not been validated, and these studies vary greatly in overall quality. A glut of CPR derivation studies with various degrees of quality may only further confuse clinical practice, a result that is contrary to the original intent of CPRs. It is clear that follow-up validation studies are a high priority for advancing clinical practice. Second, the results of this review provide a clear direction for investigators wishing to conduct high-quality CPR derivation studies. In our opinion, the areas that should be a high priority for future derivation studies aimed at CPR development are recruiting inception cohorts, performing masked assessments, including longer follow-up times, collecting larger sample sizes, and incorporating psychological or psychosocial assessments. All authors provided concept/idea/research design and writing. Dr Beneciuk and Dr Bishop provided data collection. Dr Beneciuk and Dr George provided data analysis. Dr George provided project management and fund procurement. Dr Bishop and Dr George provided consultation (including review of manuscript before submission). Dr Beneciuk was supported by a National Institutes of Health T-32 Neural Plasticity Research Training Fellowship (grant T32HD043730). Dr Bishop and Dr George were supported by grant R21 AT002796 awarded to Dr George from the National Institutes of Health/National Center for Complementary and Alternative Medicine. This article was received August 7, 2008, and was accepted October 30, 2008. DOI: 10.2522/ptj.20080239

February 2009

Clinical Prediction Rules for Physical Therapy Interventions References 1 Childs JD, Cleland JA. Development and application of clinical prediction rules to improve decision making in physical therapist practice. Phys Ther. 2006;86: 122–131. 2 Cook C. Potential pitfalls of clinical prediction rules [editorial]. J Man Manip Ther. 2008;16:69 –71. 3 McGinn TG, Guyatt GH, Wyer PC, et al. Users’ guides to the medical literature, XXII: how to use articles about clinical decision rules. Evidence-Based Medicine Working Group. JAMA. 2000;284:79 – 84. 4 McGinn TG, Guyatt GH, Wyer PC, et al. Diagnosis: clinical prediction rule. In: Guyatt GH, Rennie D, eds. Users’ Guide to the Medical Literature: A Manual for Evidence-Based Clinical Practice. Chicago, IL: AMA Press; 2002:471– 483. 5 Ingui BJ, Rogers MA. Searching for clinical prediction rules in MEDLINE. J Am Med Inform Assoc. 2001;8:391–397. 6 Laupacis A, Sekar N, Stiell IG. Clinical prediction rules: a review and suggested modifications of methodological standards. JAMA. 1997;277:488 – 494. 7 Kuijpers T, van der Windt DA, van der Heijden GJ, Bouter LM. Systematic review of prognostic cohort studies on shoulder disorders. Pain. 2004;109:420 – 431. 8 Beattie PF, Nelson RM. Evaluating research studies that address prognosis for patients receiving physical therapy care: a clinical update. Phys Ther. 2007;87:1527–1535. 9 Kuijpers T, van der Windt DA, Boeke AJ, et al. Clinical prediction rules for the prognosis of shoulder pain in general practice. Pain. 2006;120:276 –285. 10 Altman DG. Systematic reviews of evaluations of prognostic variables. BMJ. 2001;323:224 –228. 11 Shrout P, Fleiss JL. Intraclass correlations: uses in assessing rater reliability. Psychol Bull. 1979;86:420 – 428. 12 Guide to Physical Therapist Practice. 2nd ed. Phys Ther. 2001;81:9 –746. 13 Maher C. Clinical prediction rule for return to work after back pain. CMAJ. 2005;172:1575–1576. 14 Dionne CE, Bourbonnais R, Fremont P, et al. A clinical return-to-work rule for patients with back pain. CMAJ. 2005;172: 1559 –1567. 15 Hancock MJ, Maher CG, Latimer J, et al. Can rate of recovery be predicted in patients with acute low back pain? Development of a clinical prediction rule. Eur J Pain. 28 April 2008 (epub ahead of print). 16 Kuijpers T, van der Windt DA, van der Heijden GJ, et al. A prediction rule for shoulder pain related sick leave: a prospective cohort study. BMC Musculoskelet Disord. 2006;7:97. 17 Kuijpers T, van der Heijden GJ, Vergouwe Y, et al. Good generalizability of a prediction rule for prediction of persistent shoulder pain in the short term. J Clin Epidemiol. 2007;60:947–953.

February 2009

18 Suissa S, Harder S, Veilleux M. The relation between initial symptoms and signs and the prognosis of whiplash. Eur Spine J. 2001;10:44 – 49. 19 Feuerstein M, Huang GD, Haufler AJ, Miller JK. Development of a screen for predicting clinical outcomes in patients with work-related upper extremity disorders. J Occup Environ Med. 2000;42: 749 –761. 20 Heymans MWAJ, van Buuren S, Knol DL, et al. Return to work in a cohort of low back pain patients: development and validation of a clinical prediction rule [in Dutch]. Nederlands Tijdschrift Voor Fysiotherapie. 2007;117:199 –206. 21 Kongsted A, Bendix T, Qerama E, et al. Acute stress response and recovery after whiplash injuries: a one-year prospective study. Eur J Pain. 2008;12:455– 463. 22 Wolfe F, Lane NE. The long-term outcome of osteoarthritis: rates and predictors of joint space narrowing in symptomatic patients with knee osteoarthritis. J Rheumatol. 2002;29:139 –146. 23 Enthoven P, Skargren E, Kjellman G, Oberg B. Course of back pain in primary care: a prospective study of physical measures. J Rehabil Med. 2003;35:168 –173. 24 Kennedy CA, Haines T, Beaton DE. Eight predictive factors associated with response patterns during physiotherapy for soft tissue shoulder disorders were identified. J Clin Epidemiol. 2006;59:485– 496. 25 Gross DP, Battie´ MC. Predicting timely recovery and recurrence following multidisciplinary rehabilitation in patients with compensated low back pain. Spine. 2005;30:235–240. 26 Ginn KA, Cohen ML. Conservative treatment for shoulder pain: prognostic indicators of outcome. Arch Phys Med Rehabil. 2004;85:1231–1235. 27 Flynn T, Fritz J, Whitman J, et al. A clinical prediction rule for classifying patients with low back pain who demonstrate short-term improvement with spinal manipulation. Spine. 2002;27:2835–2843. 28 Childs JD, Fritz JM, Flynn TW, et al. A clinical prediction rule to identify patients with low back pain most likely to benefit from spinal manipulation: a validation study. Ann Intern Med. 2004;141: 920 –928. 29 Fritz JM, Whitman JM, Flynn TW, et al. Factors related to the inability of individuals with low back pain to improve with a spinal manipulation. Phys Ther. 2004;84: 173–190. 30 Cleland JA, Childs JD, Fritz JM, et al. Development of a clinical prediction rule for guiding treatment of a subgroup of patients with neck pain: use of thoracic spine manipulation, exercise, and patient education. Phys Ther. 2007;87:9 –23. 31 Tseng YL, Wang WT, Chen WY, et al. Predictors for the immediate responders to cervical manipulation in patients with neck pain. Man Ther. 2006;11:306 –315.

32 Jull GA, Stanton WR. Predictors of responsiveness to physiotherapy management of cervicogenic headache. Cephalalgia. 2005;25:101–108. 33 Iverson CA, Sutlive TG, Crowell MS, et al. Lumbopelvic manipulation for the treatment of patients with patellofemoral pain syndrome: development of a clinical prediction rule. J Orthop Sports Phys Ther. 2008;38:297–312. 34 Hicks GE, Fritz JM, Delitto A, McGill SM. Preliminary development of a clinical prediction rule for determining which patients with low back pain will respond to a stabilization exercise program. Arch Phys Med Rehabil. 2005;86:1753–1762. 35 Currier LL, Froehlich PJ, Carow SD, et al. Development of a clinical prediction rule to identify patients with knee pain and clinical evidence of knee osteoarthritis who demonstrate a favorable short-term response to hip mobilization. Phys Ther. 2007;87:1106 –1119. 36 Lesher JD, Sutlive TG, Miller GA, et al. Development of a clinical prediction rule for classifying patients with patellofemoral pain syndrome who respond to patellar taping. J Orthop Sports Phys Ther. 2006;36:854 – 866. 37 Cleland JA, Fritz JM, Whitman JM, Heath R. Predictors of short-term outcome in people with a clinical diagnosis of cervical radiculopathy. Phys Ther. 2007;87: 1619 –1632. 38 Fernandez-de-las-Penas C, Cleland JA, Cuadrado ML, Pareja JA. Predictor variables for identifying patients with chronic tension-type headache who are likely to achieve short-term success with muscle trigger point therapy. Cephalalgia. 2008; 28:264 –275. 39 Hancock MJ, Maher CG, Latimer J, et al. Independent evaluation of a clinical prediction rule for spinal manipulative therapy: a randomised controlled trial. Eur Spine J. 2008;17:936 –943. 40 Beattie P, Nelson R. Clinical prediction rules: what are they and what do they tell us? Aust J Physiother. 2006;52:157–163. 41 Newman D, Allison SC. Risk and physical therapy? J Orthop Sports Phys Ther. 2007;37:287–289. 42 Stiell IG, Greenberg GH, McKnight RD, et al. A study to develop clinical decision rules for the use of radiography in acute ankle injuries. Ann Emerg Med. 1992; 21:384 –390. 43 Stiell IG, Greenberg GH, Wells GA, et al. Derivation of a decision rule for the use of radiography in acute knee injuries. Ann Emerg Med. 1995;26:405– 413. 44 Stiell IG, Wells GA, Vandemheen KL, et al. The Canadian C-spine rule for radiography in alert and stable trauma patients. JAMA. 2001;286:1841–1848. 45 Stiell IG, Wells GA, Vandemheen KL, et al. The Canadian CT head rule for patients with minor head injury. Lancet. 2001;357: 1391–1396.

Volume 89

Number 2

Physical Therapy f

121

Clinical Prediction Rules for Physical Therapy Interventions 46 Carroll LJ, Cassidy JD, Cote P. Depression as a risk factor for onset of an episode of troublesome neck and low back pain. Pain. 2004;107:134 –139. 47 Crombez G, Vlaeyen JW, Heuts PH, Lysens R. Pain-related fear is more disabling than pain itself: evidence on the role of painrelated fear in chronic back pain disability. Pain. 1999;80:329 –339.

122

f

Physical Therapy

Volume 89

48 Leeuw M, Goossens ME, Linton SJ, et al. The fear-avoidance model of musculoskeletal pain: current state of scientific evidence. J Behav Med. 2007;30:77–94. 49 Vlaeyen JW, Linton SJ. Fear-avoidance and its consequences in chronic musculoskeletal pain: a state of the art. Pain. 2000; 85:317–332.

Number 2

50 Concato J, Feinstein AR, Holford TR. The risk of determining risk with multivariable models. Ann Intern Med. 1993;118: 201–210. 51 Cook C. Clinimetrics corner: use of effect sizes in describing data. J Man Manip Ther. 2008;16:E54 –E57.

February 2009

Clinical Prediction Rules for Physical Therapy Interventions Appendix. Eighteen-Item List of Criteria for Assessing the Methodological Quality of Studiesa

A. Positive if subjects were identified at an early uniform point (inception cohort) in the course of the condition (first episode, with restriction to duration of symptoms mentioned, of their respective complaint or first physical therapy–related intervention episode for their respective complaint). B. Positive if criteria were formulated for at least age, duration of symptoms, and relevant comorbidities. C. Positive if setting in which subjects were treated was described. D. Positive if information was presented about subject or condition characteristics of responders and nonresponders or if there was no selective response. E. Positive if a prospective design was used (immediate or same-day follow-up was not considered prospective). F. Positive if the follow-up period was ⱖ6 mo. G. Positive if the total number of subjects was ⱖ80% at the last moment of final follow-up compared with the number of subjects at baseline. H. Positive if demographic or clinical information (subject or condition characteristics, such as age, sex, and other potential prognostic predictors) was presented for subjects completing the study and those lost to follow-up/ dropouts at the main moment of baseline outcome measurement, or no selective dropouts/lost to follow-up, or no dropouts/lost to follow-up. I. Positive if the intervention subsequent to inclusion in a cohort was fully described or standardized (treating clinicians had to adhere to a strict protocol and were not permitted to adjust the intervention on the basis of their independent decision-making processes). J. Positive if standardized questionnaires or quantitative measurements of at least 1 of the following 5 outcome measures were used for each follow-up measurement: pain, general improvement, functional status, general health status, or lost days of work. K. Positive if masking of the outcome assessor and treating clinician was achieved. In studies in which selfadministered questionnaires were used, masking of the outcome assessor portion of this criterion would be considered acceptable but would have no bearing on the treating clinician status. L. Positive if standardized questionnaires or objective measurements were used at baseline for at least 4 of the following 6 potential prognostic factors: age, sex, pain, functional status, duration of complaints, or physical work load. M. Positive if standardized questionnaires or objective measurements were used at baseline for at least 1 of the following 7 potential prognostic factors: depression, somatization, distress, fear-avoidance, coping strategies, anxiety, or psychosocial work-related factors (social support, psychological demands, and job decision latitude). N. Positive if frequency, percentage, or mean, median, and standard deviation or confidence interval were reported for the most important outcome measures. O. Positive if frequency, percentage, or mean, median, and standard deviation or confidence interval were reported for the most important prognostic factors. (continued)

February 2009

Volume 89

Number 2

Physical Therapy f

123

Clinical Prediction Rules for Physical Therapy Interventions Appendix. Continued

P. Positive if univariate crude estimates were provided for the association of a prognostic factor with outcome. Q. Positive if an attempt was made to determine a set of prognostic factors with the highest prognostic value. R. Positive if the number of cases in the multivariate analysis was at least 10 times the number of independent variables in the multivariate analysis (on the basis of the final clinical prediction rule model, not the initial prospective variables). a Criteria B, E, H, and R required consensus agreement or clarification before the second rating process, as follows: B— clarification regarding the inclusion of duration of symptoms; E— consensus agreement on the operational definition of a prospective design; H— clarification regarding information provided for dropouts/lost to follow-up; and R— clarification regarding the number of cases in the multivariate analysis being at least 10 times the number of independent variables in the multivariate analysis (on the basis of the final clinical prediction rule model).

124

f

Physical Therapy

Volume 89

Number 2

February 2009

Research Report

Use of Standardized Outcome Measures in Physical Therapist Practice: Perceptions and Applications Diane U Jette, James Halbert, Courtney Iverson, Erin Miceli, Palak Shah

Background. Standardized instruments for measuring patients’ activity limitations and participation restrictions have been advocated for use by rehabilitation professionals for many years. The available literature provides few recent reports of the use of these measures by physical therapists in the United States.

Objective. The primary purpose of this study was to determine: (1) the extent of

DU Jette, PT, DSc, is Professor and Chair, Department of Rehabilitation and Movement Science, University of Vermont, Rowell 305, 106 Carrigan Dr, Burlington, VT 05405 (USA). Address all correspondence to Dr Jette at: diane. [email protected].

the use of standardized outcome measures and (2) perceptions regarding their benefits and barriers to their use. A secondary purpose was to examine factors associated with their use among physical therapists in clinical practice.

J Halbert, PT, BS; C Iverson, PT, BS; E Miceli, PT, BS; and P Shah, PT, MS, are students at the University of Vermont.

Design. The study used an observational design. ceived benefits and barriers of standardized outcome measures was sent to 1,000 randomly selected members of the American Physical Therapy Association (APTA).

[Jette DU, Halbert J, Iverson C, et al. Use of standardized outcome measures in physical therapist practice: perceptions and applications. Phys Ther. 2009;89: 125–135.]

Results. Forty-eight percent of participants used standardized outcome measures.

© 2009 American Physical Therapy Association

Methods. A survey questionnaire comprising items regarding the use and per-

The majority of participants (⬎90%) who used such measures believed that they enhanced communication with patients and helped direct the plan of care. The most frequently reported reasons for not using such measures included length of time for patients to complete them, length of time for clinicians to analyze the data, and difficulty for patients in completing them independently. Use of standardized outcome measures was related to specialty certification status, practice setting, and the age of the majority of patients treated.

Limitations. The limitations included an unvalidated survey for data collection and a sample limited to APTA members.

Conclusions. Despite more than a decade of development and testing of standardized outcome measures appropriate for various conditions and practice settings, physical therapists have some distance to go in implementing their use routinely in most clinical settings. Based on the perceived barriers, alterations in practice management strategies and the instruments themselves may be necessary to increase their use.

Post a Rapid Response or find The Bottom Line: www.ptjournal.org February 2009

Volume 89

Number 2

Physical Therapy f

125

Standardized Outcome Measures in Physical Therapist Practice

S

tandardized instruments measuring various aspects of health status have been advocated for use by rehabilitation professionals for many years, and much has been written about the potential benefits of, and barriers to, the use of such measures in practice.1–5 Additionally, many such instruments have been developed for use for patients with the various conditions managed by physical therapists. These instruments have been referred to in the literature using different terms such as “health status measures,” “disability measures,” “outcome measures,” and “quality-of-life measures.” In general, they assess the actual or perceived ability of an individual to carry out activities such as moving in an environment or completing personal care and to participate in life situations such as work or household management. The literature, however, also includes studies in which physical therapists have defined these measures to include assessment of body function.6 –9 Although referred to by different terms and defined at different levels, these measures, in general, are standardized in that they use closed-ended questionnaire formats or specific protocols for implementation, provide scores that allow quantitative assessment of ability, and have been evaluated for their psychometric properties. When they are used to determine the change in ability from before to after an intervention, they may be referred to as outcome measures.

know of any clinical trials that have demonstrated the direct effects of using standardized outcome measures, suggested benefits include identifying patients who are at risk for poor or adverse outcomes,4 facilitating improved continuity of care for patients transitioning from one health care setting to another,11 determining the most cost-effective settings for patients to receive rehabilitation services,11 assessing practitioner and organizational performance,4 and determining the most-effective interventions for particular conditions.4

cluded questions about use of a variety of types of outcomes measures; however, the authors included manual muscle testing and goniometric measurements in their definition of outcomes measures. In the 1998 study, a high proportion of respondents used manual muscle testing (88%) and goniometry (90%), whereas relatively low proportions used measures such as the Functional Independence Measure (FIM) (18%) or the Impairment Inventory scale of the Chedoke-McMaster Stroke Assessment (35%).

The need for physical therapists to use standardized outcome measures has been recognized at the national level in the United States. The Centers for Medicare & Medicaid Services sponsored a report in 2006 to determine the possibility of a uniform rehabilitation outcomes assessment method for patients leaving acute care.11 The authors proposed several purposes for this type of assessment, including provider decision making, patient safety, and ability to determine patients’ health and function longitudinally.11 On a smaller scale, the Commission on Accreditation in Physical Therapy Education12 supports the use of standardized outcome measures in practice by requiring all education programs to demonstrate that their graduates have some experience in using and interpreting them during their professional (entry-level) education.

The drive for use of standardized outcome measures in practice has been motivated to some extent by the recognition that goals for patients’ improvement not only must consider the traditionally measured impairments in body function (eg, range of motion, strength [force-generating capacity]) but also should consider patients’ points of view and preferences for daily activities and life participation.10 Although we do not

The literature provides relatively few reports of the overall use of standardized outcome measures by physical therapists. Physical therapists in 5 academically affiliated institutions in Toronto were surveyed in 19929 and again in 19988 to determine their use of standardized outcome measures and the perceived obstacles to their use. A second part of the latter study used qualitative methods to explicate the findings.7 The studies in-

In 1997, a study examining the use of outcome measures in rehabilitation centers in the United Kingdom showed that 77% of the centers used at least one tool; of those centers, 28% used some measures of general motor function, and 88% used at least one measure of disability.13 In 2001, 2 studies were published that examined the use of outcome measures in Europe.6,14 Haigh et al6 found that a few rehabilitation centers used a large number of tools on a small proportion of patients. For patients with orthopedic conditions, the outcomes measured were largely at the body function level. For patients with neurological conditions, disease-specific measures of disability were used more frequently. There was minimal use of generic measurement tools that can be used regardless of condition. Although specific data were not reported, Torenbeek et al14 noted low overall satisfaction with outcome measurement for patients with stroke and low back pain among rehabilitation professionals in 5 European countries. In addition, there was little consensus about which outcome measures to use. In a study of physical therapists in outpatient clinics in the United States, Russek et al15 found that only 50% of the respondents used the outcome tools they had been provided by their clinics’ corporate owner.

126

f

Physical Therapy

Volume 89

Number 2

February 2009

Standardized Outcome Measures in Physical Therapist Practice A few studies7,8,13,15,16 have examined perceptions of the benefits of and barriers to using standardized outcome measures among rehabilitation professionals, and many of the reported barriers were similar across studies. Perceptions about barriers include lack of time and inconvenience; lack of familiarity, knowhow, and training; and lack of resources such as staffing and automation. Attitudes and perceptions related to use of outcome measures among other health care providers, including mental health practitioners, oncologists, general practitioners (GPs), and nurses, also have been reported. Garland et al3 found variability in attitudes across mental health practitioners, but noted that, in general, the responses reflected ambivalence. All of the practitioners interviewed had participated in mandated outcome assessments, yet they reported being more likely to use their own intuition than standardized measures to evaluate clients’ progress. Similarly, Taylor et al17 reported that many oncologists they interviewed relied on their own impressions and informal assessments of patients’ quality of life to inform their decisions. Most respondents argued that the use of standardized measures made decision making more difficult rather than facilitating it. As in the previously mentioned studies, approximately one half of GPs and nurses interviewed in a study by Meadows et al18 said that they preferred relying on their own clinical judgment in the management of their patients. Because of the lack of recent information about the use of standardized outcome measures among physical therapists in the United States and the professional and governmental emphasis on the collection and application of data from such instruments, this study was conducted to determine the extent of their use,

February 2009

their clinical applications, perceptions of their value, and barriers to their use. Secondarily, we examined the relationships between practice setting and therapist characteristics and the use of standardized outcome measures.

the instruments as “health status questionnaires.” In an attempt to be consistent with terms used in the most recent rehabilitation literature, we use the term “standardized outcome measures” throughout this article, recognizing the various terms used to identify these measures.

Method Procedure One thousand potential participants were randomly selected from the membership list of the American Physical Therapy Association (APTA). The sample size was determined based on an estimated 50% return rate and a desire for a 95% confidence interval of 5 or less if a response was chosen by 50% of the sample. The random selection process was computer generated and stratified by geographic area. In March 2008, these individuals received a survey questionnaire and a letter explaining the purpose of the study and requesting return of the completed survey questionnaire by postage-paid return mail. Participation was presumed to indicate informed consent. The letter sent to potential participants noted that the instruments we were asking about were “referred to by various names and often include information that is related to patients’/clients’ social, physical, or psychological status as they relate to daily activities or role participation. Examples include Oswestry Low Back Pain Questionnaire, Functional Independence Measure (FIM), Arthritis Impact Questionnaire (AIM), and SF-36 [Medical Outcome Study 36-Item Short-Form Health Survey]. This study asks you to think broadly about the measures.” The questionnaire indicated that in thinking broadly, respondents should consider instruments “described with terms such as ‘health status,’ ‘quality of life,’ ‘disability,’ ‘functional status,’ or ‘outcomes measures.’” In the survey questionnaire, we referred to

Approximately 3 weeks after the initial mailing, those therapists who did not respond and who had e-mail addresses listed in the APTA Web site directory were sent a reminder e-mail, with the survey questionnaire and letter as attachments. After an additional week, another survey questionnaire was mailed to those who had not responded to the initial mailing or e-mail. Instrument The survey instrument (eAppendix 1 available at http://www.ptjournal. org) was designed by the investigators. The initial draft was sent to 14 clinician colleagues for input. Eight clinicians in various types of practice, including acute care, outpatient hospital-based care, and private practice, responded. They had between 15 and 30 years of practice as physical therapists. They were asked to assess the face and content validity of the items in the survey instrument, to indicate whether there were important gaps, and to indicate whether any items were unclear or confusing. Changes to the survey instrument were made based on their feedback. We also used the previous literature (cited in the introduction of this report) related to health care practitioners’ attitudes toward, and use of, standardized outcome measures to support the content validity of the instrument. Construct validity of the parts of the instrument that assessed beliefs about the usefulness of and barriers to using instruments in practice was assessed through factor analysis. A principal components factor analysis with varimax rotation resulted in 5 factors that explained

Volume 89

Number 2

Physical Therapy f

127

Standardized Outcome Measures in Physical Therapist Practice 57% of the variance in item responses. Cronbach alpha was determined for each of the factors to provide evidence for internal consistency. We interpreted the 5 factors to support the framework for attitudes and beliefs provided by the literature. The factors represented benefits for the management of the patient (7 items, ␣⫽.85), problems or limitations for the physical therapist (6 items, ␣⫽.77), problems or limitations for the patient (6 items, ␣⫽.77), benefits for external communication (3 items, ␣⫽.67), and limitations due to culture or language (2 items, ␣⫽.59). Taken all together, the internal consistency of the items related to beliefs about the benefits of using standardized outcome measures was good (␣⫽.84). The internal consistency of all items related to beliefs about problems of or barriers to the use of standardized outcome measures was similarly good (␣⫽.83). Data Analysis Data were analyzed using SPSS statistical software, version 15.0.* Response frequencies and means or medians for the survey items were determined and displayed in tabular and graphic formats. After examining the response frequencies, and before examining the associations among variables, some variable categories were collapsed in order to allow further analysis and derive stable models. Logistic regression analyses were conducted to examine the association of participant and practice characteristics with the use of standardized outcome measures. We used a forward selection process to derive models, requiring P⬍.05 to enter and P⬍.10 to delete. Odds ratios and their 95% confidence intervals were recorded for each level of the inde* SPSS Inc, 233 S Wacker Dr, Chicago, IL 60606.

128

f

Physical Therapy

Volume 89

pendent variables that were significant. We chose one level of each variable as a reference group to allow the most salient interpretation of results.

Results Participants Completed questionnaires were received from 498 participants, for a response rate of 49.8%. Three questionnaires were returned as undeliverable, 1 questionnaire was returned with no responses, and 38 questionnaires were returned with respondents indicating that they did not manage patient care. We, therefore, had 456 usable questionnaires. Similar response rates have been reported by Haigh et al,6 Russek et al,15 and Hatfield and Ogles.19 Sixty-eight percent of the participants were female, and 32% were male. The majority (61%) worked in an outpatient setting. A slim majority (53.4%) of participants had postbaccalaureate professional degrees. Thirty-two percent were certified clinical specialists. Although not formally tested, the sample seemed to reflect the demographics of APTA members reported in 2006 and 2007 fairly well.20 Our sample had a slightly greater proportion of those with postbaccalaureate degrees and less time in practice. Our sample also appears to have had slightly more therapists working in outpatient and acute care settings. It is difficult to determine whether these differences were due to the different time frames in which the data were collected or to bias in the sample. Participant and practice characteristics of the sample are shown in Tables 1 and 2, respectively. Overall Perceptions of Standardized Outcome Measures Of the 456 participants, 218 (47.8%) indicated that they used standardized outcome measures in practice. Table 3 shows the perceived benefits

Number 2

of and problems with using standardized outcome measures in practice among the participants who used them. More than 90% of the participants who used them agreed that standardized outcome measures enhance communication with patients and help to direct a plan of care. More than 75% of the participants who used them agreed that problems with standardized outcome measures are that they are confusing to patients, difficult for patients to complete, and too time consuming for patients. Implementation of Standardized Outcomes Measures in Practice Most frequent uses of information from standardized outcome measures were quality assurance, communicating with other health care providers, and determining progress or outcomes of individual patients (Tab. 4). Of the participants who used standardized outcome measures, 35.1% responded that they were required for all patients in their setting, and 23.8% responded that they were routinely used for all patients but not mandated. The most common means of collecting data and analyzing outcome was to have patients complete paper forms followed by therapists’ review of the raw information (80.6%). That is, the therapists did not necessarily have access to scores from the measurement tool when seeing the patient and used only their qualitative assessment of the responses. Participants were asked to list the measures that they used in their practices and to indicate whether the measures were “home grown.” The most frequently listed measures were: Oswestry Low Back Disability Index (ODI) (41.3%); facility “homegrown” measures (22%); Lower Extremity Functional Scale (LEFS) (18.8%); Disabilities of the Arm, Shoulder, and Hand (DASH) (18.3%); and Berg Balance Scale (BBS) February 2009

Standardized Outcome Measures in Physical Therapist Practice Table 1. Participant Characteristics (N⫽456)a 95% CI Variable

Percentage

Lower Bound

Upper Bound

N

Sex

(1 missing)

Male

31.9

26.9

36.9

145

34.7

Female

68.1

63.1

73.1

310

65.3

⬍3

13.7

10.4

17.0

62

3–5

11.3

8.1

14.3

51

6–10

17.9

14.1

21.7

81

Years of physical therapist practice

(4 missing) 11.1 (⬍4 y) 6.5 (4–5 y) 17.2

11–20

24.8

20.4

29.2

112

27.1

⬎20

32.3

27.3

37.3

146

38.5

4.4

2.4

6.4

20

6.9

Professional (entry-level) degree Certificate

(1 missing)

Baccalaureate

42.2

36.7

47.7

192

48.8

Master’s

40.9

35.5

46.3

186

35.6

Doctorate

12.5

9.4

15.6

57

8.1

Highest degree

(3 missing)

Professional

72.2

66.6

77.8

327

Advanced master’s

13.2

9.9

16.5

60

Transitional DPT

10.6

7.7

13.5

48

8.9

4.0

2.2

5.8

18

4

68.0

61.9

74.1

Cardiovascular-pulmonary

0.5

⫺0.1

1.1

2

Geriatric

3.9

2.1

5.7

17

Doctorate Specialty (could have more than 1) None

Neurology

Unable to determine Cannot distinguish from professional degree

(35 missing) 296

1.6

0.8

2.4

7

11.5

8.2

14.8

50

Pediatric

2.5

⫺0.7

5.7

11

Sports

1.8

0.6

3.0

8

Manual therapy

5.5

3.3

7.7

24

Hand therapy

1.1

0.1

2.1

5

Other

3.7

1.9

5.5

15

Orthopaedic

a

National Data20 (%)

CI⫽confidence interval, DPT⫽Doctor of Physical Therapy.

(17.9%). The eAppendix 2 (available at http://www.ptjournal.org) comprises a list of all measures listed by the participants. The most frequent reasons for choosing specific standardized outcome measures were: they could be completed quickly (68.7%), they were easy for patients to understand (68.2%), and they had

February 2009

been shown to be valid and reliable (64%). Fifty-two percent of participants indicated they did not use standardized outcome measures in practice, and 49% of them indicated that they did not plan to implement their use in future. The 3 most common reasons

for not using standardized outcome measures were: they are too time consuming for patients to complete (43%); they are too time consuming for clinicians to analyze, calculate, and score (30%); and they are too difficult for patients to complete independently (29.1%) (Tab. 5).

Volume 89

Number 2

Physical Therapy f

129

Standardized Outcome Measures in Physical Therapist Practice Table 2. Practice Characteristics (N⫽456)a 95% CI Variable

Percentage

Lower Bound

Upper Bound

Region

National Data20 (%)

(26 missing)

Northeast

19.1

15.0

23.2

82

Midwest

25.3

20.7

29.9

109

South

31.6

26.5

36.7

136

West

23.5

19.1

27.0

101

0.5

⫺0.1

1.1

Guam and Virgin Islands Type of work facility

2 (21 missing)

Acute care Inpatient rehabilitation (including subacute care) Extended care

15.4

11.8

19.0

67

13.1

6.0

3.8

8.2

26

3.5 5.6 (including SNF)

3.0

1.4

4.6

13

61.2

55.2

67.0

266

Home health

7.8

3.1

12.5

34

7.9

School system

3.4

1.6

5.2

15

4.1

Other

3.2

1.6

4.8

14

9.8

Outpatient

Age, y (majority of patients)

56

(6 missing)

No majority

70.0

64.2

75.8

315

⬍21

8.7

6.0

11.4

39

21–40

1.8

0.6

3.0

8

41–60

7.1

4.7

9.5

32

61–75

5.1

3.1

7.1

23

⬎75

7.3

4.9

9.7

33

No majority

30.7

25.8

35.8

139

Musculoskeletal

56.1

50.5

62.1

254

Neuromuscular

6.4

4.2

8.6

29

Cardiovascular-pulmonary

1.5

0.3

2.7

7

Women’s health

0.4

⫺0.2

1.0

2

Integumentary

0.4

⫺0.2

1.0

2

Other

4.4

2.4

6.4

20

Conditions (majority of patients)

(3 missing)

X Treatment sessions per 8-h day a

N

10.9

95% CI 10.5

11.3

CI⫽confidence interval, SNF⫽skilled nursing facility.

Odds of Using Standardized Outcome Measures The type of facility in which the participant practiced, whether or not the participant had a clinical specialty certification, and the age of the majority of patients managed in the practice were related to the likelihood of using standardized outcome measures. Compared with physical 130

f

Physical Therapy

Volume 89

therapists working in acute care settings, those working in outpatient settings were nearly 7 times more likely to use standardized outcomes measures and those working in home care settings were approximately 12 times more likely to use standardized outcome measures. Participants with a clinical specialty were nearly 2 times more likely to

Number 2

use standardized outcome measures than those who did not have a specialty (Tab. 6).

Discussion More than 50% of the respondents in this study reported that they did not use standardized outcome measures, and only a small proportion of those indicated that they intended to use February 2009

Standardized Outcome Measures in Physical Therapist Practice them in the future. There are no comparable data reported from similar samples to assist in the interpretation of this number. The use of standardized measures was variable across settings, and the greater likelihood of use in the outpatient and home health care settings compared with the acute care setting was not surprising. Abrams et al16 reported that among physical therapists who participated in their survey, with most managing a majority of patients with orthopedic conditions, usage of standardized outcome measures was fairly high. In the home health care setting, the Outcome Assessment and Information Set (OASIS) is mandated. Huijbregts et al7 reported the perception that it would be difficult to find suitable measures for patients who might have fluctuating conditions, such as those in intensive care units. Hanekom et al,21 in a 2007 systematic review of outcomes measures used by physical therapists in intensive care units, reported that only one case study measured function using the modified Borg scale. No other functional measures or measures of health-related quality of life were found as outcome measures in any of the studies they reviewed. In our study, the finding that 3 of the most frequently used measures are useful in orthopedic conditions is not surprising given the fact that a majority of the participants practiced in outpatient settings and approximately 11% had orthopedic clinical specialty certification. Among Australian physical therapists who managed mostly patients with orthopedic conditions, Abrams et al16 found a relatively high use of the ODI; approximately 50% of the therapists indicated that they used the ODI frequently or always. Measures specific to other body regions were used less frequently, but the authors did not indicate the specific names of most of those measures. Haigh et al6 reported that the ODI was used in February 2009

Table 3. Perceived Benefits and Problems Among Physical Therapists Who Used Standardized Outcome Measures (n⫽218) N

Percentage

Enhance communication with patient

206

94.5

Help direct the plan of care

204

93.6

Perceived benefits

Enhances communication with payers

190

87.2

Enhance thoroughness of physical therapist examination

190

87.2

Improve patient outcomes

184

84.4

Help focus the intervention

182

83.5

Helps to motivate patient

172

78.9

Enhance efficiency of physical therapist examination

170

78.0

Help to decrease insurance denials

150

68.8

Enhance marketing of practice

115

52.8

11

5.0

Other Perceived problems Confusing to patients

174

79.8

Difficult for patients to complete

166

76.1

Take too much time for patients

164

75.2

Often are not completed at discharge, so cannot give information about response to treatment

144

66.1

Take too much of clinicians’ time

113

51.8

Make patients/clients anxious

110

50.5

Are difficult to interpret

102

46.8

Are not culturally sensitive

100

45.9

Require too high a reading level

97

44.5

Provide information that is too subjective

87

39.9

Do not help to direct the plan of care

71

32.6

Items are not relevant for my patients

71

32.6

Require more effort than they are worth

67

30.7

English is a language in which many of my patients/clients are not fluent

58

26.6

only approximately 4% of assessments done for patients with low back pain across 418 rehabilitation centers in Europe in 1998. Torenbeek et al14 indicated that the ODI was used in rehabilitation facilities in 4 out of 5 European countries; they reported the highest use in Ireland (12.5% of facilities). The date of their survey was not reported. The ODI is available in the public domain, and the ISI Web of Science citation index22 identifies 1,035 citations of the article in which it was originally reported in 1980.23 The ISI Web of

Science citation index also indicates that the articles in which the LEFS and DASH were originally reported have been cited 74 and 431 times since their original publications in 1999 and 1996, respectively.24,25 These data suggest that the measures are fairly well known, at least among those publishing articles in scientific journals. Many standardized outcome measures have been developed within the last decade or so, and this timing may explain why the participants who had been practicing for more than 20 years were

Volume 89

Number 2

Physical Therapy f

131

Standardized Outcome Measures in Physical Therapist Practice Table 4. Uses of Information Among Physical Therapists Who Used Standardized Outcome Measures (n⫽218) Use Quality assurance

N

Percentage

173

79.4

Communicating with other health care providers

167

76.6

Determining progress/outcomes of individual patients

163

74.8

Determining average patient improvement to examine practice effectiveness

154

70.6

Determining average patient improvement to examine clinician effectiveness

125

57.3

Traditional research

113

51.8

Comparing patient outcomes across conditions

108

49.5

Determining case mix

89

40.8

Comparing clinicians’ performances

82

37.6

Comparing clinics’ performances

72

33.0

much more likely than their younger colleagues to learn about them from continuing education workshops and other therapists than from formal, professional education. One surprise is the relatively high use (22%) of “home-grown” measures. Similarly, Kay et al8 reported that 18% of the physical therapists surveyed in their study used departmentally developed instruments. This practice seems unnecessary given the large number of existing measures that cover all body regions and many specific conditions. The finding also is somewhat contradictory, given that 68% of those who used standardized outcome measures indicated that one reason for choosing an instrument was its documented validity and reliability. We also found that participants defined outcome measures broadly to include not only measures of activity and participation but also some measures of body function such as the BBS. This finding is reflected in the literature in that previous reports of use of outcome measures by physical therapists have included references to measures of body function.6 – 8 The problems perceived by physical therapists who used standardized outcomes measures and the reasons 132

f

Physical Therapy

Volume 89

given for not using them among those who did not use them were fairly similar and included issues that have been discussed in the literature for more than a decade.13,15,16 Even in the most recent study,16 the majority of participants indicated lack of familiarity with, lack of training in, and lack of access to measures were barriers. Practitioners in other health care specialties have reported the same types of barriers as those reported by physical therapists. Meadows et al18 reported that 39% of GPs and 28% of nurses indicated having insufficient time to discuss health outcome data with their patients. Logistical problems such as time, additional paperwork, and costs of personnel were cited as the most important reason for not using the measures among psychologists.19 Based on our results, it appears that many physical therapist practices may not yet have determined how best to address these barriers. Twenty-seven percent of the participants in our study who did not use standardized outcome measures cited the lack of a support system in terms of technology and staffing as a reason, and only 11% of those who used the measures indicated that office staff helped patients to complete them. Similarly, more than 10 years

Number 2

ago, Russek et al15 reported that the physical therapists in their study identified lack of personnel to assist in data management as a barrier to implementation of these measures. Kay et al8 reported that approximately 42% of the physical therapists they surveyed in 1998 thought that lack of resources was an obstacle. The study of nurses and GPs indicated that they, too, would be more willing to use standardized measures if the data were collected and analyzed by someone else.18 In our study, approximately 7% of the participants indicated that computers, and not paper, were used for completion and analysis of measures, and slightly fewer than 10% of participants indicated that they chose measures based on their ability to analyze data electronically. Recent literature has suggested that implementation of computerized systems is critical to clinical practice in terms of evaluating both individual patients and overall practice performance. For example, in 1994, Shields et al26 described the development of a computer-based clinical database in the acute care setting and urged its implementation to better measure outcomes of physical therapy interventions. More recently, Jette et al,27 reporting on a new standardized outcome measure that uses a computerized adaptive testing format, suggested that challenges for implementation included assisting clinicians in carrying out the testing as well as understanding and interpreting the data derived from such measures. They stressed the need for training, technical support, and access to software. In our study, 18% of the participants who did not use standardized outcome measures cited the lack of relevance to the plan of care as a reason. Kay et al8 found that 39% of physical therapists surveyed in 1998 thought that outcome measures did February 2009

Standardized Outcome Measures in Physical Therapist Practice not meet the needs of their patients. Researchers reporting the perceptions of nurses, GPs, psychologists, and oncologists also cite lack of clinical relevance as a barrier to use of standardized outcome measures. For example, Hatfield and Ogles19 reported that a substantial number of psychologists felt that standardized outcome measures could “distort” the effects of treatment. General practitioners and nurses stated that they were more likely to use standardized outcomes measures if they helped in the care of the individual patient,18 and oncologists indicated that informal collection of data seemed a better way to understand individual patient needs than using standardized outcome measures.17 Among the physical therapists in our study who used standardized outcome measures, however, the majority believed that these measures could aid in directing the plan of care and enhancing the thoroughness of their examinations. Similarly, previous studies7,14 have shown that physical therapists perceived planning of care and monitoring the effects of treatment as benefits of standardized outcome measures. Although it is likely that many physical therapists are similar to other health care practitioners in valuing and applying the qualitative information gathered from patients, differences in perceptions regarding the usefulness of standardized outcome measures may be due to the fact that physical therapists have better tools for measuring the constructs that provide a basis for evaluating the effectiveness of their care. Limitations One limitation of our study is that our data reflect what has been reported by physical therapists rather than what has been observed, and although we provided our participants with a definition of standardized outcome measures, they may have thought about the measures February 2009

Table 5. Reasons Among Participants Who Did Not Use Standardized Outcome Measures (n⫽238, May Indicate More Than 1 Reason) Reason Take too much time for patients/clients to complete

N

Percentage

102

43.0

Take too much of clinicians’ time to analyze/calculate/score

71

30.0

Are difficult for patients/clients to complete independently

69

29.1

Require a support system that I do not have (eg, technology, staffing)

64

27.0

Often are not completed at discharge, so are not useful in determining patients’/clients’ response to treatment

58

24.5

Do not contain the types of items or questions that are relevant for the types of patients/clients who I see

57

24.1

Other reason

54

21.2

Are confusing for patients/clients

48

20.3

Require more effort than they are worth

47

19.8

Do not contain information that helps direct the plan of care

43

18.1

Are difficult to interpret (eg, do not know what norms are, how score relates to severity, or what a clinically important change might be)

40

16.9

Require too high a reading level for my patients/clients

27

11.4

Make patients/clients anxious

22

9.3

Provide information that is too subjective to be useful

22

9.3

Require training that I do not have

18

7.6

Are in English, a language in which many of my patients/clients are not fluent

16

6.8

Are not sensitive to the cultural/ethnic concerns of many patients/clients

10

4.2

Cost too much

7

3.0

Are really only useful for research purposes

7

3.0

Are not relevant because my practice involves consultation, case management, or discharge planning only

6

2.5

Plan to implement? No

110

49.3

Maybe

93

41.4

Yes

20

9.0

they used in different ways. Additionally, the validity and test-retest reliability of our survey data were not tested. We attempted, however, to demonstrate content validity through use of previous literature on the topic and construct validity through factor analysis. There was good internal consistency within the items assessing the perceived benefits and barriers to using outcome measures. Another limitation was that we sent survey questionnaires only to members of APTA. Therefore, the results of this study may be biased and not representative of the

entire profession of physical therapy. Given that APTA members may be more likely than nonmembers to attend national meetings, they may be more likely to have been exposed to issues related to measuring outcomes. Therefore, we might speculate that those who are members would be more likely than nonmembers to use standardized outcome measures. We considered our response rate to be adequate in that it was comparable to that reported in similar studies; however, there is the possibility that the sample was biased.

Volume 89

Number 2

Physical Therapy f

133

Standardized Outcome Measures in Physical Therapist Practice Table 6. Odds of Using Standardized Outcome Measures by Participant and Practice Characteristicsa 95% CI Factor

Odds Ratio

Lower

Upper

Percentage Using

Facility Acute care

Reference

16.4

Inpatient rehabilitation

2.63

0.80

8.67

30.8

Extended care

2.21

0.47

10.30

23.1

Outpatient

6.80

2.99

15.48

60.5

Home care

12.56

4.36

36.18

64.7

School system

1.04

0.10

11.07

6.7

Other type of facility

5.46

1.27

23.43

50.0

1.03

2.88

Specialty No

Reference

Yes

1.72

43.2 59.2

Age of majority (⬎50%) of patients, y ⬍21

a

Reference

17.9

21–60

3.58

1.16

11.05

57.9

⬎60

2.42

0.74

7.92

35.0

No majority

6.57

2.03

21.31

59.1

CI⫽confidence interval.

Implications Despite more than a decade of development and testing of measures appropriate for various conditions and practice settings, the physical therapy profession appears to have some distance to go in implementing standardized outcome measurement routinely in most clinical settings. The development of such measures for acute care settings may need to be a particular focus. Regardless of setting, practices will need to help clinicians to manage time so that collection of data can become routine despite productivity expectations. Given the perceived time-consuming nature of standardized outcome measurement, investment in computerized systems for quick data entry and analysis may be warranted. Although the content, properties, and applicability of many standardized outcome measures have been reported in the literature for more than a decade, clinicians continue to 134

f

Physical Therapy

Volume 89

report that the measures are not used because they are not applicable to their patients or that they cannot interpret the scores. It appears, therefore, that disseminating information through the professional literature may not be an efficient or effective mechanism. Further instruction and enculturation through continuing education as well as professional and graduate professional education may increase the use of standardized outcome measures. Education should include the use of hardware and software to facilitate their usage. In addition, software should be made readily available to provide analyses that assist in the interpretation of scores. Interpretation could include comparing patients’ scores with norms; using scores to qualify severity of condition or predict outcome or duration of an episode of care; or categorizing changes in scores as worse, stable, or improved. Such data could assist physical therapists in making decisions

Number 2

about change in management strategies, referral, or discharge from services. As noted by Jette et al,27 the essential strategies to improve use of standardized outcome measures may well require new funding mechanisms. Given that many of our participants believed that standardized outcome measures are confusing and difficult for patients to complete, efforts should be made to ensure readability and interpretability by patients. Reading level, font size, and general appearance of measurement tools need to be considered. Language and cultural concerns were cited by relatively few of our participants; however, given the changing nature of the US population, these concerns may become magnified and necessitate adaptations to the commonly used instruments.

February 2009

Standardized Outcome Measures in Physical Therapist Practice

Conclusion Most participants in our study did not use standardized outcome measures. There was a relationship between specialty certification, age of patients seen, and practice setting and likelihood of using standardized outcome measures. Most participants who used standardized outcome measures perceived that their use enhanced communication with patients and helped to direct the plan of care. More than 70% of the participants, however, felt that these tools could be confusing, difficult, and time consuming for patients. It appears that outcome measures are used largely without computerized systems for either administration or analysis. Participants with fewer years in practice reported learning to use standardized outcome measures in their professional education programs. Focus on education in the use of outcome measures in professional education programs and continuing education, as well as alterations in practice management strategies, may lead to increasing use of standardized outcome measures in the future. Dr Jette provided concept/idea/research design, data analysis, and project management. All authors provided writing. Mr Halbert, Ms Iverson, Ms Miceli, and Ms Shah provided data collection. The study was approved by the Institutional Review Board of the University of Vermont. This article was received August 3, 2008, and was accepted October 30, 2008. DOI: 10.2522/ptj.20080234

References 1 Deyo RA, Carter WB. Strategies for improving and expanding the application of health status measures in clinical settings. Med Care. 1992;30:MS176 –MS186.

February 2009

2 Deyo RA, Patrick DL. Barriers to the use of health status measures in clinical investigation, patient care, and policy research. Med Care. 1989;27:S254 –S268. 3 Garland AF, Kruse M, Aarons GA. Clinicians and outcome measurement: What’s the use? J Behav Health Serv Res. 2003;30:393– 405. 4 Lansky D, Butler JBV, Waller FT. Using health status measures in the hospital setting: from acute care to outcomes management. Med Care. 1992;30:MS57–MS73. 5 Nelson EC, Berwick DM. The measurement of health status in clinical practice. Med Care. 1989;27:S77–S90. 6 Haigh R, Tennant A, Biering-Sorensen F, et al. The use of outcome measures in physical medicine and rehabilitation within Europe. J Rehabil Med. 2001;33: 273–278. 7 Huijbregts MPJ, Myers AM, Kay TM, Gavin TS. Systematic outcome measurement in clinical practice: challenges experienced by physiotherapists. Physiother Can. Winter 2002:25–31, 36. 8 Kay TM, Myers AM, Huijbregts MPJ. How far have we come since 1992? a comparative survey of physiotherapists’ use of outcome measures. Physiother Can. Fall 2001:268 –275. 9 Mayo N, Cole B, Dowler J, et al. Use of outcome measurement in physiotherapy: survey of current practice. Can J Rehab. 1993;7:81– 82. 10 Thier S. Forces motivating the use of health status assessment measures in clinical settings and related clinical research. Med Care. 1992;30:MS15–MS22. 11 Kramer AM, Holthaus D. Uniform Patient Assessment for Post-Acute Care. 2006. Available at: http://www.cms.hhs.gov/Quality InitiativesGenInfo/downloads/QualityPAC FullReport.pdf. Accessed August 1, 2008. 12 Commission on Accreditation in Physical Therapy Education. Evaluative Criteria for Accreditation of Education Programs for the Preparation of Physical Therapists. Alexandria, VA: American Physical Therapy Association; 2006. 13 Turner-Stokes L, Turner-Stokes T. The use of standardized outcome measures in rehabilitation centres in the UK. Clin Rehabil. 1997;11:306 –313. 14 Torenbeek M, Caulfield B, Garrett M, Van Harten W. Current use of outcome measures for stroke and low back pain rehabilitation in five European countries: first results of the ACROSS project. Int J Rehabil Res. 2001;24:95–101. 15 Russek L, Wooden M, Ekedahl S, Bush A. Attitudes toward standardized data collection. Phys Ther. 1997;77:714 –729.

16 Abrams D, Davidson M, Harrick J, et al. Monitoring the change: current trends in outcome measure usage in physiotherapy. Man Ther. 2006;11:46 –53. 17 Taylor KM, Macdonald KG, Bezjak A, et al. Physicians’ perspective on quality of life: an exploratory study of oncologists. Qual Life Res. 1996;5:5–14. 18 Meadows KA, Rogers D, Greene T. Attitudes to the use of health outcome questionnaires in the routine care of patients with diabetes: a survey of general practitioners and practice nurses. Br J Gen Pract. 1998;48:1555–1559. 19 Hatfield DR, Ogles BM. Why some clinicians use outcomes measures and others do not. Adm Policy Ment Health Ment Health Serv Res. 2007;34:283–291. 20 PT Demographics. Available at: http:// www.apta.org/AM/Template.cfm?Section⫽ Demographics&Template⫽/TaggedPage/ TaggedPageDisplay.cfm&TPLID⫽101& ContentID⫽14332. Accessed August 1, 2008. 21 Hanekom SD, Faure M, Coetzee A. Outcomes research in the ICU: an aid in defining the role of physiotherapy. Physiother Theory Pract. 2007;23:125–135. 22 ISI Web of Science. Available at: http://apps. isiknowledge.com/WOS_GeneralSearch_ input.do?highlighted_tab⫽WOS&product⫽ WOS&last_prod⫽WOS&SID⫽3FjiaaoCe9 FEile2c4k&search_mode⫽GeneralSearch. Accessed June 30, 2008. 23 Fairbank CT, Couper J, Davies JB, O’Brien JP. The Oswestry low back pain disability questionnaire. Physiotherapy. 1980;66: 271–273. 24 Binkley JM, Stratford PW, Lott SA, Riddle DL. The Lower Extremity Functional Scale (LEFS): scale development, measurement properties, and clinical application. Phys Ther. 1999;79:371–383. 25 Hudak PL, Amadio PC, Bombardier C. Development of an upper extremity outcome measure: the DASH (Disabilities of the arm, Shoulder and Hand). Am J Ind Med. 1996;29:602– 608. 26 Shields RK, Leo KC, Miller B, et al. An acute care physical therapy practice database for outcomes research. Phys Ther. 1994;74:463– 470. 27 Jette AM, Haley SM, Tao W, et al. Prospective evaluation of the AM-PAC CAT in outpatient rehabilitation settings. Phys Ther. 2007;87:385–398.

Volume 89

Number 2

Physical Therapy f

135

Research Report Gastrocnemius-Soleus Muscle Tendon Unit Changes Over the First 12 Weeks of Adjusted Age in Infants Born Preterm Marybeth Grant-Beuttler, Robert J Palisano, Debra P Miller, Barbara Reddien Wagner, Carolyn B Heriza, Patricia A Shewokis M Grant-Beuttler, PT, PhD, PCS, is Assistant Professor, Department of Physical Therapy, Chapman University, One University Dr, Orange, CA 92866 (USA). Address all correspondence to Dr Grant-Beuttler at: beuttler@ chapman.edu. RJ Palisano, PT, ScD, is Professor, Department of Physical Therapy and Rehabilitation Sciences, Drexel University, Philadelphia, Pennsylvania. DP Miller, PT, DPT, is Assistant Director of Clinical Education, Department of Physical Therapy, University of Scranton, Scranton, Pennsylvania. B Reddien Wagner, PT, DPT, MHA, is Director of Clinical Education, Department of Physical Therapy, University of Scranton. CB Heriza, PT, EdD, FAPTA, is Professor, Doctor of Science Program, Pediatric Therapy, Rocky Mountain University of Health Professions, Provo, Utah.

Background and Purpose. Differences in the gastrocnemius-soleus muscle and tendon have been documented shortly after birth in infants born preterm compared with infants born at term. Knowledge of muscle tendon unit lengths at term age to 12 weeks of age in infants born preterm may be useful in understanding motor development. Participants and Method. Gastrocnemius-soleus muscle tendon unit lengths were compared at term age, at 6 weeks of age, and at 12 weeks of age (preterm adjusted age) in 20 infants born full term and 22 infants born preterm.

Results. Significant differences were found between the 2 groups on taut tendon, relaxed muscle length (AO); taut tendon, stretched muscle length (AMax); and muscle stretch (AO to AMax). Infants born preterm demonstrated measures of AO and AMax in positions of greater plantar flexion compared with infants born full term. Significant differences in measurements of AO were found between term age and 12 weeks of age, indicating that the tendon lengthens during this period for both groups. Discussion and Conclusion. These results provide knowledge of musculoskeletal development of the gastrocnemius-soleus muscle and tendon. Differences in musculoskeletal measurements are consistent with uterine confinement in the last weeks of full-term gestation. These findings have implications when examining the musculoskeletal system in infants born preterm who are demonstrating functional changes.

PA Shewokis, PhD, is Associate Professor, College of Nursing and Health Professions and the School of Biomedical Engineering, Science and Health Systems, Drexel University.

For an Invited Commentary and the Author Response, visit www.ptjournal.org.

[Grant-Beuttler M, Palisano RJ, Miller DP, et al. Gastrocnemiussoleus muscle tendon unit changes over the first 12 weeks of adjusted age in infants born preterm. Phys Ther. 2009;89:136 –148.] © 2009 American Physical Therapy Association Post a Rapid Response or find The Bottom Line: www.ptjournal.org 136

f

Physical Therapy

Volume 89

Number 2

February 2009

Muscle and Tendon Changes in Infants Born Preterm

M

otor development in infants traditionally has focused on maturation of the central nervous system, and less focus has been placed on posture of the infant and examination of the musculoskeletal system. Posture in infants born preterm, both those at low risk and those at high risk for developmental issues, is generally characterized by more extension1,2 and, specifically, by increased trunk extension,3 decreased elevation of the hips in a prone position,3–5 and lateral (external) rotation of the hips3–5 compared with infants born full term. Prolonged positioning in extension has the potential to alter the musculoskeletal system over the long term.6 – 8 Davis and colleagues9 found increased lateral rotation range of motion and increased foot angle during gait in children 3 to 4.5 years of age who had been born preterm and defined as either low risk or high risk. These authors suggested that these changes in range of motion at the hip and in gait are related to a “frog-leg” positioning in the neonatal intensive care unit (NICU).9 Heriza10 examined supine kicking and reported more dorsiflexion at peak leg flexion in infants born full term compared with infants at low risk born preterm at 40 weeks of gestation. Although the ankle difference was greater than 25 degrees, the statistical power of this analysis was low and differences did not reach significance in that study. The degree of ankle motion during supine kicking found in infants born preterm may potentially be explained by changes in musculoskeletal system development at the ankle. Posture and positioning observed at birth have been reported to gradually change as infants at low risk and infants at high risk born preterm approach term age (40 weeks of gestation).3,4,11 Palmer and colleagues2 found that differences in arm recoil, arm traction, and leg recoil, which February 2009

were significantly different at birth between infants born preterm and infants born full term, were no longer significant when the infants born preterm reached term age. By term age, infants born preterm demonstrated increased flexion in their posture. These changes in posture were found to be strongest in infants born preterm with higher birth weights, fewer medical interventions, and later gestational age at birth.2 Harris and colleagues11 suggested that, over the first year, ankle dorsiflexion decreases a mean of 10 degrees in infants with very low birth weight born before 35 weeks of gestation. Their findings of increased extension at the ankle were consistent with those of Desmond et al12 and Grenier,13 who suggested that extension increases during the period between birth and term age due to positioning in extension and the force of gravity on the joints. Infants born full term also demonstrated a decrease in dorsiflexion range of motion after birth, perhaps as a consequence of gravity.2,14,15 Hoffer14 suggested that these changes in range of motion at the ankle in infants born full term were complete by 3 months of age. Specific changes in the muscle tendon unit (MTU) as a result of positioning could depend on the age of the infants when positioning occurs and not necessarily where these changes occur, in utero or out of utero. Tardieu et al6 and Williams and Goldspink7 documented structural changes in the MTU when prolonged shortening or lengthening occurred differently in very young versus adult animals.6,7 In very young rabbits6 and mice,7 tendon lengthening occurred when a muscle was immobilized (18 –21 days) in either a lengthened or shortened position, whereas the number of sacromeres decreased in both scenarios.6,7 These findings were in contrast to those of

adolescent or adult animals that underwent decreased sarcomere number with prolonged immobilization in a shortened position and increased sarcomere number with immobilization in a lengthened position. During the last 4 weeks (28 days) of gestation, the human fetus is confined in a tight uterine space, which decreases the amount of fetal movement and promotes dorsiflexion, putting the gastrocnemiussoleus muscle in a position of prolonged lengthening.16,17 Tardieu et al18 developed a reliable and valid method of measuring MTU length and muscle belly stretch in the gastrocnemius-soleus muscle that has been applied to infants born preterm and infants born full term.19 Their method of measuring MTU length requires 2 goniometric measures. The first measure, AO, requires palpation of the Achilles tendon while measuring the ankle range with the goniometer.18 The measurement is recorded at the first point of maximal Achilles tendon tension while the muscle is relaxed. This measure has been suggested by Tardieu, et al18 to be the point of full tendon lengthening with a relaxed muscle belly. AMax is a measure of full tendon and muscle belly lengthening. This measurement is taken when the muscle and tendon are moved to maximal stretch, the same way a physical therapist would measure maximum dorsiflexion at the ankle.18 Grant-Beuttler et al19 found interrater reliability (intraclass correlation coefficient [ICC (2,2)]) of .95 to .86 for AMax and .96 to .97 for AO and test-retest reliability (ICC [3, 2]) of .91 to .97 for AMax and .97 to .98 for AO in infants born preterm and infants born full term. In previous research using these methods of measuring MTU length, Grant-Beuttler and Shewokis20 found that when infants born preterm were examined soon after birth, they demonstrated less maximum dorsiflexion

Volume 89

Number 2

Physical Therapy f

137

Muscle and Tendon Changes in Infants Born Preterm compared with infants born full term. Twenty infants at low risk born between 26 and 36 weeks of gestation demonstrated a shorter taut tendon, relaxed muscle length (AO) following birth compared with 21 infants born full term measured within 48 hours of birth. Although differences in both AO and AMax reached significance, muscle belly stretch (AO to AMax) did not reach significance. These results suggest that there is a difference in the gastrocnemius-soleus MTU length between infants born preterm and infants born full term. However, MTU length and muscle belly stretch have not been measured over time to determine whether differences found immediately after birth persist at term age, at 6 weeks of age, or at 12 weeks of age, when the fullterm infant’s ankle range of motion stabilizes. Documenting changes in preterm infants at low risk is especially important secondary to fewer medical issues occurring in these infants and documentation of motor issues occurring at later ages of development. Fewer medical complications and medical interventions increase the likelihood that motor differences are due to preterm birth. In addition, when motor issues are documented, they are found after specific skills have developed in an infant born full term and may not be detected with simple observation.21–25 Detecting subtle motor changes is challenging and requires sophisticated systems used in laboratories.21–25 Subtle motor issues in infants at low risk are more likely to be missed during a standard physical therapy examination and evaluation, especially early in motor development before skills have developed. Finding a relevant assessment method to detect changes early will increase the likelihood that physical therapy will identify and intervene with subtle issues that may persist. 138

f

Physical Therapy

Volume 89

Examination and evaluation of the gastrocnemius-soleus MTU may be important to detect differences that could alter development of a lowerextremity coordination pattern. During supine kicking, a coordinated lower-extremity movement pattern is learned through repetition.10,25,26 This coordinated movement pattern will include increased plantar flexion if the MTU is shortened. As documented by Heriza,10 an infant at low risk born preterm may be repeating more plantar flexion during kicking. When the infant is developing a new skill, such as walking, this altered movement pattern may be used.27 An altered ankle movement pattern is consistent with the increased frequency of toe-touch foot contact during walking, as documented by Cioni and colleagues.28 The relationship between decreased dorsiflexion and persistent toe-walking in infants born preterm also is suggested in the work of Georgieff and associates.29 Evaluating the MTU may be useful to physical therapists for predicting which infants born preterm are at risk for altered ankle movements during lower-extremity kicking and active movement during walking. The purpose of this study was to compare passive gastrocnemius-soleus MTU measurements at the ankle obtained at birth and at 6 weeks and 12 weeks of age between infants born preterm and infants born full term to identify: (1) whether differences in MTU length and muscle belly stretch exist between infants born preterm and infants born full term; (2) whether differences are present in the tendon or the muscle belly, or both; and (3) at what ages changes in the tendon length or the muscle belly, or both, are found. Based on pilot work by Grant-Beuttler and Shewokis,20 we hypothesized that differences in the gastrocnemiussoleus MTU would be observed at term age. Whether differences in the MTU between the infants born pre-

Number 2

term and the infants born full term are present at 6 weeks of age and at 12 weeks of age is less clear. Ferrari et al30 presented evidence that by 12 weeks of age, infants at low risk born preterm without neurological impairment demonstrate extremity movements that are similar to movements observed in infants born full term. These observations were based on “general movement assessment,” and specific lower-limb kinematics were not measured quantitatively. However, differences in ankle kinematics and the delay in age of walking that have been reported9,30 –32 suggest that differences in gastrocnemius-soleus MTU length and muscle belly stretch may be present at 12 weeks after term age.

Method Participants Sample size estimates were made prospectively based on a significance criterion of .05, a power of .80, and the lowest meaningful effect size (d) of 1.59 for AO and 1.63 for AMax using G-power (version 2.0) statistical software.33 Effect size estimates for measures of muscle length were based on data from the pilot study by Grant-Beuttler and Shewokis.20 A sample of 40 infants, with 20 infants born preterm and 20 infants born full term, was required for this study’s repeated-measures, longitudinal design. All infants were recruited to participate while in a hospital setting. Families of infants born full term were recruited from 3 local hospitals. Families of infants born preterm were recruited from NICUs in 2 of the 3 local hospitals. The third local hospital did not have an NICU. Forty-two infants were included in this study. Twenty infants born following full-term gestation of 38 to 41 weeks were included in a fullterm group. Twenty-two infants born following premature birth at 26 to 36 February 2009

Muscle and Tendon Changes in Infants Born Preterm weeks of gestation were included in a preterm group. Data collection of infants from preterm births was larger than the original 20 infants secondary to inclusion of infants from multiple births. All infants had normal newborn examinations, with no orthopedic, genetic, or neurological impairments. Gestational age, birth weight, sex, and method of delivery for both groups of infants are described in Table 1. Mean gestational age for the infants born full term was 39 weeks 3 days (SD⫽6.44 days, range⫽38 weeks 2 days– 41 weeks 0 days), and mean birth weight was 3,342 g (SD⫽456.21, range⫽2,750 – 4,720). Eleven of the infants born full term were male, and 9 were female. Five of the infants born full term were born via cesarian section, and 15 were born via normal vaginal delivery. All infants in the full-term group were singleton pregnancies. Mean gestational age for the infants born preterm was 32 weeks 3 days (SD⫽1 week 5.43 days, range⫽26 weeks 3 days–34 weeks 5 days), and mean birth weight was 1,854 g (SD⫽463.14, range⫽794 –2,608). Fourteen of the infants born preterm were male, and 8 were female. Fourteen of the infants born preterm were born via cesarian section, and 8 were born by vaginal delivery. Eleven infants born preterm were singleton pregnancies, 3 infants born preterm were a triplet pregnancy, and 8 infants born preterm were twin pregnancies. The infants in the preterm group were eligible to participate in the study if the neonatologist classified them as low risk for a movement disorder. Classification of low risk included limited time on ventilator support, no evidence of orthopedic or genetic impairment, no evidence of neurological complications, and a benign medical course during their stay in the NICU. All infants born preterm were discharged February 2009

with no indication of continued medical problems. Infants born full term were recruited within 2 days of birth. Infants born preterm were recruited prior to discharge from the NICU. Posters advertising the project were present in both NICUs and the maternity areas in the 3 hospitals. Nurses in all of the areas were aware of the project and provided with letters for parents that described the project. Of the 120 families of infants born full term that were contacted by the researchers, 21 families agreed to participate. The family of one infant born full term dropped out prior to data collection. All other infants born full term were measured 3 times, with no attrition in the group. Of the 33 families of infants born preterm that were contacted by the researchers, 18 families (23 infants, 5 multiple-birth groups) agreed to participate. One family could not be contacted to schedule the first session. The remaining 22 infants in the preterm group were measured 3 times, with no attrition in the group. Instrumentation A 10.16-cm-diameter (4-in-diameter) goniometer* was used to measure ankle range of motion. The goniometer had line designations for each degree. Dorsiflexion measurements were documented as positive angles, and plantar-flexion measurements were documented as negative angles, with a 90-degree angle between the tibia and the lateral foot designated as a neutral position, or 0 degrees. Procedure Informed consent was obtained from a parent prior to data collection, which occurred in the Physical Therapy Movement Laboratory at the Uni-

versity of Scranton. Each infant in the full-term group was scheduled for data collection within 7 days after birth, at 6 weeks, and at 12 weeks. Each infant in the pre-term group was scheduled for data collection within 7 days of term age, at 6 weeks adjusted age, and at 12 weeks adjusted age. Families were offered a ride with a professional driving service to and from the laboratory for all 3 visits. In order to accurately measure gastrocnemius-soleus muscle length and muscle belly stretch, the infant’s gastrocnemius-soleus MTU must be relaxed. If the infant was asleep when he or she came to the laboratory, all attempts were made to allow the infant to continue to sleep. If the infant was not asleep, all attempts were made to relax the infant. In an attempt to relax the infant, a parent was asked to cuddle or swaddle him or her in a blanket. In addition, the laboratory was kept warm with a small room heater. All infants were measured when in behavioral state 1 (deep sleep), state 2 (light sleep), or state 4 (quiet, alert)34,35 because they would most likely have relaxed limbs and less movement when in these states. If any tension was palpated in the limb during measurement, the foot was moved into and out of dorsiflexion and plantar flexion, and measurements were delayed until the infant relaxed. All infant states demonstrated during the measure of muscle length are reported in Table 2 for both groups. Infants were not measured in behavioral state 3 (drowsy), 5 (active, alert), or 6 (crying).34,35 If an infant moved into an undesired state, the mother or the researchers assisted the infant in moving into a desired state. No infants were removed from this study because of an inability to get them into or keep them in the desired behavioral state.

* Pro-Med Products, 6445 Powers Ferry Rd, #199, Atlanta, GA 30339.

Volume 89

Number 2

Physical Therapy f

139

Muscle and Tendon Changes in Infants Born Preterm Table 1. Sex, Gestational Age, Method of Delivery, and Birth Weight for Infants Born Full Term and Infants Born Preterma Participant

Sex

Gestational Age (wk)

Method of Delivery

Birth Weight (g)

FT-01

M

38 3/7

NVD

3,572

FT-02

M

40

NVD

3,345

FT-03

F

39

NVD

2,862

FT-04

F

38 5/7

NVD

3,402

FT-05

F

40 2/7

NVD

3,615

FT-06

M

40 1/7

NVD

3,062

FT-07

M

39 2/7

NVD

3,827

FT-08

F

39 5/7

NVD

2,750

FT-09

M

40 3/7

NVD

4,026

FT-10

F

40 4/7

NVD

4,720

FT-11

F

38 3/7

C

4,267

FT-12

M

38 2/7

NVD

3,445

FT-13

F

38 3/7

C

3,260

FT-14

M

40 5/7

NVD

3,856

FT-15

M

40 4/7

NVD

4,082

FT-16

M

38 6/7

C

3,118

FT-17

M

38 3/7

C

3,232

FT-18

F

38 6/7

C

3,230

FT-19

M

39 1/7

NVD

3,175

FT-20

F

41

C

3,827

PT-01b

F

31 3/7

C

1,191

PT-02b

F

31 3/7

C

1,446

b

F

31 3/7

C

PT-04

M

32 4/7

NVD

2,410

PT-05

F

32 1/7

C

2,353

PT-06

M

30 4/7

NVD

1,474

PT-07c

M

33 3/7

NVD

2,410

c

PT-03

M

33 3/7

NVD

2,155

PT-09

M

32 3/7

C

1,616

PT-10

M

34 1/7

NVD

2,155

PT-11

M

26 3/7

NVD

1,049

PT-12c

M

33 5/7

C

2,126

c

M

33 5/7

C

2,296

PT-14c

M

34 1/7

NVD

1,899

PT-15c

M

34 1/7

C

2,041

PT-16

M

31 4/7

C

1,588

PT-17c

M

32 6/7

C

1,928

PT-18c

M

32 6/7

C

1,673

PT-19

F

32 4/7

C

2,070

PT-20

F

31

C

1,219

PT-21

F

34 5/7

NVD

2,608

PT-22

F

33 4/7

C

2,296

PT-08

PT-13

a b c

794

FT⫽full term, PT⫽preterm, M⫽male, F⫽female, NVD⫽normal vaginal delivery, C⫽cesarean section. Triplet birth. Twin birth.

140

f

Physical Therapy

Volume 89

Number 2

February 2009

Muscle and Tendon Changes in Infants Born Preterm Table 2. Behavioral States Exhibited by Infants Born Full Term (n⫽20) and Infants Born Preterm (n⫽22) During Measurement of Muscle Length at Each Age Measureda Term Age Behavioral States

Full-term Group

1

1 (5%)

2

2 (10%)

4 1, 2

12 (60%)

1–3

1 (5%)

6 wk of Age Preterm Group

3 (13.5%)

c

2 (9%)

1 (5%) 15 (75%)

8 (40%)

10 (46%)

3 (15%)

2 (9%)

Preterm Group

16 (73%)

1 (4.5%) 1 (5%)

2 (10%)

3 (13.5%)

1 (4.5%)

2, 4

1 (4.5%)

1 (5%)

2–4b

1 (4.5%)

1 (5%)

4 (18%)

1 (5%)

3, 4

b

Full-term Group

1 (4.5%)

2–5

a

12 wk of Age

Preterm Group

12 (55%)

1, 2, 4 2, 3

Full-term Group

1 (4.5%)

3–5c

1 (5%)

4, 5

1 (5%)

3 (15%)

1 (4.5%)

1 (4.5%) 2 (9%)

2 (10%)

1 (4.5%)

4 (20%)

3 (13.5%)

Number of infants in each group of behavioral states listed, with percentage of infants in parentheses. Primarily in behavioral states 2 and 4. Primarily in behavioral state 4.

As shown in Table 2, infants in the full-term group exhibited more sleep states (states 1 and 2) during the term measurement period and more alert states (states 4 and 5) by the 12-week measurement period. The infants in the preterm group demonstrated similar trends toward sleep states in the term measurement period, and most infants demonstrated alert states in the 12-week measurement period. Infants born preterm demonstrated an ability to regulate their states in a manner similar to that of the infants born full term. This ability of the infants born preterm to maintain and regulate their states is another indication of their physiological stability. No infants in either group demonstrated state 6 (crying), and no infant spent more than brief periods in state 5 (active, alert). Documentation of the states by both groups of infants suggests that measurement could be completed when the infant was in a relaxed state at all ages. February 2009

Each infant’s ankle was passively moved through the available range of motion before taking measurements in an attempt to ensure the infant was relaxed. Ankle range of motion was measured as described by Norkin and White.36 Figures 1 and

2 illustrate measurement of MTU length. The first measurement, AO, is the angle between the foot and the leg when the foot is moved from plantar flexion to dorsiflexion and the palpation of the Achilles tendon reveals the first point of maximum

Figure 1. Measurement of AO muscle length. The AO goniometric measurement is recorded when the Achilles tendon becomes taut. This measure represents slack removed from the tendon, with a relaxed muscle belly.

Volume 89

Number 2

Physical Therapy f

141

Muscle and Tendon Changes in Infants Born Preterm

Figure 2. Measurement of AMax muscle length. The AMax goniometric measurement is recorded when the muscle belly and the tendon are stretched to their maximum. This measure represents slack removed from the tendon and full stretch of the muscle belly.

tautness (Fig. 1). This measurement reflects the point in ankle range where slack is removed from the Achilles tendon and the muscle belly of the gastrocnemius-soleus muscle is relaxed. A second measurement, AMax, then was taken after the foot was gently moved into maximum dorsiflexion (Fig. 2). When the foot was moved into maximum dorsiflexion, the infant’s behavior was closely monitored to ensure no discomfort resulted from this movement. This measurement reflected a fully lengthened Achilles tendon and maximal stretch in the muscle belly of the gastrocnemius-soleus muscle. The knee was positioned into maximum extension during measurement of both AO and AMax. Knee extension limitations can be present in newborn infants who are developmentally normal, and have been documented as a mean of ⫺15.3 degrees37 to ⫺21.4 degrees.15 Brown and Swenson38 examined dorsiflexion in newborns with knee extension and knee flexion to 90 degrees. In girls, dorsiflexion was 60 degrees with the knee extended and 63 degrees with the knee flexed. The difference between dorsiflexion with and without knee flexion was 142

f

Physical Therapy

Volume 89

57 and 59 degrees, respectively, in boys. Based on the data from Brown and Swenson38 and knowing knee extension limitation is less than 45 degrees, we suggest the change in the MTU length measures with the knee extended as far as possible would be a difference of no more than half of that found with 90 degrees of knee flexion,38 or 1 to 1.5 degrees. The accepted standard for measuring ankle range of motion in the newborn is to extend the knee as far as possible.11,14,15,37 Applying this clinical standard to this study allowed greater ease in transferring applications into clinical practice. Each measurement was taken twice by 2 examiners who were blinded to the other examiner’s measurements. For each measurement, the infant’s foot was placed in a plantar-flexed position, moved toward dorsiflexion for the measurement of AO, and then moved into maximum dorsiflexion for the measurement of AMax. The foot was returned to a plantar-flexed position, and then the second measurements were taken in the same manner. Only one examiner measured at a time, and the examiners did not watch each other measure. Each examiner recorded 2 measures

Number 2

of AO and AMax on a sticky note that was added to the research folder for each infant after measurements were complete. If the examiner recorded 2 measurements for either AO or AMax that were more than 4 degrees different, a third measurement was taken, and the examiner recorded the 2 measurements that were most similar. A third measurement was obtained during 15% of the measurements for both examiners on either AO or AMax. Once the measurements were complete, each infant was dressed, and the next session was scheduled. Blinding of infant gestational age group was not possible because each examiner was able to visually distinguish among infants in each group based on experience, particularly at term age. The ICC (2,1) for reliability of AO between the 2 raters on 100 of the 126 measurements was .99 (95% confidence interval⫽.98 –.99). The ICC (2,1) for reliability of AMax between the 2 raters on 99 of the 126 measurements was .95 (95% confidence interval⫽ .93–.97). Data Analysis Data for measurements of AO and AMax were entered into SPSS version 12.0.† The 2 measurements of AO and AMax were averaged after they were entered into SPSS. If a third measurement was taken, the 2 most similar measurements were averaged. A third variable, muscle belly stretch (AO to AMax), was created by subtracting the measurement of AO from the measurement of AMax. This variable represents muscle extensibility, the distance the muscle belly fibers stretched. All 3 values were entered for each infant at each age measured, resulting in 126 measurements for each measure of MTU length.

† SPSS Inc, 233 S Wacker Dr, Chicago, IL 60606.

February 2009

Muscle and Tendon Changes in Infants Born Preterm Differences between infants born preterm and infants born full term for each of the 3 measures (AO, AMax, and AO to AMax) were analyzed using a 2 ⫻ 3 (term ⫻ age), mixed-model analysis of variance (ANOVA). In this statistical model, the first factor (term) was a between-subjects factor that was a random effect, and the second factor (age) was a withinsubjects factor that was a fixed effect. Because 3 separate ANOVAs were performed to analyze these data, a Bonferroni adjustment was made to the alpha level (.05/3⫽ .0167). Follow-up analysis for main effects of age was performed using the Tukey honestly significant difference test. Effect size for the main effects and interactions in the ANOVA were reported using partial eta2. Partial eta2 represents the amount of the measure and error variance that can be attributed to term, age, or the interaction of these main effects. Cohen d effect sizes and 95% confidence interval for these effect sizes were determined for the differences in means between the full-term infants and the preterm infants using the Effect Size Calculator.39 Role of Funding Source This study was funded by the Edward J. Leahy Center for Faculty Research Awards through the University of Scranton. The funding from this grant provided transportation to and from the laboratory setting for all 3 visits, if agreed to by the family. An approved car seat was provided for the families to use when using the provided transportation. Following completion of all 3 data collection meetings, each family was given a gift card for a local store that sold various newborn items. This funding source in no way biased the outcome of this investigation.

Results The means and standard deviations for the 3 measures at term age, 6 weeks of age, and 12 weeks of age February 2009

Table 3. Mean Values and Standard Deviations (in Degrees) of Muscle Tendon Unit Measures at Term Age, 6 Weeks of Age, and 12 Weeks of Agea Full-term Group (nⴝ20)

Preterm Group (nⴝ22)

X (SD)

X (SD)

Cohen d (95% Confidence Interval)

Term

⫺3.7 (3.1)

⫺17.0 (3.8)

3.82 (2.74–4.75)

6 wk

⫺3.3 (2.0)

⫺14.8 (4.2)

3.44 (2.43–4.32)

12 wk

⫺2.5 (0.8)

⫺14.1 (2.3)

6.61 (4.97–8.00)

48.0 (3.8)

39.3 (6.0)

1.71 (0.98–2.39)

Measure AO

AMax Term 6 wk

46.5 (3.2)

41.1 (5.7)

1.15 (0.48–1.78)

12 wk

47.4 (3.3)

43.0 (4.7)

1.07 (0.41–1.70)

51.7 (4.9)

56.3 (7.4)

0.73 (0.09–1.34)

AO to AMax Term 6 wk

49.8 (3.6)

55.9 (5.2)

1.35 (0.66–2.00)

12 wk

49.9 (3.3)

57.1 (5.6)

1.55 (0.83–2.21)

a Plantar flexion is denoted by negative degrees, and dorsiflexion is denoted by positive degrees. AO⫽taut tendon, relaxed muscle belly; AMax⫽taut tendon, stretched muscle belly; AO to AMax⫽muscle belly stretch.

are presented in Table 3. Cohen d effect sizes ranged from 0.73 to 6.61 (Tab. 3), suggesting differences in the means between the groups were medium-large to very large.40 Measurements of AO, AMax, and AO to AMax for both groups over all 3 ages are represented in Figures 3 and 4. Differences in MTU length (AO and AMax) and muscle belly stretch (AO to AMax) between infants born preterm and infants born full term were all significant (Tab. 4). At term age, 6 weeks of age, and 12 weeks of age, infants born preterm demonstrated values of AO and AMax in positions of greater plantar flexion, with a larger range of AO to AMax, compared with values in infants born full term (Tab. 3). Partial eta2 ranges were as high as .90 for AO, suggesting that full term versus preterm explained 90% of the variability observed in AO (Tab. 4). A significant difference in the main effect of age was found for AO (Tab. 4). As infants aged, the measure of AO was in a position of more dorsiflexion,

representing an increase in tendon length. A Tukey honestly significant difference post hoc analysis revealed a significant difference in AO only between term age and 12 weeks of age.

Discussion The results support the hypothesis that, at term age, infants born preterm and infants born full term differ in gastrocnemius-soleus MTU length and muscle belly stretch and that these differences persist at 6 weeks and 12 weeks of age. Infants born preterm demonstrate less MTU length with a taut tendon, relaxed muscle belly (AO), and a taut tendon, stretched muscle belly (AMax) at term age and at 12 weeks of age. The increased range of muscle belly stretch (AO to AMax) indicates that the muscle belly stretched more in the infants born preterm from term age to 12 weeks of age. Assuming more muscle belly stretch is primarily the result of a longer muscle belly with more sarcomeres or longer sarcomere length,6,7,18,41 the tendon may be shorter in the infant born preterm.

Volume 89

Number 2

Physical Therapy f

143

Muscle and Tendon Changes in Infants Born Preterm

Figure 3. Measurements of taut tendon, relaxed muscle belly (AO) and taut tendon, stretched muscle belly (AMax) at term age, 6 weeks of age, and 12 weeks of age in infants born full term and infants born preterm. Plantar flexion is represented by negative degrees, and dorsiflexion is represented by positive degrees.

Figure 4. Measurements of muscle belly stretch (taut tendon, relaxed muscle belly [AO] to taut tendon, stretched muscle belly [AMax]) at term age, 6 weeks of age, and 12 weeks of age in infants born full term and infants born preterm.

144

f

Physical Therapy

Volume 89

Number 2

February 2009

Muscle and Tendon Changes in Infants Born Preterm Table 4. Main Effects for Term, Age, and Interaction on Analyses of Variance for 3 Measures of the Muscle Tendon Unita Measure AO

AMax

AO to AMax

Between Terms F1,40⫽358.27

Between Ages F2,39⫽6.23

Interaction F2,39⫽1.33

P⬍.001

P⫽.005

P⫽.28

Power⫽1.00

Power⫽.87

Power⫽.27

Partial eta2⫽.90

Partial eta2⫽.24

Partial eta2⫽.064

F1,40⫽30.65

F2,39⫽2.92

F2,39⫽3.65

P⬍.001

P⫽.066

P⫽.035

Power⫽1.00

Power⫽.54

Power⫽.56

Partial eta2⫽.43

Partial eta2⫽.13

Partial eta2⫽.14

F1,40⫽25.20

F2,39⫽.89

F2,39⫽.72

P⬍.001

P⫽.42

P⫽.49

Power⫽1.00

Power⫽.19

Power⫽.16

Partial eta2⫽.39

Partial eta2⫽.044

Partial eta2⫽.036

AO⫽taut tendon, relaxed muscle belly; AMax⫽taut tendon, stretched muscle belly; AO to AMax⫽muscle belly stretch. Power was determined retrospectively.

a

Our results suggest that only one measure (AO) demonstrated a change over age. Lengthening of the tendon occurred in all infants, both preterm and full term, between term age and 12 weeks of age. This change could be explained by numerous factors, including decreased knee flexion contraction from term age to 12 weeks of age, tendon growth, and an MTU adaptation to more kicking. Increased tendon length and decreased knee flexion contraction are not convincing explanations because AMax also would have changed from term age to 12 weeks of age.42 An increase in range of dorsiflexion movements between 6 and 9 weeks of age has been reported by Ferrari et al30 and is consistent with the increase in tendon length we found in passive measures. The increase in tendon length could be an adaptive response to inertial forces on the ankle during kicking between term age and 12 weeks of age. Group ⫻ age interactions were not statistically significant at a level of .0167 for any of the 3 measures of the MTU. However, the interaction February 2009

for AMax was .035, which would have been accepted as significant if we had not adjusted our alpha level. These measures are found to move in opposite directions, decreasing in the full-term group and increasing in the preterm group, when the means are examined (Tab. 3), suggesting that these measures could be moving in different patterns. The patterns of changes in MTU over the 3 ages should be interpreted cautiously due to a possible type II error. Power is reported in Table 4 and should be examined when interpreting the results of the statistical analyses and applying the results. The results from this study are consistent with the findings from our previous study.20 As shown in Table 5 and Figure 5, our results for AO, AMax, and AO to AMax are very similar and support the reliability of our data using the procedure developed by Tardieu and colleagues.18 An unexpected finding was the similarity between the preterm infants at birth in our previous study20 and the preterm infants at term age in the current study. Although some authors2,4

have reported postural changes from birth to term age in infants born preterm, we found no evidence of a similar change in the gastrocnemiussoleus MTU. Our findings at the ankle do not correspond with either the assumption that gastrocnemiussoleus MTU changes from birth to term age as the result of gravity causing more ankle extension12,13 or the assumption that gastrocnemiussoleus MTU changes from birth to term age as the result of increased flexor tone causing more ankle flexion.2– 4 The significant difference in muscle belly stretch (AO to AMax) between the 2 groups in this study differs from the findings of our previous study.20 Post hoc statistical power for the previous study’s comparison of AO to AMax between full-term infants and preterm infants was low (.06), suggesting the possibility of a type II error. In this study, we had a similar number of infants, but measurements were repeated 3 times, resulting in higher statistical power. If a goal of therapy is to re-create the MTU length observed in infants born full term while intervening with the infant born preterm, the age window for successful therapeutic implementation may be limited. The changes we found in this study between the infants born full term, who underwent a prolonged stretch in utero during the last 4 to 6 weeks of gestation, and the infants born preterm, who did not undergo this prolonged stretch, is the same response Tardieu et al6 and Williams and Goldspink7 found in the animal studies. Our data suggest that the adaptation of muscle in the full-term group results in lengthening of the tendon and shortening of the muscle belly in comparison with the infants born preterm. Because the response in these infants is similar to the animal model for very young animals, there may be a period where the adaptation to stretch will change and an intervention will no longer lead to the same

Volume 89

Number 2

Physical Therapy f

145

Muscle and Tendon Changes in Infants Born Preterm Table 5. Means, Standard Deviations, Medians, Interquartile Ranges, and 95% Confidence Intervals (in Degrees) for 3 Measures of the Muscle Tendon Unit in Our Pilot Study20 and the Current Studya Full term Measure

Preterm

Pilot Study

Current Study

Pilot Study

Current Study

⫺6.04

⫺3.70

⫺18.05

⫺17.00

AO X SD Median Interquartile range

6.92

3.12

8.17

3.77

⫺4.00

⫺2.75

⫺16.50

⫺16.25

3.50

95% confidence interval

⫺9.20 to ⫺2.90

3.25 ⫺5.16 to ⫺2.24

7.00 ⫺21.87 to ⫺14.23

3.63 ⫺18.67 to ⫺15.33

AMax X

49.19

SD Median Interquartile range

37.85

39.30

4.59

3.78

8.76

6.00

48.00

48.75

40.00

38.75

7.50

95% confidence interval

48.03

4.38

47.10 to 51.28

46.26 to 49.79

55.24

51.73

7.75 33.75 to 41.95

9.13 36.63 to 41.96

AO to AMax X SD Median Interquartile range

a

56.30

7.16

4.88

7.85

7.39

54.00

51.00

55.00

56.50

10.50

95% confidence interval

55.90

51.98 to 58.50

6.88 49.44 to 54.01

6.50 52.22 to 59.58

11.00 53.02 to 59.57

AO⫽taut tendon, relaxed muscle belly; AMax⫽taut tendon, stretched muscle belly; AO to AMax⫽muscle belly stretch.

adaptive change found in the very young infant, but will switch to the adaptation found in adult animals and humans. Different responses to immobilization at different ages suggest that intervention to mimic the lengthening in infants born preterm should occur as soon as possible if it is to be similar to the lengthening that occurs in utero in infants born full term. Muscle tendon unit length measures at the ankle should be used to document gastrocnemius-soleus muscle changes in infants born preterm from birth to term age and beyond. Successful dorsiflexion positioning using a splint that limits plantar flexion and allows active dorsiflexion in the NICU or shortly after discharge should result in MTU lengths closer to those found in the full-term group.43,44 The preterm infants in this 146

f

Physical Therapy

Volume 89

study and in our previous study20 did not receive physical therapy positioning programs in the NICU, so positioning in neutral and dorsiflexed ankle positions by a physical therapist needs to be explored. This study does support the idea that MTU length and muscle belly stretch can be used as reliable clinical measures to document preterm MTU changes in the NICU and during early intervention. In addition, to increase reliability when performing measurements of ankle range of motion, therapists should repeat each measurement twice to compare consistency, monitor behavioral state during measurements, and move the ankle through the full available range of motion prior to taking the measurements to decrease creep. The gastrocnemius-soleus MTU length in infants born preterm may poten-

Number 2

tially affect lower-extremity coordinated movement patterns and motor learning. A shorter gastrocnemiussoleus MTU would lead to movement in more plantar flexion during lower-extremity kicking. Use of a lower-extremity kicking pattern with increased ankle plantar flexion could explain, at least partially, ankle plantar flexion during supine kicking,10 new walkers’ initial foot contact in plantar flexion (toe touch),28 and persistent toe walking in preterm infants with decreased passive dorsiflexion.29 This method of measuring gastrocnemius-soleus MTU length may be a quick method available to physical therapists to predict which infants are at continued risk for changes in learning coordinated movement patterns such as kicking, which ultimately could alter the development of motor skills, such as walking, at later ages. Research is February 2009

Muscle and Tendon Changes in Infants Born Preterm

Figure 5. Mean values and 95% confidence intervals for measures of taut tendon, relaxed muscle belly (AO); taut tendon, stretched muscle belly (AMax); muscle belly stretch (AO to AMax) from pilot study20 and current study.

needed to investigate changes in MTU length at later ages and the effects of these changes on development of supine kicking, standing, and walking function. There are limitations in the design of this study. In addition to the lack of full knee extension and the inability to blind the examiners to the groups described earlier, another possible limitation of the results of this study is the low statistical power for the main effect of age. Our a priori power was based on the effect sizes found in our initial study20 with respect to differences between infants born preterm and infants born full term and not on changes over age. To adequately power a follow-up study, a minimum of 40 subjects would be needed in each of the groups of infants, or larger age intervals could be compared to increase the likelihood of detecting a difference over age. The lack of signifiFebruary 2009

cance found in this study over age needs to be interpreted with caution, considering the possibility of a type II error. Another limitation of this study is the lack of understanding of MTU length measures past the age of 12 weeks or in infants born preterm who are not low risk. It is possible that the MTU issues resolve or continue at older ages and similar changes are found in infants who are at higher risk for developmental issues. In addition, comparison of the same infants born preterm from birth to term age would be more convincing evidence that MTU lengths from birth until term age did not change during this period.

Conclusions Muscle tendon unit lengths in the gastrocnemius-soleus muscle are different between infants born full term and infants born preterm. In-

fants born preterm demonstrate shorter tendon length and shorter overall tendon length and muscle belly stretch, but more muscle belly stretch. These differences are present at term age and remain until 12 weeks of age (adjusted for the preterm group). Measurements of MTU length at term age in this study are similar to previous measurements obtained in preterm infants at birth. These measures can be used reliably to document changes in the MTU during an infant’s stay in the NICU and during early intervention followup post-discharge. Additional research is needed to document the effects of the MTU lengths on active movement at these ages and beyond. Dr Grant-Beuttler and Dr Palisano provided concept/idea/research design. Dr GrantBeuttler, Dr Palisano, Dr Heriza, and Dr Shewokis provided writing. Dr GrantBeuttler, Dr Miller, and Dr Reddien Wagner provided data collection, project manage-

Volume 89

Number 2

Physical Therapy f

147

Muscle and Tendon Changes in Infants Born Preterm ment, and participants. All authors provided data analysis. Dr Grant-Beuttler provided fund procurement and facilities/equipment. This project was approved by institutional review boards at the University of Scranton, Scranton-Temple Residency Program, and Drexel University. It also received approval from the Community Medical Center’s Executive Board. This research was presented at the International Conference for Society for Chaos Theory in Psychology & Life Sciences; July 27– 29, 2007; Orange, California, and as a poster presentation at the Combined Sections Meeting of the American Physical Therapy Association; February 1–5, 2006; San Diego, California. This study was funded by an Edward R Leahy Jr Center Research Award. This article was received October 9, 2007, and was accepted November 8, 2008. DOI: 10.2522/ptj.20070306

References 1 Carter RE, Campbell SK. Early neuromuscular development of the premature infant. Phys Ther. 1975;55:1332–1341. 2 Palmer PG, Duowitz LM, Verghote M, Dubowitz V. Neurological and neurobehavioral differences between preterm infants at term and full-term newborn infants. Neuropediatrics. 1982;13:183–189. 3 Vaivre-Douret L, Ennouri K, Jrad I, et al. Effects of positioning on the incidence of abnormalities of muscle tone in low-risk, preterm infants. Eur J Paediatr Neurol. 2004;8:21–34. 4 Lacey JL, Henderson-Smart DJ, Edwards DA. A longitudinal study of early leg postures of preterm infants. Dev Med Child Neurol. 1990;32:151–163. 5 Downs JA, Edwards AD, McCormick DC, et al. Effect of intervention on development of hip posture in very preterm babies. Arch Dis Child. 1991;66:797– 801. 6 Tardieu C, Tabary JC, Tabary C, Heut de la Tour E. Comparison of the sarcomere number and adaptation in young and adult animals: influence of tendon adaptation. J Physiol (Paris). 1977;73:1045–1055. 7 Williams PE, Goldspink G. Changes in sarcomere length and physiological properties in immobilized muscle. J Anat. 1978; 127:459 – 468. 8 Tabary JC, Tabary C, Tardieu C, et al. Physiological and structural changes in the cat’s soleus muscle due to immobilization at different lengths by plaster casts. J Physiol. 1972;224:231–244. 9 Davis DW, Thelen E, Keck J. Treadmill stepping in infants born prematurely. Early Hum Dev. 1994;39:211–223.

148

f

Physical Therapy

Volume 89

10 Heriza C. Comparisons of leg movements in preterm infants at term with healthy full-term infants. Phys Ther. 1988;68: 1687–1693. 11 Harris MB, Simons CJR, Ritchie SK, et al. Joint range of motion development in premature infants. Ped Phys Ther. 1990;9: 185–191. 12 Desmond MM, Wilson GS, Alt EJ, Fisher ES. The very low birth weight infant after discharge from intensive care: anticipatory health care and developmental course. Curr Probl Pediatr. 1980;10:1–59. 13 Grenier A. Prevention of early hip deformities in the brain-damaged newborn infant: Little’s disease without scissoring? Ann Pediatr (Paris). 1988;35:423– 427. 14 Hoffer MM. Joint limitations in newborns. Clin Orthop Relat Res. 1980;148:94 –96. 15 Broughton NS, Wright J, Menelaus MB. Range of knee motion in normal neonates. J Pediatr Orthop. 1993;13:263–264. 16 D’Elia A, Pighetti M, Moccia G, Santangelo N. Spontaneous motor activity in normal fetuses. Early Hum Dev. 2001;65:139 –147. 17 Roodenburg PJ, Wladimiroff JW, van Es A, Prechtl HFR. Classification and quantitative aspects of fetal movements during the second half of normal pregnancy. Early Hum Dev. 1991;25:19 –35. 18 Tardieu C, Heut de la Tour, Bret MD, Tardieu G. Muscle hypoextensibility in children with cerebral palsy, I: clinical and experimental observations. Arch Phys Med Rehabil. 1982;63:97–102. 19 Grant-Beuttler M, Leininger PM, Palisano RJ. Reliability of a measure of muscle extensibility in full-term and preterm newborns. Phys Occup Ther Pediatr. 2004;24: 173–185. 20 Grant-Beuttler M, Shewokis PA. Muscle tendon unit comparisons between infants born preterm and infants born full term: a pilot study. Ped Phys Ther. 2007;19: 309 –314. 21 Fallang B, Saugstad O, Gregaard J, HaddersAlgra M. Kinematic quality of reaching movements in preterm infants. Pediatr Res. 2003;54:429 – 435. 22 Van der Fits I, Flikweert E, Stremmelaar E, et al. Development of postural adjustments during reaching in preterm infants. Pediatr Res. 1999;46:1–7. 23 Hadders-Algra M, Brogren E, Katz-Salamon M, Forssberg H. Periventricular leucomalacia and preterm birth have different detrimental effects on postural adjustments. Brain. 1999;22:727–740. 24 Sagnol C, Debillon T, Debu ˆ B. Assessment of motor control using kinematic analysis in preschool children born very preterm. Dev Psychobiol. 2007;49:421– 432. 25 Geerdink J, Hopkins B, Beek W, Heriza C. The organization of leg movements in preterm and full-term infants after term age. Dev Psychobiol. 1996;29:335–351. 26 Droit S, Boldrini A, Cioni G. Rhythmic leg movements in low-risk and brain-damaged preterm infants. Early Hum Dev. 1996; 44:201–213.

Number 2

27 Thelen E, Bradshaw G, Ward J. Spontaneous kicking in month-old infants: manifestation of a human central locomotor pattern. Behav Neural Biol. 1981;32:45–53. 28 Cioni G, Duchini F, Milianti B, et al. Differences and variations in the patterns of early independent walking. Early Hum Dev. 1993;35:193–205. 29 Georgieff M, Bernbaum J, HoffmanWilliamson M, Daft A. Abnormal truncal muscle tone as a useful early marker for developmental delay in low birth weight infants. Pediatrics. 1986;77:659 – 663. 30 Ferrari F, Cioni G, Einspieler C, et al. Cramped synchronized general movements in preterm infants as an early marker for cerebral palsy. Arch Pediatr Adolesc Med. 2002;156:422– 423. 31 Jeng SF, Yau KI, Liao HF, et al. Prognostic factors for walking attainment in very lowbirthweight preterm infants. Early Hum Dev. 2000;59:159 –173. 32 Lacey J. Very low-birthweight preterm infants walk later than term infants, but most are walking by 18 months. Aust J Physiother. 2001;47:65. 33 Buchner A, Faul F, Erdfelder E. G-Power: A Priori, Post Hoc, and Compromise Power Analysis for Windows-based Operating System. Trier, Germany: University of Trier; 1996. 34 Brazelton TB. Neonatal Behavioral Assessment Scale. 2nd ed. Philadelphia, PA: JB Lippincott Co; 1984. Clinics in Developmental Medicine, No. 88. 35 Als H. Neonatal Individualized Developmental Care and Assessment Program (NIDCAP). Boston, MA: Children’s Hospital; 1984. 36 Norkin CC, White JD. Measurement of Joint Motion: A Guide to Goniometry. 3rd ed. Philadelphia, PA: FA Davis Co; 2003. 37 Waugh KG, Minkel JL, Parker R, Coon VA. Measurement of selected hip, knee, and ankle joint motions in newborns. Phys Ther. 1983;63:1616 –1621. 38 Brown GA, Swenson DR. A descriptive system for lower extremity evaluation in children: data for the newborn infant. Orthopedics. 2000;23:111–115. 39 Effect Size Calculator. Available at: http:// davidmlane.com/hyperstat/effect-size.html. Accessed March 27, 2003. 40 Cohen J. Statistical Power Analysis for the Behavior Sciences. 2nd ed. Hillsdale, NJ: Erlbaum; 1988. 41 Brand PW, Beach RB, Thompson DE. Relative tension and potential excursion of muscles in the forearm and hand. J Hand Surg. 1981;6:209 –219. 42 Huijing PAJBM, Rozendal RH, Heslinga JW, Willems MET. Skeletal muscle reaction to growth and immobilization. J Rehabil Res Dev. 1994;43:30 –31. 43 Noonan K, Richards B. Nonsurgical management of idiopathic clubfoot. J Am Acad Orthop Surg. 2003;11:392– 402. 44 Bensahel H, Guillaume A, Czukonyi Z, Desgrippes Y. Results of physical therapy for idiopathic clubfoot: a long-term follow-up study. J Pediatr Orthop. 1990;10:189 –192.

February 2009

Research Report Physical Therapy Health Human Resource Ratios: A Comparative Analysis of the United States and Canada Michel D Landry, Thomas C Ricketts, Erin Fraher, Molly C Verrier

Background and Purpose. Health human resource (HHR) ratios are a measure of workforce supply and are expressed as a ratio of the number of health care practitioners to a subset of the population. Health human resource ratios for physical therapists have been described for Canada but have not been fully described for the United States. In this study, HHR ratios for physical therapists across the United States were estimated in order to conduct a comparative analysis of the United States and Canada.

Methods. National US Census Bureau data were linked to jurisdictional estimates of registered physical therapists to create HHR ratios at 3 time points: 1995, 1999, and 2005. These results then were compared with the results of a similar study conducted by the same authors in Canada.

Results. The national HHR ratio across the United States in 1995 was 3.8 per 10,000 people; the ratio increased to 4.3 in 1999 and then to 6.2 in 2005. The aggregated results indicated that HHR ratios across the United States increased by 61.3% between 1995 and 2005. In contrast, the rate of evolution of HHR ratios in Canada was lower, with an estimated growth of 11.6% between 1991 and 2005. Although there were wide variations across jurisdictions, the data indicated that HHR ratios across the United States increased more rapidly than overall population growth in 49 of 51 jurisdictions (96.1%). In contrast, in Canada, the increase in HHR ratios surpassed population growth in only 7 of 10 jurisdictions (70.0%). Discussion and Conclusion. Despite their close proximity, there are differences between the United States and Canada in overall population and HHR ratio growth rates. Possible reasons for these differences and the policy implications of the findings of this study are explored in the context of forecasted growth in demand for health care and rehabilitation services.

MD Landry, PT, PhD, is Adjunct Assistant Professor, Department of Health Policy and Administration, University of North Carolina at Chapel Hill, Chapel Hill, North Carolina, and Assistant Professor, Department of Physical Therapy, Faculty of Medicine, University of Toronto, Rehabilitation Sciences Building, 8th Floor, 160 –500 University Ave, Toronto, Ontario, Canada M5G 1V7. Address all correspondence to Dr Landry at: [email protected]. TC Ricketts, PhD, is Professor, Department of Health Policy and Administration, and Director, North Carolina Rural Health Research Program, Cecil G Sheps Center for Health Services Research, University of North Carolina at Chapel Hill. E Fraher, MPP, PhD Candidate, is Director, NC Health Professions Data System, Cecil G Sheps Center for Health Services Research, University of North Carolina at Chapel Hill. MC Verrier, PT, MSc, is Associate Professor, Department of Physical Therapy and Graduate Department of Rehabilitation Sciences, University of Toronto. [Landry MD, Ricketts TC, Fraher E, Verrier MC. Physical therapy health human resource ratios: a comparative analysis of the United States and Canada. Phys Ther. 2009;89:149 –161.] © 2009 American Physical Therapy Association

Post a Rapid Response or find The Bottom Line: www.ptjournal.org February 2009

Volume 89

Number 2

Physical Therapy f

149

Physical Therapy Health Human Resource Ratios

H

ealth human resources (HHR) continue to emerge as critical factors in health care policy planning.1–3 An overall measure of supply within a workforce, the HHR ratio, is generally expressed as the number of health care practitioners relative to the population or a subset of the population.4 The origins of the use of HHR ratios for workforce policy can be traced back to the 1930s, when pioneering work done in the United States reported that a target of 134.7 physicians per 100,000 people was a desirable benchmark (TC Ricketts, unpublished data, 2003). Since that time, the use of HHR ratios has become a common measure of the density of health care practitioners in a given geographical area. In Canada, according to the PanCanadian Health Human Resource Strategy, “appropriate planning and management of HHR are key to developing a health-care workforce that has the right number and mix of health professionals.”5 Overall, the published literature has focused on estimating the HHR ratio for larger groups of health care practitioners, such as physicians6 – 8 and nurses,9 –11 across multiple time periods. The publication of such workforce ratios has created an evidence base from which health care disciplines have drawn for public policy discussions and professional advocacy activities. In contrast, international estimates for physical therapists are relatively sparse. Landry12 estimated the Canadian HHR ratio for physical therapists to be 5.0 per 10,000 people in 2000, which represented a 16.3% increase from the value in 1991. Landry et al13 followed up on this earlier work and estimated that the Canadian HHR ratio for physical therapists dropped to 4.8 per 10,000 people by 2005. They reported that although there was an 11.6% increase in the HHR ratio between 1991 and 2005, there was an alarming decline in the HHR ratio between 150

f

Physical Therapy

Volume 89

2000 and 2005. Although the reasons for this decline were not discussed, their research demonstrated that the increase in the HHR ratio outstripped overall population growth in only 7 of 10 provincial jurisdictions (70.0%) between 1991 and 2005. In the United States, there has been no similar national-level trend analysis regarding HHR ratios for physical therapists, and the existing data concern ratios at the state or county level.14 To address the gap in nationallevel analyses of HHR ratios across the United States, we carried out this study with the aim of achieving 2 objectives. The first objective was to estimate HHR ratios for physical therapists across jurisdictions by combining US Census Bureau data with the total numbers of active and inactive registered physical therapists across the United States at 3 time points: 1995, 1999, and 2005. The second objective was to compare the HHR ratio estimates in the United States with those in Canada to explore trends in the 2 countries.

Method The method used in this study was similar to that previously applied by Landry12 and Landry et al13 to estimate HHR ratios for physical therapists in Canada. In brief, for the determination of HHR ratios across the United States, 2 sources of data were linked to generate estimates of the number of physical therapists relative to the population. First, publicly available historical population data from the US Census Bureau were retrieved for the years 1995, 1999, and 2005 in each of the 50 states and the District of Columbia.15 Second, publicly available estimates of the total number of active and inactive registered physical therapists in each US jurisdiction were obtained from the American Physical Therapy Association (APTA) for the years 1995, 1999, and 2005. For calculation of the

Number 2

number of physical therapists per 10,000 people in each jurisdiction for each year, a simple mathematical equation was applied: The number of physical therapists in each jurisdiction was multiplied by 10,000, and the result was divided by the overall jurisdictional population. This equation yielded the density of physical therapists per 10,000 people in each jurisdiction at 3 time points. Change scores for the overall population, absolute numbers of physical therapists, and HHR ratios for each jurisdiction then were determined by calculating the percent changes between 1995 and 1999, between 1999 and 2005, and between 1995 and 2005. In this study, both active physical therapists (defined by APTA as currently practicing) and inactive physical therapists (defined by APTA as not practicing) were included in the sample of physical therapists for 3 reasons. First, APTA’s publicly available files collapse both categories, making it difficult to distinguish between them. Second, an objective of this study was to compare US data with Canadian data, and a comparative Canadian study13 included both active and inactive physical therapists in the data analysis. Third, the overall goal of this study was to provide a macro-level perspective on the number of physical therapists relative to the population, and much of the reviewed literature included the total numbers of active and inactive health care professionals as a measure of the density of health care professionals.6 –11 The APTA does not represent all physical therapists in the United States, but this national professional organization triangulates multiple data sources to estimate the total number of licensed physical therapists in the United States. The APTA gathers data on human resources across the United States through anFebruary 2009

Physical Therapy Health Human Resource Ratios Table 1. Total Population by State or Jurisdiction in 1995, 1999, and 2005 Population in: State or Jurisdiction Alabama Alaska Arizona

% Change

1995

1999

2005

4,253,000

4,369,000

4,548,000

From 1995 to 1999 2.75

From 1999 to 2005 4.08

From 1995 to 2005 6.94

604,000

619,000

663,000

2.57

7.06

9.81

4,218,000

4,778,000

5,953,000

13.28

24.58

41.13

Arkansas

2,484,000

2,551,000

2,775,000

2.71

8.79

11.74

California

31,589,000

33,145,000

36,154,000

4.93

9.08

14.45

Colorado

3,747,000

4,056,000

4,663,000

8.25

14.97

24.45

Connecticut

3,275,000

3,282,000

3,500,000

0.21

6.66

6.89

Delaware Florida

717,000

753,000

841,000

5.10

11.71

17.40

14,166,000

15,111,000

17,768,000

6.67

17.58

25.43

Georgia

7,201,000

7,788,000

9,132,000

8.15

17.26

26.82

Hawaii

1,187,000

1,185,000

1,273,000

⫺0.13

7.40

7.27

Idaho

1,163,000

1,251,000

1,429,000

7.63

14.19

22.90

Illinois

11,830,000

12,128,000

12,765,000

2.52

5.25

7.91

Indiana

5,803,000

5,942,000

6,266,000

2.41

5.44

7.98

Iowa

2,842,000

2,869,000

2,965,000

0.96

3.35

4.35

Kansas

2,565,000

2,654,000

2,748,000

3.47

3.55

7.14

Kentucky

3,860,000

3,960,000

4,172,000

2.61

5.35

8.10

Louisiana

4,342,000

4,372,000

4,507,000

0.69

3.09

3.81

Maine

1,241,000

1,253,000

1,318,000

0.97

5.20

6.22

Maryland

5,042,000

5,171,000

5,589,000

2.57

8.08

10.86

Massachusetts

6,074,000

6,175,000

6,433,000

1.67

4.18

5.92

Michigan

9,549,000

9,863,000

10,100,000

3.30

2.40

5.78

Minnesota

4,610,000

4,775,000

5,126,000

3.59

7.35

11.21

Mississippi

2,697,000

2,768,000

2,908,000

2.66

5.05

7.84

Missouri

5,324,000

5,468,000

5,797,000

2.71

6.02

8.90

Montana

870,000

882,000

934,000

1.47

5.89

7.44

Nebraska

1,637,000

1,666,000

1,758,000

1.77

5.53

7.40

Nevada

1,530,000

1,809,000

2,412,000

18.25

33.33

57.67

New Hampshire

1,148,000

1,201,000

1,306,000

4.63

8.80

13.83

New Jersey

7,945,000

8,143,000

8,703,000

2.50

6.87

9.54

New Mexico

1,685,000

1,739,000

1,925,000

3.25

10.70

14.30

18,136,000

18,196,000

19,315,000

0.33

6.15

6.50

7,195,000

7,650,000

8,672,000

6.33

13.35

20.53

New York North Carolina

641,000

633,000

634,000

⫺1.14

0.15

⫺1.00

11,151,000

11,256,000

11,470,000

0.95

1.90

2.87

Oklahoma

3,278,000

3,358,000

3,543,000

2.44

5.52

8.10

Oregon

3,141,000

3,316,000

3,638,000

5.58

9.73

15.85

Pennsylvania

12,072,000

11,994,000

12,405,000

⫺0.65

3.43

2.76

Rhode Island

990,000

990,000

1,073,000

0.08

8.35

8.44

3,673,000

3,885,000

4,246,000

5.79

9.30

15.63

729,000

733,000

774,000

0.57

5.69

6.29

North Dakota Ohio

South Carolina South Dakota

(Continued)

February 2009

Volume 89

Number 2

Physical Therapy f

151

Physical Therapy Health Human Resource Ratios Table 1. Continued Population in: State or Jurisdiction Tennessee

1995

1999

% Change 2005

From 1995 to 1999

From 1995 to 2005

5,256,000

5,483,000

5,955,000

4.33

8.61

13.31

Texas

18,724,000

20,044,000

22,928,000

7.05

14.39

22.46

Utah

1,951,000

2,129,000

2,490,000

9.17

16.93

27.64

Vermont

585,000

593,000

622,000

1.49

4.82

6.39

Virginia

6,618,000

6,872,000

7,564,000

3.85

10.06

14.30

Washington

5,431,000

5,756,000

6,291,000

5.99

9.30

15.85

554,000

519,000

582,000

⫺6.32

12.15

5.06

Washington, DC West Virginia

1,828,000

1,806,000

1,814,000

⫺1.15

0.40

⫺0.76

Wisconsin

5,123,000

5,250,000

5,527,000

2.49

5.28

7.90

Wyoming United States

480,000

479,000

508,000

⫺0.08

6.09

6.00

262,755,000

272,690,000

296,507,000

3.78

8.73

12.85

nual requests to all licensing bodies in each of the 50 states and the District of Columbia. In large measure, the licensing bodies provide HHR information to APTA; however, because a physical therapist can be licensed to practice in more than one state, relying solely on individual state-level data would overestimate the number of licensed physical therapists in the United States. Therefore, each state’s licensing board is asked to report the total number of physical therapists licensed and the total number of physical therapists residing in that state. The APTA then cross-references these data sources to estimate the total number of physical therapists in each state. Although there is an error rate associated with these estimates, APTA has estimated HHR data by using the same methodology at the 3 time periods of interest in this study, and these data represent the only publicly available national-level data.

Results

Population growth, absolute numbers of physical therapists, and HHR ratios obtained from the US sample were used, along with published Canadian data, in a comparative analysis of the neighboring countries. For the comparative analysis, the US data

Population Growth in the United States The population of the United States in 2005 was 296.5 million, representing an increase of 8.7% from 1999 and an increase of 12.9% from 1995.15,16 Although the population of the United States increased be-

152

From 1999 to 2005

f

Physical Therapy

Volume 89

in this study ranged from 1995 to 2005 (the 3 data points were 1995, 1999, and 2005), but the Canadian data ranged from 1991 to 2005 (the 3 data points were 1991, 2000, and 2005). Although the final data points were the same for both countries, there were slight differences in the first and second points. The differences resulted from the availability of publicly available data from the original sources. The first and second data points were within 4 years of each other, a fact that in and of itself introduces complexity to the analysis, but these data were the best available. The data from the United States and the comparative analysis of the United States and Canada are presented in the “Results” section. Explorations into the reasons for differences and policy implications of our findings are presented in the “Discussion” section.

Number 2

tween 1995 and 2005, the positive growth was not equal across all jurisdictions. As shown in Table 1, 49 of 51 jurisdictions had positive growth between 1995 and 2005; however, West Virginia and North Dakota reported slight decreases in their populations— 0.8% and 1.0%, respectively. Growth in the Absolute Number of Physical Therapists in the United States The data indicated that the absolute number of physical therapists in the United States increased from 98,696 in 1995 to 167,810 in 2005, representing an increase of 70.0%. As shown in Table 2, the growth in the number of physical therapists was not equal across the United States; there were differences among jurisdictions. Although each of the 51 jurisdictions experienced overall positive growth in the absolute number of physical therapists from 1995 to 2005, there was a range from a low of 13.7% in Florida to a high of 151.0% in West Virginia. However, in the latter 5-year period (1999 –2005), Michigan and the District of Columbia experienced decreases of 2.5% and 11.7%, respectively. February 2009

Physical Therapy Health Human Resource Ratios Table 2. Number of Physical Therapists by State or Jurisdiction in 1995, 1999, and 2005 No. of Physical Therapists in: State or Jurisdiction Alabama Alaska Arizona Arkansas

1995 942

% Change

1999

2005

1,218

1,524

From 1995 to 1999 29.30

From 1999 to 2005

From 1995 to 2005

25.15

61.78

246

258

391

4.88

51.55

58.94

1,670

1,772

2,655

6.11

49.83

58.98

723

765

1,308

5.81

70.98

80.91

California

11,400

13,511

16,402

18.52

21.40

43.88

Colorado

2,516

3,716

3,955

47.69

6.43

57.19

Connecticut

2,467

2,720

3,198

10.26

17.57

29.63

Delaware

289

329

713

13.84

116.72

146.71

Florida

8,072

6,113

9,178

⫺24.27

50.14

13.70

Georgia

1,880

2,072

3,825

10.21

84.60

103.46

425

449

699

5.65

55.68

64.47

Hawaii Idaho

432

520

863

20.37

65.96

99.77

Illinois

4,552

4,576

6,877

0.53

50.28

51.08

Indiana

1,840

2,100

3,715

14.13

76.90

101.90

822

1,218

1,444

48.18

18.56

75.67

Iowa Kansas

707

959

1,423

35.64

48.38

101.27

Kentucky

1,128

1,335

2,032

18.35

52.21

80.14

Louisiana

1,194

1,547

1,958

29.56

26.57

63.99

644

765

1,224

18.79

60.00

90.06

Maine Maryland

1,948

2,791

3,562

43.28

27.62

82.85

Massachusetts

4,259

4,490

6,622

5.42

47.48

55.48

Michigan

3,036

6,454

6,291

112.58

⫺2.53

107.21

Minnesota

2,048

2,281

3,140

11.38

37.66

53.32

Mississippi

573

943

1,146

64.57

21.53

100.00

1,851

2,390

3,357

29.12

40.46

81.36

Montana

428

501

778

17.06

55.29

81.78

Nebraska

554

703

1,101

26.90

56.61

98.74

Missouri

Nevada

374

374

846

0.00

126.20

126.20

New Hampshire

674

786

1,315

16.62

67.30

95.10

2,785

3,976

6,221

42.76

56.46

123.38

589

745

984

26.49

32.08

67.06

New York

7,448

7,209

14,098

⫺3.21

95.56

89.29

North Carolina

2,382

3,119

4,365

30.94

39.95

83.25

New Jersey New Mexico

272

266

453

⫺2.21

70.30

66.54

Ohio

3,191

3,713

5,848

16.36

57.50

83.27

Oklahoma

1,002

1,210

1,489

20.76

23.06

48.60

Oregon

1,700

1,655

2,314

⫺2.65

39.82

36.12

Pennsylvania

5,404

5,613

9,201

3.87

63.92

70.26

Rhode Island

567

616

925

8.64

50.16

63.14

South Carolina

873

1,155

1,775

32.30

53.68

103.32

North Dakota

(Continued)

February 2009

Volume 89

Number 2

Physical Therapy f

153

Physical Therapy Health Human Resource Ratios Table 2. Continued No. of Physical Therapists in: State or Jurisdiction South Dakota

1995

1999

278

390

2005 586

From 1995 to 1999 40.29

From 1999 to 2005 50.26

From 1995 to 2005 110.79

Tennessee

1,671

2,422

3,055

44.94

26.14

82.82

Texas

5,336

7,074

8,863

32.57

25.29

66.10

Utah

542

723

1,259

33.39

74.14

132.29

Vermont

383

532

652

38.90

22.56

70.23

Virginia

2,071

2,800

4,939

35.20

76.39

138.48

Washington

1,988

2,732

3,888

37.42

42.31

95.57

Washington, DC

316

429

379

35.76

⫺11.66

19.94

West Virginia

361

468

906

29.64

93.59

150.97

Wisconsin

1,676

2,602

3,766

55.25

44.73

124.70

Wyoming

167

247

304

47.90

23.08

82.04

98,696

117,352

167,810

18.90

43.00

70.03

United States

HHR Ratios for Physical Therapists in the United States For exploration of trends over time, HHR ratios for physical therapists per 10,000 people in each of the 50 states and the District of Columbia were determined for 3 time points: 1995, 1999, and 2005. The national averages for physical therapists per 10,000 people across the United States were 3.8 in 1995, 4.3 in 1999, and 6.2 in 2005. The trend for HHR ratios thus represented an aggregated increase of 61.3% between 1995 and 2005. As shown in Table 3, the increases in HHR ratios between 1995 and 2005 ranged from a low of 6.6% in Florida to a high of 153.9% in West Virginia. However, during the period from 1995 to 1999, 7 of 51 jurisdictions showed decreases in HHR ratios. These states included Arizona (6.3%), Florida (29.0%), Illinois (2.0%), Nevada (15.4%), New York (3.5%), North Dakota (1.1%), and Oregon (7.8%). Moreover, during the period from 1999 to 2005, Michigan and the District of Columbia experienced decreases of 2.5% and 11.7%, respectively. In aggregate, even through some jurisdictions showed periods of decline, HHR ratios across 154

% Change

f

Physical Therapy

Volume 89

the United States showed positive growth over the 10-year period from 1995 to 2005. Change Scores for Population Growth and HHR Ratios for Physical Therapists in the United States To more fully appreciate the association between trends in overall population growth and trends in HHR ratios over time, we plotted change scores for population growth and change scores for HHR ratios for physical therapists per 10,000 people between 1995 and 2005 for each of the 9 US Census Bureau regions (Fig. 1). We decided to present these data according to US Census Bureau regions rather than the 51 individual jurisdictions to provide a macro-level perspective on the relationship between HHR ratios and population growth. The data indicated that the increase in HHR ratios surpassed population growth in all US Census Bureau regions. However, analysis at the state level demonstrated that the increase in HHR ratios surpassed population growth in all but Arizona and Florida. In Arizona, the population grew by 41.1%, whereas HHR

Number 2

ratios increased by 40.3%; in Florida, the population grew by 25.4%, whereas HHR ratios increased by only 6.6%. Moreover, the data indicated that HHR ratios across the United States increased more rapidly than overall population growth in 49 of 51 jurisdictions (96.1%). In contrast, in Canada, the increase in HHR ratios surpassed population growth in only 7 of 10 jurisdictions (70.0%).13 Comparison of Overall Population Growth in the United States and Canada To contextualize the US data, we compared the data from the present study with those of a similar study conducted in Canada.13 The population of Canada in 2005 was 32.6 million, representing 19.5% growth from 1991 and 5.9% growth from 2000.17,18 Like that of the United States, the population of Canada increased during the study period, but positive population growth was not found across all individual provincial jurisdictions. Although the 15-year period from 1991 to 2005 generally showed a positive growth pattern in all jurisdictions, population growth February 2009

Physical Therapy Health Human Resource Ratios Table 3. Health Human Resource (HHR) Ratios for Physical Therapists per 10,000 Population by State or Jurisdiction in 1995, 1999, and 2005 HHR Ratio in: State or Jurisdiction

% Change From 1995 to 1999

From 1999 to 2005

From 1995 to 2005

1995

1999

2005

Alabama

2.21

2.79

3.49

25.84

25.12

57.46

Alaska

4.07

4.16

6.31

2.25

51.55

54.97

Arizona

3.96

3.71

5.56

⫺6.33

49.83

40.34

Arkansas

2.91

3.00

5.13

3.02

70.98

76.14

California

3.61

4.08

4.95

12.95

21.40

37.12

Colorado

6.71

9.16

9.75

36.44

6.43

45.21

Connecticut

7.53

8.29

9.74

10.02

17.57

29.35

Delaware

4.03

4.37

9.46

8.32

116.72

134.75

Florida

5.70

4.05

6.07

⫺29.01

50.14

6.59

Georgia

2.61

2.66

4.91

1.90

84.60

88.12

Hawaii

3.58

3.79

5.90

5.78

55.68

64.68

Idaho

3.71

4.15

6.89

11.84

65.96

85.61

Illinois

3.85

3.77

5.67

⫺1.95

50.28

47.36

Indiana

3.17

3.53

6.25

11.44

76.90

97.15

Iowa

2.89

4.24

5.03

46.76

18.56

73.99

Kansas

2.76

3.61

5.36

31.09

48.38

94.52

Kentucky

2.92

3.37

5.13

15.34

52.21

75.56

Louisiana

2.75

3.54

4.48

28.67

26.57

62.86

Maine

5.19

6.11

9.77

17.65

60.00

88.24

Maryland

3.86

5.40

6.89

39.68

27.62

78.27

Massachusetts

7.01

7.27

10.72

3.70

47.48

52.94

Michigan

3.18

6.54

6.38

105.80

⫺2.53

100.60

Minnesota

4.44

4.78

6.58

7.52

37.66

48.01

Mississippi

2.12

3.41

4.14

60.32

21.53

94.83

Missouri

3.48

4.37

6.14

25.71

40.46

76.57

Montana

4.92

5.68

8.81

15.36

55.29

79.14

Nebraska

3.38

4.22

6.61

24.68

56.61

95.27

Nevada

2.44

2.07

4.68

⫺15.43

126.20

91.29

New Hampshire

5.87

6.54

10.95

11.46

67.30

86.47

New Jersey

3.51

4.88

7.64

39.29

56.46

117.93

New Mexico

3.50

4.28

5.66

22.50

32.08

61.80

New York

4.11

3.96

7.75

⫺3.53

95.56

88.66

North Carolina

3.31

4.08

5.71

23.14

39.95

72.33

North Dakota

4.24

4.20

7.15

⫺1.07

70.30

68.47

Ohio

2.86

3.30

5.20

15.27

57.50

81.55

Oklahoma

3.06

3.60

4.43

17.88

23.06

45.06

Oregon

5.41

4.99

6.98

⫺7.79

39.82

28.93

Pennsylvania

4.48

4.68

7.67

4.54

63.92

71.37

Rhode Island

5.73

6.22

9.34

8.55

50.16

63.00

South Carolina

2.38

2.97

4.57

25.06

53.68

92.19 (Continued)

February 2009

Volume 89

Number 2

Physical Therapy f

155

Physical Therapy Health Human Resource Ratios Table 3. Continued No. of Physical Therapists in: State or Jurisdiction

1995

1999

2005

From 1995 to 1999

South Dakota

3.81

5.32

7.99

39.50

50.26

109.60

Tennessee

3.18

4.42

5.57

38.93

26.14

75.24

Texas

2.85

3.53

4.42

23.84

25.29

55.16

Utah

2.78

3.39

5.91

22.19

74.14

112.78

Vermont

6.55

8.96

10.98

36.86

22.56

67.73

Virginia

3.13

4.07

7.19

30.19

76.39

129.64

Washington

3.66

4.75

6.75

29.66

42.31

84.52

From 1999 to 2005

From 1995 to 2005

Washington, DC

5.70

8.27

7.30

44.91

⫺11.66

28.02

West Virginia

1.97

2.59

5.01

31.15

93.59

153.90

Wisconsin

3.27

4.96

7.17

51.48

44.73

119.25

Wyoming

3.48

5.15

6.34

48.03

23.08

82.19

United States

3.76

4.30

6.15

14.57

43.00

63.83

in the latter 5 years of this period (2000 –2005) showed a slightly negative trend in 5 of 10 provincial jurisdictions.13 As shown in Figure 2, the change in the total US population was 12.9% (1995–2005), and the change in the total Canadian population was 19.5% (1991–2005)13; these data indicated that growth rates in Canada surpassed those in the United States. Although the Canadian population showed greater proportional growth, the study period did include an additional 4 years. Comparison of Absolute Numbers of Physical Therapists in the United States and Canada As in the United States, the absolute number of physical therapists in Canada also increased between 1991 and 2005. The total numbers of active and inactive physical therapists increased from 11,794 in 1991 to 15,772 in 2005, representing a 33.7% increase in Canada.19 As in the United States, all 10 provincial jurisdictions experienced positive growth in the absolute number of physical therapists from 1991 to 2005.13 However, growth rates between 2000 and 2005 represented a 156

% Change

f

Physical Therapy

Volume 89

different scenario, with the provinces of Newfoundland & Labrador and Ontario showing declines of 0.5% and 3.1%, respectively. Overall, the aggregated national increases in the absolute numbers of physical therapists were 70.0% in the United States and 33.7% in Canada. Table 4 shows a summary of the comparisons of the 2 countries with regard to proportional changes in population growth, absolute numbers of physical therapists, and HHR ratios. Comparison of HHR Ratios for Physical Therapists in the United States and Canada The US data indicated a 63.8% increase in HHR ratios for physical therapists between 1995 and 2005. In Canada, the national averages of HHR ratios for physical therapists per 10,000 people were 4.3 in 1991 and 5.0 in 2000; however, in 2005, the ratio dropped to 4.8. The overall increase in HHR ratios in Canada between 1991 and 2005 was 11.6%. As shown in Figure 3, the trend for HHR ratios was an upward slope in the United States from 1995 to 2005; in contrast, the data for Canada showed an initial upward slope between

Number 2

1991 and 2000 and a slight downward slope between 2000 and 2005.13 Therefore, it appeared that the growth trends for HHR ratios were similar in both countries until the midpoint of our study period, at which time HHR ratios continued to increase in the United States but began to decrease in Canada.

Discussion The results of the present study highlight macro-level trend data for HHR ratios across the United States and provide an interpretive context for exploring the meaning of these data through a comparative analysis with Canadian data. Our analysis suggests that the numbers of physical therapists relative to the overall population have increased in the United States and Canada, but the proportional increase appears to be much higher in the United States. The findings of the present study also indicate that most jurisdictions in the United States showed increased HHR ratios for physical therapists relative to the overall population but that jurisdictions had different rates of growth. Although the research design does not enable an exploration February 2009

Physical Therapy Health Human Resource Ratios

Figure 1. Comparison of the change in population with the change in health human resource ratios for physical therapists (PTs) per 10,000 people in the United States from 1995 to 2005. Jurisdictions included in each US Census Bureau region were as follows: New England—Connecticut, Maine, Massachusetts, New Hampshire, Rhode Island, and Vermont; Middle Atlantic—New Jersey, New York, and Pennsylvania; East North Central—Illinois, Indiana, Michigan, Ohio, and Wisconsin; West North Central—Iowa, Kansas, Minnesota, Missouri, Nebraska, North Dakota, and South Dakota; South Atlantic—Delaware, Florida, Georgia, Maryland, North Carolina, South Carolina, Virginia, Washington, DC, and West Virginia; East South Central—Alabama, Kentucky, Mississippi, and Tennessee; West South Central—Arkansas, Louisiana, Oklahoma, and Texas; Mountain—Arizona, Colorado, Idaho, Montana, Nevada, New Mexico, Utah, and Wyoming; and Pacific—Alaska, California, Hawaii, Oregon, and Washington.

of causation, future investigations with the concept of small-area analysis in health care services are critical for a greater understanding of the policy and environmental drivers in individual jurisdictions.20 –25 The implications and policy interpretations of our findings are complex and not linear. As in the study of Landry et al,13 the data collected within this research do not establish causation. However, critical questions regarding the optimal number of physical therapists in a given jurisdiction emerged from the present study. For instance, in our view, it would be inappropriate to conclude that the United States is “doing better” or “doing worse” than Canada in terms of HHR ratios because the benchmark for the optimal number of physical therapists relative to a population has not been established. February 2009

To our knowledge, there are no need-based, evidence-based targets or benchmarks for the number of physical therapists relative to a population across clinical settings, dis-

ease conditions, countries, or a combination of these. That is, we were not able to determine whether the national average of 6.2 physical therapists per 10,000 people across the

Figure 2. Population growth in the United States and Canada. Times were as follows: 1—1995 for United States and 1991 for Canada; 2—1999 for United States and 2000 for Canada; and 3—2005 for United States and for Canada.

Volume 89

Number 2

Physical Therapy f

157

Physical Therapy Health Human Resource Ratios Table 4. Summary of Comparative Analysis of Physical Therapy Health Human Resource (HRR) Ratios Across the United States and Canada

Location (Time Span)

% National Population Growth

% Growth in Absolute Number of Physical Therapists

% Growth in HHR Ratios for Physical Therapists Relative to Population

United States (1995–2005)

12.85

70.03

61.29

Canada (1991–2005)

19.50

33.70

11.60

United States in 2005 was high, low, or “just right” because there are no established human resource benchmarks. We suggest that until benchmarks across the care continuum are established, the usefulness of HHR ratios in the policy planning cycle is likely to be somewhat limited. Demand for health care and rehabilitation services is projected to increase in the next decade as a result of factors such as an aging population, increased public expectations, and advances in technology.18,26,27 Ensuring that there is a sufficient human resource supply to meet future demand is a critical policy issue for which further research on health care services is required; the next step is to establish benchmarks. Although evidence-based benchmarks regarding appropriate HHR ratios are not yet known, the propor-

tional growth for physical therapists in both the United States and Canada is impressive compared with that for other disciplines. Although other health care disciplines in the United States tend to have a higher absolute ratio of practitioner to population, their rates of growth in the last decade have been modest compared with the results obtained in the present study. For instance, Shih28 reported that, in the United States, the national ratio for nurses per 10,000 people was 78.2 in 2000; the ratio increased to 82.5 in 2004, a change representing a 5.5% increase between 2000 and 2004. Moreover, McEldowney and Berry29 reported that there were 22.4 physicians per 10,000 people in 1992 and suggested that should that trend continue, there would be 30.0 physicians per 10,000 people by 2020. The forecasted trend suggested by McEl-

Figure 3. Trends in physical therapy health human resource ratios in the United States and Canada (number of physical therapists [# PT] per 10,000 people). Times were as follows: 1—1995 for United States and 1991 for Canada; 2—1999 for United States and 2000 for Canada; and 3—2005 for United States and for Canada.

158

f

Physical Therapy

Volume 89

Number 2

downey and Berry29 would indicate a 25.6% increase over a 30-year period. The trend for an increase of 61.3% across the 10-year period for physical therapists in the United States represents much more impressive growth for therapists than for nurses or physicians in the United States. As mentioned elsewhere, the reasons for such increases in HHR ratios for physical therapists were not directly explored in the present study. However, according to Alameddine et al,30 HHR ratios may be affected by production variables (eg, education and training) and policy variables (eg, health system change). In terms of production, or the rates of educating new practitioners, there were 13 education programs in Canada during the period from 1995 to 2005. In 1995, the production of new physical therapist graduates in Canada was 665; this value dropped to 620 in 2000 and then slightly rebounded to 631 in 2005.19 Thus, there was a 5.1% overall decline in the number of new graduates between 1995 and 2004.19 Although the precise reasons for these findings are not yet fully known, they may be related to the transition toward masters degree entry-to-practice programs, in which university programs decrease or limit enrollment during a transition period. The production of physical therapists in Canada in the period from 2005 on may be quite different because a new physical therapist education program was developed in the province of Quebec, and some training programs have doubled their enrollment since 2005. In contrast, in the United States, there are 210 physical therapist education programs that graduate approximately 6,000 physical therapists per year (Marc S Goldstein, EdD, Director of Research Services, American Physical Therapy Association; personal communication; September 24, February 2009

Physical Therapy Health Human Resource Ratios 2008). Although the populations of the United States and Canada are vastly different (ie, the population of Canada is 10% that of the United States), the United States has an appreciably higher production of physical therapists than Canada, and the number of educational programs in the United States also far exceeds the number in Canada. We found no publicly available reports that outline national trends in new physical therapist graduates in the United States. The entry-to-practice physical therapy degree in the United States during the period from 1995 to 2005 shifted from a Master of Physical Therapy to a Doctor of Physical Therapy. In contrast, in Canada, the entry-level degree shifted from a Bachelor of Physiotherapy to a Master of Science in Physical Therapy. The level of education or training may not have a direct effect on the supply of physical therapists in either country, but the United States may ultimately be attracting more potential students into a “doctoring” profession, whereas Canada’s programs grant master’s degrees. Although it is speculative and surely requires further research, the presence of doctorate-level training may ultimately increase demand by potential students and result in higher production. Other factors related to HHR production include rates of attrition (rates at which practitioners leave the profession), the movement of practitioners into and out of the workforce, and the influx of foreign-trained practitioners.31 No national-level analysis that explored these factors within the physical therapy profession was found in either the United States or Canada. As such, the extent to which graduates of physical therapy programs move out of the workforce (and how many move into the workforce) is not known, and the rates of February 2009

foreign-trained physical therapists entering the profession and contributing to the overall professional population in either country are not known. The absence of data on these 2 important factors related to HHR production represents an opportunity for further research. There are several relevant policy changes that may have affected HHR. A particularly relevant policy shift occurred in the United States during the study period. The Balanced Budget Act (BBA) of 1997 was a monumental policy shift that altered Medicare reimbursement. The history and impact of the BBA on publicly funded physical therapy reimbursement have been described by Enchelmayer et al32 and Latham et al33 and will not be repeated here; however, it is important to remark that the BBA has been described as the most important change in Medicare since its inception in 1965.34 In brief, the BBA was implemented as a way in which to reduce rapid growth in Medicare post-acute-care expenditures, including chronic care services provided by physical therapists (and other health care professionals, such as occupational therapists and speech-language pathologists), in various settings across the health care continuum. Enchelmayer et al32 conducted a survey of physical therapists in Florida in 1999 and concluded that implementation of the BBA resulted in reductions in employment settings and human resources in physical therapy. Latham et al33 used national-level data that seemed to refute the findings of Enchelmayer et al32 by reporting that, in aggregate, there was no decline in physical therapist services for people with conditions for which rehabilitation services were indicated in the United States. We suggest that, on the basis of our macro-level findings, the BBA may have had different effects depending

on the age distributions of residents in certain jurisdictions, particularly states with higher proportions of people eligible for Medicare. Jurisdictions such as Florida may have been most likely to have been affected by the BBA because of a higher proportion of people over the age of 65 years. The data from the present study indicated that Florida had a 24.27% decrease in the physical therapist-to-population ratio between 1995 and 1999. However, this period of decline was followed by an increase of 50.1% between 1999 and 2005 and an aggregated overall increase of 6.6% between 1995 and 2005. Six other states experienced declines in HHR ratios between 1995 and 1999: Arizona (6.3%), Illinois (2.0%), Nevada (15.4%), New York (3.5%), North Dakota (1.1%), and Oregon (7.8%). Like Florida, all of these states rebounded in the period from 1999 to 2005 and showed positive growth, ranging from a low of 29.9% in Oregon to a high of 91.3% in Nevada. The BBA may have had a shortterm effect on HHR ratios in some states with higher proportions of people eligible for Medicare, but between 1995 and 2005, all jurisdictions experienced aggregated positive growth at rates that surpassed those in Canada. Our data offer support for the findings of Latham et al33 in that there appeared to be no longlasting effects of the BBA on physical therapist HHR ratios across the United States. However, a policy factor that may have complicated the suggested relationship between reimbursement and HHR ratios was the 1999 Balanced Budget Refinement Act, which delayed the cap on outpatient physical therapy services.35 In previous policy research in Canada, Gordon et al36 and Paul et al37 explored the outcomes of a policy shift that may not have been as dramatic as the shift resulting from the BBA but that nonetheless significantly altered reimbursement for

Volume 89

Number 2

Physical Therapy f

159

Physical Therapy Health Human Resource Ratios publicly funded physical therapy services in the Canadian province of Ontario. Overall, the published findings regarding the partial delisting of publicly funded, community-based physical therapy services provided within a network of designated physical therapy centers in Ontario suggests that when policy changes are initiated to reduce reimbursement for publicly funded physical therapy services, the protective rebound effect may be a shift from public to private financing, resulting in some degree of human resource preservation. Although we believe that more research is needed to explore the short- and long-term effects of reimbursement policies, such as the BBA, on HHR strategies, we suggest that significant policy shifts affecting publicly funded physical therapy services may ultimately create the necessary underlying structure to encourage a shift from public to private financing for physical therapy services. Limitations of US/Canada HHR Comparative Analysis There are a number of inherent limitations of the present study and, although this research represents a foundation for additional research, there are details that must be considered. The sample of physical therapists in the present (US) study (as well as the sample in the Canadian study) included active and inactive physical therapists. We acknowledge that future investigations must separate active and inactive physical therapists, along with other factors, such as productivity measures, to yield more robust data regarding the clinical workforce. However, because the present study is the first, to our knowledge, to generate nationallevel data in the United States, it was most feasible to include both active and inactive registrants across the United States in our estimates of HHR ratios. Moreover, the methodological approach of including active 160

f

Physical Therapy

Volume 89

and inactive registrants has been used by others to estimate HHR ratios.2,4,7 We concede that there are limitations associated with this particular inclusion criterion, most notably, that active and inactive therapists would have different workloads and that therapists working full time also would differ in terms of productivity. Nevertheless, the integration of both groups allowed for moredirect comparisons of the 2 countries. Once again, we argue that this research provides the foundation for future research in an emerging investigative field. Another important limitation may have been the ways in which jurisdictions reported their data over the same time period. Further research should validate the ways in which national HHR ratios are calculated and reported; for instance, we do not know whether, over time, more physical therapists were registered in multiple jurisdictions. The fact that APTA provides the only publicly available consolidated list of physical therapists in the United States represents an inherent limitation of the present study. It also is important that the starting points for the trend analyses were different in the 2 countries. Another important limitation in the interpretation of HHR ratios is the “denominator effect”—that is, a slight change in the denominator can have an important effect on the numerator and thus the overall ratio. For instance, the HHR ratio can change because the numerator (physical therapists) increases at a higher or lower rate or because the denominator (population) increases at a higher or lower rate. A slight change in the denominator can have a greater effect than a change in the numerator. Thus, it will be instructive in future research to separate the change in the ratio into its com-

Number 2

ponent parts in order to more fully interpret the outcome.

Conclusion We have found that there are 6.2 physical therapists per 10,000 people across the United States. The findings of the present study also indicate that there are differences between the United States and Canada in terms of population growth, growth in the absolute number of physical therapists, and HHR ratios for physical therapists relative to the population. Overall, our analysis indicated that despite policy shifts and changes in education or training, there has been a larger increase in the HHR ratios for physical therapists relative to the population in the United States than in Canada. Moreover, our results underscore the need to examine the policy and environmental factors that drive the supply of physical therapists, to develop evidence-based targets or benchmarks regarding optimal HHR ratios, and to explore the effects of public and private funding on HHR ratios. Novel research methods must be developed within a health policy framework to plan for a future stable supply of physical therapists to meet emerging health care and rehabilitation demands across the United States and Canada. All authors provided concept/idea/research design, writing, and data analysis. Dr Landry and Ms Verrier provided data collection. Dr Landry holds a Career Scientist Personnel Award through the Ontario Ministry of Health and Long Term Care. This article was submitted March 12, 2008, and was accepted November 10, 2008. DOI: 10.2522/ptj.20080075

References 1 Duffield C, Franks H. Career paths beyond nursing and the contribution of nursing experience and skills in attaining these positions. Int J Nurs Stud. 2002;39:601– 609. 2 Zurn P, Dal Poz M, Stilwell B, Adams O. Imbalance in the health workforce. Hum Resour Health. 2004;2:13.

February 2009

Physical Therapy Health Human Resource Ratios 3 Dreesch N, Dolea C, Dal Poz M, et al. An approach to estimating human resource requirements to achieve the Millennium Development Goals. Health Policy Plan. 2005;20:267–276. 4 Diallo K, Zurn P, Gupta N, Dal Poz M. Monitoring and evaluation of human resources for health: an international perspective. Hum Resour Health. 2003; 3:113. 5 Health Care System. Health human resource strategy. Health Canada Web site. Available at: http://www.hc-sc.gc.ca/ hcs-sss/hhr-rhs/strateg/index_e.html. Modified October 11, 2006. Accessed November 24, 2008. 6 Stoddard GL, Barer M. Will increased medical school entry solve Canada’s physician supply problem? Can Med Assoc J. 1999;161:983–984. 7 Yang H, Bryck R, Doren W. Analysis of anesthesia physician supply: projected deficits in 2005. Can J Anaesth. 2000;47:179 –184. 8 Tepper J. The evolving role of Canadas family physicians, 19922001. Canadian Institute for Health Information Web site. Available at: http://secure.cihi.ca/cihiweb/ products/PhysiciansREPORT_eng.pdf. Published 2004. Accessed November 24, 2008. 9 OBrien-Pallas L, Baumann A, Donner G, et al. Forecasting models for human resources in health care. J Adv Nurs. 2001; 33:120 –129. 10 Alameddine M, Laporte A, Baumann A, et al. Where are nurses working? Employment patterns by sub-sector in Ontario, Canada. Healthc Policy. 2006;1:65– 86. 11 Registered Nurses Database. Workforce trends of registered nurses in Canada, 2005. Canadian Institute for Health Information Web site. Available at: http://secure. cihi.ca / cihiweb / products / ndb _ work force _ trends_registered_nurses_canada _2005_e.pdf. Published 2006. Accessed November 24, 2008. 12 Landry MD. Physical therapy human resources in Canada: 1991 to 2000. Physiother Can. 2004;56:39 – 42. 13 Landry MD, Ricketts TC, Verrier MC. The precarious supply of physical therapists across Canada: exploring national trends in health human resources (1991 to 2005). Hum Resour Health. 2007;5:23.

February 2009

14 North Carolina Health Professions Data System. Trends in licensed health professionals in North Carolina: 1979 –2005. Cecil G. Sheps Center for Health Services Research Web site. Available at: http:// www.shepscenter.unc.edu/hp/publications/ NCHPDS_27yrbk.pdf. Published June 2007. Accessed November 24, 2008. 15 Cheeseman Day J. Population profile of the United States. National population projections. US Census Bureau Web site. Available at: http://www.census.gov/pop ulation / www/pop-profile/natproj.html. Accessed November 24, 2008. 16 Barondess JA. Toward healthy aging: the preservation of health. J Am Geriatr Soc. 2008;56:145–148. 17 Tables by subject: population estimates and projections. Statistics Canada Web site. Available at: http://www40.statcan. gc.ca/l01/ind01/l3_3867_3433-eng.htm? hili_demo23. Modified November 24, 2008. Accessed November 24, 2008. 18 The sustainability report. Institute for Research and Innovation in Sustainability Web site. Available at: http://www.sust report.org. Published 2004. Accessed November 24, 2008. 19 Health Personnel Database. Health personnel trends in Canada, 1995 to 2004 (Revised July 2006). Canadian Institutes for Health Information Web site. Available at: http://secure.cihi.ca/cihiweb/products/ Health_Personnel_Trend_1995–2004_e. pdf. Published 2006. Accessed November 24, 2008. 20 Walsh D, Whyte B, Gordon DS. Changing places? A comparative analysis of areabased health trends in Scotland through the 1980s and 1990s. Public Health. 2007;121:889 – 897. 21 Mendez-Luck CA, Yu H, Meng YY, et al. Estimating health conditions for small areas: asthma symptom prevalence for state legislative districts. Health Serv Res. 2007;42:2389 –2409. 22 Wennberg JE. Practice variation: implications for our health care system. Manag Care. 2004;13(9 suppl):3–7. 23 Fisher ES, Wennberg JE. Health care quality, geographic variations, and the challenge of supply-sensitive care. Perspect Biol Med. 2003;46:69 –79. 24 Wennberg JE. On the appropriateness of small-area analysis for cost containment. Health Aff. 1996;15:164 –167. 25 Wennberg JE. Future directions for small area variations. Med Care. 1993;31(5 suppl):YS75–YS80.

26 Landry MD, Jaglal SB, Wodchis WP, et al. Rehabilitation services after total joint replacement in Ontario, Canada: can prehabilitation programs mediate an increasing demand? Int J Rehabil Res. 2007;30: 297–303. 27 Landry MD, Jaglal S, Wodchis WP, et al. Analysis of factors affecting demand for rehabilitation services in Ontario, Canada: a health policy perspective. Disabil Rehabil. In press. 28 Shih YT. Growth and geographic distribution of selected health professions, 19711996. J Allied Health. 1999;28:61–70. 29 McEldowney RP, Berry A. Physician supply and distribution in the USA. J Manag Med. 1995;9:68 –74. 30 Alameddine M, Laporte A, Baumann A, et al. “Stickiness” and “inflow” as proxy measures of the relative attractiveness of various sub-sectors of nursing employment. Soc Sci Med. 2006;63:2310 –2319. 31 Baumann A, Hunsberger M, Blythe J, Crea M. Sustainability of the workforce: government policies and the rural fit. Health Policy. 2008;85:372–379. 32 Enchelmayer KB, Hamby EF, Martindale C. The impact of the Balanced Budget Act of 1997 on the physical therapy profession. Health Care Man (Frederick). 2001; 19:58 – 69. 33 Latham NK, Jette AM, Ngo LH, et al. Did the 1997 Balanced Budget Act reduce use of physical therapy and occupational therapy services? Arch Phys Med Rehabil. 2008;89:807– 814. 34 Salsberg E, Rockey PH, Rivers KL, et al. US residency training before and after the 1997 Balanced Budget Act. J Am Med Assoc. 2008;300:1174 –1180. 35 Rao P. Selecting a rehabilitation program for people with stroke. Clin Geriatr Med. 1999;15:857– 868. 36 Gordon M, Waines B, Englehart J, et al. The consequences of delisting publiclyfunded, community-based physical therapy services in Ontario: a health policy analysis. Physiother Can. 2007;59:58 – 69. 37 Paul J, Park L, Ryter E, et al. Delisting publicly-funded community-based physical therapy services in Ontario, Canada: a 12-month follow-up study. Physiother Theory Pract. 2008;24:329 –343.

Volume 89

Number 2

Physical Therapy f

161

Research Report

External Focus Instructions Reduce Postural Instability in Individuals With Parkinson Disease Gabriele Wulf, Merrill Landers, Rebecca Lewthwaite, Thomas To ¨ llner G Wulf, PhD, is Professor, Department of Kinesiology and Nutrition Sciences, University of Nevada, Las Vegas, 4505 Maryland Pkwy, Las Vegas, NV 89154-3034 (USA). Address all correspondence to Dr Wulf at: [email protected]. M Landers, PT, DPT, OCS, is Associate Professor, Department of Physical Therapy, University of Nevada, Las Vegas. R Lewthwaite, PhD, is Director, Research and Education in Physical Therapy, and Director, Rehabilitation Outcomes Management, Rancho Los Amigos National Rehabilitation Center. She also is Adjunct Faculty, Department of Biokinesiology and Physical Therapy, University of Southern California, Los Angeles, California. T To¨llner, PhD, is Assistant Professor, Department of NeuroCognitive Psychology, LudwigMaximilians University, Munich, Germany. [Wulf G, Landers M, Lewthwaite R, To¨llner T. External focus instructions reduce postural instability in individuals with Parkinson disease. Phys Ther. 2009;89:162–168.] © 2009 American Physical Therapy Association

Background. Postural instability while standing, walking, and interacting with objects or the environment places individuals with Parkinson disease (PD) at risk for falls, injuries, and self-imposed restrictions in activity. Recent research with motor skills, including those demanding postural stability, has demonstrated performance and learning advantages when performers are instructed to adopt an external rather than an internal focus of attention. Despite the potential benefits in stability-related risk reduction and enhanced movement effectiveness, attentional focus research in individuals challenged with postural instability is limited.

Objective. The present translational research study examined the generalizability of the attentional focus effect to balance in older adults with PD.

Design. A within-participant design was used to account for potentially substantial individual variations in balancing capabilities. Methods. Fourteen participants diagnosed with idiopathic PD (Hoehn and Yahr stages II and III) participated in the experiment. They were asked to balance on an unstable surface (inflated rubber disk). In counterbalanced orders, they were instructed to focus on reducing movements of their feet (internal focus) or the disk (external focus), or they were not given attentional focus instructions (control).

Results. The adoption of an external focus resulted in less postural sway relative to both internal focus and control conditions. There was no difference between the internal focus and control conditions. Limitations. Mental functioning was not formally assessed, and comprehensive clinical profiles of participants were not obtained. Conclusions. The results are consistent with previous findings on attentional focus in samples of patients and people without disabilities. Subtle wording distinctions that direct attention to movement effects external to the mover reduce postural instability during standing for individuals with PD relative to an internal focus. The findings have potentially important implications for instructions given by clinicians and the reduction of fall risk.

Post a Rapid Response or find The Bottom Line: www.ptjournal.org 162

f

Physical Therapy

Volume 89

Number 2

February 2009

Parkinson Disease and Focus of Attention

P

ostural instability is a cardinal feature of Parkinson disease (PD)1 and a primary risk factor for falls.2 About two thirds of individuals with PD reported falling within the past 12 months,3 and 90% of people with PD will fall at some point in their lives.4 Some falls lead to severe injuries (eg, head injuries, fractures), which may result in hospitalization5 or further limitations to mobility. In one study,6 a cohort of people with PD experienced a hip fracture within 10 years after diagnosis—a 20-fold increase in risk compared with an age-matched control group. Although most falls do not result in serious injuries, they will, at the least, affect the individual’s confidence and quality of life.1 Those factors illustrate the need for developing interventions that can enhance balance in people with PD and, consequently, reduce their risk for falls. One research approach supports the benefits of providing external cues in individuals with PD. External cues such as visible target lines on the floor or paced auditory signals improve performance in movements during activities such as gait, buttonpressing, and handwriting.7–9 The present study originates from a separate line of work in motor learning10 that focuses on the preperformance wording of instructions provided to movers, rather than the continuing presence of external cues. Numerous studies over the past 10 years have shown that an individual’s focus of attention has an important influence on both the performance and learning of a variety of motor skills, including balance skills.10 Specifically, instructions that direct a performer’s attention to the effects that his or her movements have on the environment (external focus) have been demonstrated to lead to more-effective learning than directing attention to the movements themselves (internal focus). For example, in learning to balance February 2009

on a moving platform (stabilometer), asking participants to direct their attention to markers attached to the balance platform in front of their feet (external focus) has been shown to be more beneficial for learning than directing attention to the feet themselves (internal focus).11,12 It should be noted that participants were instructed not to look at the markers or their feet, but rather to look straight ahead and to simply concentrate on the markers or feet, respectively. Advantages of adopting an external focus also have been found for a variety of other balance skills13,14 as well as for sport and movement skills such as golf,15 tennis,16 volleyball and soccer,17 and vertical jumping.18 Importantly, external focus benefits have been shown not only relative to internal focus conditions but also relative to control conditions.11,19,20 This finding suggests that, left to their own devices, people may direct their attention to less-optimal (possibly internal) foci. Studies of patient populations also have demonstrated benefits of inducing an external focus.21,22 Fasoli et al21 examined the impact of attentional focus on the performance of reaching tasks in patients who had a cerebrovascular accident and in agematched control participants without impairments. They found that both groups performed various common tasks (eg, taking an apple off a shelf and putting it into a basket, moving an empty coffee mug from a table onto a saucer) more effectively if given external rather than internal focus instructions. Specifically, movement times were shorter and peak velocities were greater on all tasks, suggesting that these patients as well as the control participants preplanned their movements to a greater extent and used more automatic control processes when they focused externally.

Mediolateral posturographic differences have been found between individuals with idiopathic PD and controls who were healthy, even in quiet standing with eyes open.23 A recent study19 showed that, for individuals with PD and a history of falls, balance was improved when participants were given external focus instructions as opposed to an internal focus instructions or no instructions. Specifically, in that study, the postural stability of individuals with PD was measured in 3 conditions: (1) standing quietly with eyes open on a stable support surface, (2) standing quietly with eyes closed on a stable support surface, and (3) standing with eyes open on a sway-referenced support surface that tilted forward or backward in accord with shifts in the individual’s center of mass. Under all 3 conditions, participants were instructed to focus on rectangles under their feet (external focus) or on their feet (internal focus), or they were not given any focus instructions (baseline). No differences among attentional focus conditions were found on the relatively easy tasks with a stable support surface (conditions 1 and 2). However, on the more challenging task with the swayable support surface, an external focus produced less sway than both internal focus and control conditions. Significant external focus benefits were only found for individuals with a history of falls, whereas this effect did not reach significance for those without a history of falls. Thus, that study19 provided preliminary evidence for improved balance through external focus instructions in individuals with PD. The fact that performance advantages were found only for the most difficult condition (sway-referenced support surface) and for those participants who were most challenged (those with a history of falls) is in line with recent evidence that a certain degree of relative task difficulty is a precondition

Volume 89

Number 2

Physical Therapy f

163

Parkinson Disease and Focus of Attention for attentional focus effects to occur.24 Nevertheless, we deemed it important to have additional evidence for the influence of attentional focus on balance for those affected by PD. Therefore, the purpose of the present study was to replicate the external focus advantages found by Landers et al,19 using a different and uniformly more-challenging balance activity. Participants were asked to balance on an unstable surface (inflated rubber disk). We measured the amount of postural sway under different attentional focus conditions. Postural sway, or instability, has been shown to increase with age,25 to be higher in “fallers” than in “nonfallers,”26,27 and to increase with the demands of the balance task24 or of a secondary task in elderly people25 and people with a history of falls.26 We used a within-participant design to account for potentially substantial individual variations in balancing capabilities. Participants, in counterbalanced orders, were instructed to focus on reducing movements of either the disk (external focus condition) or their feet (internal focus condition), or they were not given attentional focus instructions (control condition). We expected to see more-effective balance, or less postural sway, under the external focus condition relative to internal focus and control conditions.

Method Participants Fourteen (10 male, 4 female) community-dwelling individuals who were diagnosed with idiopathic PD by a neurologist participated in this experiment. They were recruited from a local PD support group and were aged 52 to 80 years (mean⫽ 71.1). As all of the participants were Hoehn and Yahr stage II or III,28 they exhibited bilateral PD involvement, with minimal to moderate balance impairment. None of the partici164

f

Physical Therapy

Volume 89

pants exhibited dyskinesias. All participants were independent with ambulation (ie, no use of assistive devices); however, 7 of the participants had a history of falling within the past year. All participants followed their normal medication regimen during testing. Participants were excluded from the study if they had symptoms of dizziness or lightheadedness or other neurological or orthopedic disorders that would have negatively affected balance. Participants also were excluded if they had young-onset PD (⬍50 years of age), parkinsonian disorders (progressive supranuclear palsy, ShyDrager syndrome, corticobasal degeneration, nigrostriatal degeneration, olivopontocerebellar atrophy, secondary parkinsonism, or familial parkinsonism), or a history of dementia as reported by the family or caregiver or if they were unable to stand unassisted for 10 minutes without an assistive device. In addition, the participants’ ability to follow simple directions, as determined by their responses to questions and instructions, was informally assessed throughout the interview and consent process. Informed consent was received from each participant. Apparatus, Task, and Procedure Participants were scheduled for their testing approximately 1 hour after their medication had been taken. Testing was conducted at each participant’s home, using a portable force platform (model 9286AA*) situated under the rubber disk. The task required participants to balance on an inflated rubber disk (Disc ‘O’ Sit†) with a diameter of 33.02 cm (13 in). The disk was placed on the force platform to record data on center of pressure (COP). Participants were

* Kistler Instruments AG, Winterthur, Switzerland. † Perform Better, 11 Amflex Dr, Cranston, RI 02921.

Number 2

instructed to look straight ahead while balancing on the disk (similar to procedures used in previous studies14,24). A repeated-measures, within-subject design was used to assess the differences among conditions. Each participant performed four 15-second trials under each of the 3 attentional focus conditions. Specifically, participants were instructed to “stand still” (control condition), to “focus on minimizing movements of your feet” (internal focus condition), or to “focus on minimizing movements of the disk” (external focus condition). The order of attentional focus conditions was counterbalanced across participants to control for possible order and carryover effects. Participants were randomly assigned to perform under 1 of 3 orders of attentional focus conditions: control-internalexternal, internal-external-control, external-control-internal.14,24 Precautions were taken because of participants’ known balance impairments. A standard wheel-less walker was placed around the balance platform for safety. In addition, a spotter provided standby assistance from behind the participant using a standard gait belt in the event of a fall (Fig. 1). Dependent Variables and Data Analysis Center-of-pressure data were recorded at 500 Hz. The data were converted to ASCII format and processed using custom-designed laboratory software. The COP data were adjusted so that the central coordinates were (0, 0). Data then were converted from Cartesian to polar coordinates with the magnitude vector analyzed by calculating the root-mean-square error (RMSE). The RMSE of the COP vector magnitude served as a measure of postural sway. The RMSE is a commonly used measure of postural February 2009

Parkinson Disease and Focus of Attention sway14,24,25,29 –31 that represents the amount of postural sway, or sway area. The relationship of COP RMSE to functional abilities is unknown, although COP sway area recently has been identified as a risk factor for falls.2 Most participants were not able to complete all 15 seconds of each trial. That is, they would lose their balance and had to support themselves by holding on to the walker, or they had to be supported by the spotter. Therefore, we analyzed the longest segment of each trial during which the participant was able to stand on the disk without support. The average length of the analyzed segments was similar for the 3 attentional focus conditions (see “Results” section). The RMSE on each trial (independent of whether the entire 15 seconds or a shorter segment was analyzed) was submitted to a 3 (attentional focus condition) ⫻ 4 (trial) analysis of variance (ANOVA) for repeated measures on both factors. As all participants performed under all attentional focus conditions (internal, external, control), there was a possibility that carryover effects might occur. That is, performance under a given condition could be influenced by the previous condition. In contrast, if participants were able to adopt the instructed attentional focus, no such carryover effects should occur. To assess the possible influence of carryover effects, we conducted additional analyses, with attentional focus condition order included as a factor: 2 (group) ⫻ 3 (attentional focus condition) ⫻ 3 (attentional focus condition order: control-internal-external, internalexternal-control, external-controlinternal) ANOVAs for repeated measures on the last 2 factors.

Results The average length of the analyzed trial segments was 11.9 seconds February 2009

(SD⫽4.0, range⫽4 – 15). The length did not differ significantly among attentional focus conditions (control: 11.9 seconds, internal focus: 11.7 seconds, external focus: 12.0 seconds; F2,26⫽⬍1). In addition, there was no effect of trial (F3,39⫽⬍1) or interaction of focus and trial (F6,68⫽⬍1). Figure 2 shows the mean RMSE for each attentional focus condition. Postural sway was less under the external focus condition than it was under either the internal focus or control condition. The main effect of atFigure 1. tentional focus was sigParticipant performing the balance task. nificant (F2,26⫽5.07, P⬍.05, Eta2⫽0.28, observed power⫽.77). Post hoc tests (Fisher least significant when they were instructed to focus difference tests) indicated that RMSEs on their feet (internal focus condiwere smaller for the external focus tion) or when they were not given condition relative to the internal focus focus instructions (control condicondition (P⬍.001) and the control tion). Thus, on this relatively chalcondition (P⬍.05), whereas there was lenging task, participants demonno difference between the internal fo- strated the typical attentional focus cus and control conditions. The main effect (found in more than 50 studeffect of trial (F3,39⫽1.27, P⬎.05) and ies10), with the external focus condithe interaction of attentional focus tion resulting in more-effective percondition and trial (F6,78⬍1) were not formance than either the internal significant. The interaction between focus or control condition and with attentional focus condition and condi- no difference between the latter 2 tion order was not significant conditions.11,18,20 These results repli(F4,22⬍1). Thus, participants re- cate the external focus advantages sponded to the distinction between seen in previous studies with young instructional sets, and there were no adults who were healthy10 and with carryover effects among attentional fo- people affected by neurological discus conditions. orders19,21 or other disorders22 affecting balance. The fact that an exDiscussion ternal focus reduced postural sway Participants with PD showed en- in participants with PD may have imhanced balance when they were in- portant implications for balance structed to focus on the disk (exter- training, particularly considering renal focus condition) compared with cent findings showing that increased Volume 89

Number 2

Physical Therapy f

165

Parkinson Disease and Focus of Attention Because the former type of attentional focus increased walking speed, while the latter type of focus degraded it, she concluded that “the general suggestion that directing learners’ attention to the effects of their movements be incorporated into rehabilitation practice. . .may not be appropriate in all circumstances for people with PD.”32(p98)

Figure 2. Magnitude of sway (root-mean-square error [RMSE]) for participants with Parkinson disease as a function of the type of attentional focus (control, internal, or external). Error bars indicate 95% confidence intervals.

postural sway is associated with increased risk for falling in people with PD.2 It might be expected that individuals would spontaneously adopt the optimal focus of attention. Interestingly, however, this does not seem to be the case. Several studies11,19,20 as well as the present study have shown that when participants are not given attentional focus instruction and are left to adopt their own focus (control conditions), their performance is typically similar to that seen under internal focus conditions and less effective than under external focus conditions. These findings suggest that individuals tend to choose a less-than-optimal type of focus. One reason for this might be that individuals adopt the content of self-instructions that are modeled by others. In practical settings that involve the learning or relearning of motor skills (eg, sports, music, physical therapy), instructions that refer to the performer’s body movements are common. It may not be surprising, therefore, that individuals spontaneously focus on their own movements. Furthermore, people are presumably inclined to be relatively cautious when confronted with novel and complex motor tasks, es166

f

Physical Therapy

Volume 89

pecially those involving balance. The problem is that this cautiousness does not result in optimal performance. Ironically, it even exacerbates postural instabilities and balance problems. It should be noted that, in apparent contrast to our findings, one study involving individuals with PD32 purported to show performance benefits of an internal focus. Canning32 examined how directing attention affected participants’ gait when carrying a tray with glasses. Specifically, she instructed her participants to either focus on “maintaining big steps while walking” or focus on “balancing the tray and glasses.” She found that when participants focused on walking, they walked faster and with longer strides compared with a baseline condition without focus instructions and compared with when they focused on the tray and glasses. In contrast, when participants focused on balancing the tray and glasses, they walked more slowly and with shorter strides than under the baseline condition. Canning argued that the instructions not only directed attention to one task (walking) or the other (carrying the tray), but also induced an internal focus (walking) or an external focus (carrying the tray).

Number 2

This conclusion, however, is problematic. Importantly, both walking and carrying a tray can be executed under either external or internal attentional focus conditions (walking: distance covered/stride targets on floor or coordination of body segments, respectively; tray carrying: tray surface/objects on tray or hands holding the tray, respectively). Without further insight into the specific foci adopted by participants, it cannot be concluded on the basis of activity type that an internal focus of attention was better than an external focus of attention. Thus, although the conditions and results of the study by Canning32 do not pose a challenge to the contention that instructions inducing an external focus can enhance balance in people with PD, they do show that there is a need to further examine this issue and to extend the range of balance tasks used in those studies. Furthermore, the particular external foci that might be chosen to optimize performance for any given task remain to be determined empirically. Some work suggests that a focus on a more-distal movement effect results in better performance than effects just outside the body.33 Of potential value to the ultimate construction of effective balance training programs, however, is the sorting out of the theoretical or explanatory basis of the superiority of external cueing7,8 and external focus of attention effects found here. Interestingly, the benefits of external cues are attributed by some authors8,34 to February 2009

Parkinson Disease and Focus of Attention the desirability of using conscious control mechanisms to guide the movements of individuals with PD whose automatic movement control capacities have been reduced due to basal ganglia damage; external cues are said to render the tasks less regulated by automatic control.8,34 In contrast, attentional focus researchers10 contend that instructions that direct the performer’s attention to external movement-related effects act to support a more automatic form of motor control, consistent with that seen from expert performers. Internal foci, such as body movements, are thought to be associated with more-conscious and lessautomatic, and presumably lesseffective, forms of motor control.10,12 The explanatory difference may be a function of differing definitions of the concept of automaticity. Some authors8,34 appear to link the concept of automaticity to “mindless” movements associated with less forethought or systematic planning for potentially risky actions. Attentional focus and motor control researchers use the term “automaticity” to refer to the relatively effortless governance of wellcoordinated, fluent movements directed at environmental goals that is demonstrated by expert performers or well-practiced movers. The explanatory difference also might result from an assumption that the intrinsic regulation of sequential movements by the basal ganglia cannot be adequately compensated for by the parkinsonian nervous system, although recent evidence suggests that this assumption may need revision.35 Support for the beneficial automaticity notion comes from studies showing faster probe reaction times, indicating reduced attentional demands (or a greater degree of automaticity), as well as higherfrequency movement adjustments when individuals adopted an external rather than internal focus.12 Both February 2009

are seen as an indication of a more automatic, reflex-type mode of control that is based on faster and more finely tuned integrated movement responses. In addition, electromyographic activity connected with superior performance has been found to be reduced with an external focus.36 This is seen as an indication of greater movement efficiency, presumably accomplished through more discriminate motor unit recruitment and a reduction of noise in the motor system that hampers fine movement control.37 One limitation of this study is that no standardized method of formally assessing mental function was used. As mental focus was a key component of our research design, it is important to ensure that all of the participations have the ability to follow directions and maintain focus. Future researchers might consider formal assessment of cognitive capability (eg, Mini-Mental State Exam38) to exclude participants with substantial mental impairment. Additionally, researchers should consider posing post-trial questions to the participants to ensure that the instructions are understood and the focus is maintained. In this translational research39 study aimed at evaluating potential generalizability of a body of research to this clinical population, we did not obtain the kind of comprehensive clinical profile of participants that would be important in a trial of a clinical intervention. Other clinical measures of balance function, balance confidence, and PD severity would be helpful in characterizing the individuals who might benefit from the application of this research with clinical intervention intent.

Conclusions Future studies should examine the relative permanency of attentional focus effects (ie, learning), as well as the potential transfer to novel

skills.22,40 As the goal of balance enhancement training is to prepare individuals to more safely and effectively meet the demands of the kinds of balance situations that they encounter during their daily lives, determining the sustainability and generalizability of the practice setting training effects to less-supervised home and community contexts would be important. The potential for individuals with PD at varying levels of severity to regulate or selfmanage through self-instruction their own external attentional focus in posture and movement activities deserves investigation and may have the concomitant effect of enhancing a sense of control in their lives. Caregiver as well as clinician training in optimal attentional focus instructions may be of additional benefit in risk reduction for individuals with PD. Finally, it might be fruitful to examine whether the effects generalize to situations in which the individual is prevented from adopting the specific focus on balance used during practice, by using dual-task procedures.13,41 Given that, in real-life situations, people will not always maintain an external focus, it would be important to determine whether the effects persist when the focus in no longer cued or consciously adopted. Dr Wulf provided concept/idea/research design, data analysis, project management, and facilities/equipment. Dr Wulf and Dr Lewthwaite provided writing. Dr Wulf, Dr Landers, and Dr To¨llner provided data collection. Dr Landers provided participants. Dr Lewthwaite provided consultation (including review of manuscript before submission). All experimental work was carried out under the approval of the University of Nevada Institutional Review Board. This article was received February 7, 2008, and was accepted October 28, 2008. DOI: 10.2522/ptj.20080045

Volume 89

Number 2

Physical Therapy f

167

Parkinson Disease and Focus of Attention References 1 Playfer JR. Falls and Parkinson’s disease. Age Ageing. 2001;30:3– 4. 2 Matinolli M, Korpelainen JT, Korpelainen R, et al. Postural sway and falls in Parkinson’s disease: a regression approach. Mov Disord. 2007;22:1927–1935. 3 Ashburn A, Stack E, Pickering AM, Ward CD. A community-dwelling sample of people with Parkinson’s disease: characteristics of fallers and non-fallers. Age Ageing. 2001;30:47–52. 4 Koller WC, Glatt S, Vetere-Overfield B, Hassanien R. Falls and Parkinson’s disease. Clin Neuropharmacol. 1989;12:98 –105. 5 Temlett JA, Thompson PD. Reasons for admission to hospital for Parkinson’s disease. Intern Med J. 2006;36:524 –546. 6 Johnell O, Melton LJ III, Atkinson EJ, et al. Fracture risk in patients with parkinsonism: a population-based study in Olmsted County, Minnesota. Age Ageing. 1992;21: 32–38. 7 Morris ME, Iansek R, Matyas TA, Summers JJ. Stride length regulation in Parkinson’s disease: normalization strategies and underlying mechanisms. Brain. 1996;119: 551–558. 8 Oliveira RM, Gurd JM, Nixon P, et al. Micrographia in Parkinson’s disease: the effect of providing external cues. J Neurol Neurosurg Psychiatry. 1997;63:429 – 433. 9 Rogers MA, Phillips JG, Bradshaw JL, et al. Provision of external cues and movement sequencing in Parkinson’s disease. Motor Control. 1998;2:125–132. 10 Wulf, G. Attention and Motor Skill Learning. Champaign, IL: Human Kinetics Inc; 2007. 11 Wulf G, Ho ¨ ␤ M, Prinz W. Instructions for motor learning: differential effects of internal versus external focus of attention. J Mot Behav. 1998;30:169 –179. 12 Wulf G, McNevin N, Shea CH. The automaticity of complex motor skill learning as a function of attentional focus. Q J Exp Psychol A. 2001;54:1143–1154. 13 Totsika V, Wulf G. An external focus of attention enhances transfer to novel situations and skills. Res Q Exerc Sport. 2003;74:220 –225. 14 Wulf G, Mercer J, McNevin NH, Guadagnoli MA. Reciprocal influences of attentional focus on postural and supra-postural task performance. J Mot Behav. 2004;36: 189 –199. 15 Wulf G, Su J. An external focus of attention enhances golf shot accuracy in beginners and experts. Res Q Exerc Sport. 2007;78: 384 –389.

168

f

Physical Therapy

Volume 89

16 Wulf G, McNevin NH, Fuchs T, et al. Attentional focus in complex motor skill learning. Res Q Exerc Sport. 2000;71:229 – 239. 17 Wulf G, McConnel N, Ga¨rtner M, Schwarz A. Feedback and attentional focus: enhancing the learning of sport skills through external-focus feedback. J Mot Behav. 2002;34:171–182. 18 Wulf G, Zachry T, Granados C, Dufek JS. Increases in jump-and-reach height through an external focus of attention. International Journal of Sports Science & Coaching. 2007;2:275–284. 19 Landers M, Wulf G, Wallmann H, Guadagnoli MA. An external focus of attention attenuates balance impairment in Parkinson’s disease. Physiotherapy. 2005;91: 152–185. 20 Wulf G, Weigelt M, Poulter D, McNevin N. Attentional focus on supra-postural tasks affects balance learning. Q J Exp Psychol A. 2003;56:1191–1211. 21 Fasoli SE, Trombly CA, Tickle-Degnen L, Verfaellie MH. Effect of instructions on functional reach in persons with and without cerebrovascular accident. Am J Occup Ther. 2002;56:380 –390. 22 Rotem-Lehrer N, Laufer Y. Effect of focus of attention on transfer of a postural control task following an ankle sprain. J Orthop Sport Phys Ther. 2007;37:564 –569. 23 Mitchell SL, Collins JJ, De Luca CS, et al. Open-loop and closed-loop postural control mechanisms in Parkinson’s disease: increased mediolateral activity during quiet standing. Neurosci Lett. 1995;8:133–136. 24 Wulf G, To ¨ llner T, Shea CH. Attentional focus effects as a function of task difficulty. Res Q Exerc Sport. 2007;78: 257–264. 25 Huxhold O, Li S-C, Schmiedek F, Lindenberger U. Dual-tasking postural control: aging and the effects of cognitive demand in conjunction with focus of attention. Brain Res Bull. 2006;69:294 –305. 26 Shumway-Cook A, Woollacott M, Kerns KA, Baldwin M. The effects of two types of cognitive tasks on postural stability in older adults with and without a history of falls. J Gerontol Med Sci. 1997;52:M232– M240. 27 Melzer I, Benjuya N, Kaplanski J. Postural stability in the elderly: a comparison between fallers and non-fallers. Age Aging. 2004;33:602– 607. 28 Hoehn MM, Yahr M. Parkinsonism: onset, progression, and mortality. Neurology. 1967;17:427– 442.

Number 2

29 Balasubramaniam R, Riley M, Turvey MT. Specificity of postural sway to the demands of a precision task. Gait Posture. 2000;11:12–24. 30 Balasubramaniam R, Turvey M.T. The handedness of postural fluctuations. Hum Mov Sci. 2000;19:667– 684. 31 Riley MA, Baker AA, Schmit JM. Inverse relation between postural variability and difficulty of a concurrent short-term memory task. Brain Res Bull. 2003;62: 191–195. 32 Canning CG. The effect of directing attention during walking under dual-task conditions in Parkinson’s disease. Parkinsonism Relat Disord. 2005;11:95–99. 33 McNevin NH, Shea CH, Wulf G. Increasing the distance of an external focus of attention enhances learning. Psychol Res. 2003;67:22–29. 34 Cunnington R, Iansek R, Bradshaw JL. Movement-related potentials in Parkinson’s disease: external cues and attentional strategies. Mov Disord. 1999;14:63– 68. 35 Wu T, Hallett M. A functional MRI study of automatic movements in patients with Parkinson’s disease. Brain. 2005;128:2250 – 2259. 36 Vance J, Wulf G, To ¨ llner T, et al. EMG activity as a function of the performer’s focus of attention. J Mot Behav. 2004;36:450 – 459. 37 Zachry T, Wulf G, Mercer J, Bezodis N. Increased movement accuracy and reduced EMG activity as the result of adopting an external focus of attention. Brain Res Bull. 2005;67:304 –309. 38 Folstein MF, Folstein SE, McHugh PR. “Mini-mental state”: a practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res. 1975;12: 189 –198. 39 Woolf SH. The meaning of translational research and why it matters. JAMA. 2008;299:211–213. 40 Sidaway B, Anderson J, Danielson G, et al. Effects of long-term gait training using visual cues in an individual with Parkinson disease. Phys Ther. 2006;86:186 –194. 41 Laufer Y. Effect of cognitive demand during training on acquisition, retention and transfer of a postural skill. Hum Mov Sci. 2007 Oct 15 [Epub ahead of print].

February 2009

Parkinson Disease and Focus of Attention

Invited Commentary Wulf and colleagues1 are to be congratulated for adding new data to the emerging body of knowledge showing that the instructions given by physical therapists are a powerful determinant of motor performance in people with Parkinson disease (PD). Despite the finding that more than 4 million people worldwide have PD, the evidence for physical therapy interventions is still comparatively sparse. Most trials have described the signs, symptoms, and short-term progression of the disease rather than investigating how individuals respond to different physical therapy interventions. Because people with PD vary considerably with respect to their rate of progression, impairments, activity limitations, participation restrictions, and quality of life, there can be no single recipe for physical therapy treatment.2 Rather, each person needs to be assessed individually so that the physical therapist can provide a tailored program, suited to the needs of the individual and those of close family members. The “external focus of attention” advocated in Wulf and colleagues’ article is likely to be beneficial for some people with PD, and methods that require a person to focus his or her attention “internally” on movements or postural alignment might be useful for others.3 Throughout the 10- to 45-year time course of progression, physical therapy goals and strategies need to be adapted to ensure that the person receives interventions that are suitable at the time.2 The needs of a newly diagnosed person who is at stage I on the Hoehn and Yahr scale4 and has mild slowing and underscaling of movements are very different from those of the person at stage IV who has long-standing disease and is experiencing loss of balance, falls,

February 2009

Meg E Morris

hypokinesia, and possibly other movement disorders such as rigidity, tremor, dyskinesia, and dystonia. The “external focus of attention” strategy tested in the study by Wulf and colleagues is likely to be most suitable for people in the early to mid stages of PD, when cognition and the capacity for motor skill learning are not compromised. Wulf et al claim that they have shown that, in people at Hoehn and Yahr stage II or III, focusing attention on external cues from a moving support surface disk reduces postural sway. This finding supports existing evidence that people with PD can improve performance when they consciously attend to key aspects of a motor task or action sequence. For example, it is well known that step length improves when a person with PD focuses his or her attention on stepping over visual cues on the floor or thinks about walking with large steps.5 The seminal study by Behrman et al6 showed that the instruction set delivered to people with PD has powerful effects on performance. How the instructions are phrased before a task is performed has considerable impact on the quality and outcomes of a motor task. Likewise, the scholarly work by Canning et al3 demonstrated that when people with PD were instructed to divide their attention between competing tasks, the performance of the task not receiving attention deteriorated, whereas the primary task was performed relatively well. Several aspects of Wulf and colleagues’ investigation are controversial and warrant debate. The first is the choice of dependent variables selected for this trial. In upright standing, many people with PD have re-

duced postural sway due to hypokinesia (under-scaling of movement speed and size). Unlike people who have experienced a stroke, traumatic brain injury, or multiple sclerosis and have increased postural sway in standing, there is underscaling, over-constraint, and reduced variability in postural responses in people with PD. This over-constraint and reduced variability is argued to be a major contributing factor that predisposes people with PD to falls. Posture and movements are stereotyped and lack the usual flexibility and adaptability that enable a person to quickly adjust his or her posture and movement to changing task demands. Therefore, the aim of physical therapy in people with PD often is to increase the variability of performance to enable more adaptability, rather to constrain postural responses even further, as appears to be the aim of Wulf and colleagues in this article. A second concern is the type of motor task selected for analysis. The basal ganglia regulate the performance of well-learned, sequential motor skills, such as walking, turning, writing, speaking, swallowing, dressing, and turning over in bed. The performance of simple movements (such as steady standing) or novel tasks (such as standing on an inflatable disk, as in Wulf and colleagues’ study) is not compromised to the same extent as goal-directed activities. From the outset, it would be predicted that little difference in postural sway would be detected between focusing on the disk or the feet. Therefore, it is not surprising that “there was no difference between the internal focus and control conditions.” This finding probably was due to the novel nature of the

Volume 89

Number 2

Physical Therapy f

169

Parkinson Disease and Focus of Attention task, in addition to the small sample size. Although the findings of this study represent an important step forward, caution needs to be exercised when considering the generalizability of the findings to the population of people with PD as a whole. This is because only a small sample of 14 people with PD was tested and there were no individuals who were mildly affected (Hoehn and Yahr stage I) or, due to the nature of the task, individuals who were severely affected (Hoehn and Yahr stage V). To conclude, this trial on a novel laboratory-based task provides some

Author Response

f

Physical Therapy

Volume 89

ME Morris, PT, PhD, FACP, is Professor of Physiotherapy, The University of Melbourne, 3010, Melbourne, Victoria, Australia, and Director of Allied Health Research, Rehabilitation & Aged Care Program, Southern Health, Kingston Centre, Warrigal Rd, Cheltenham, 3192, Australia. Address all correspondence to Dr Morris at: [email protected]. DOI: 10.2522/ptj.20080045.ic

References 1 Wulf G, Landers M, Lewthwaite R, To ¨ llner T. External focus instructions reduce postural instability in individuals with Parkinson disease. Phys Ther. 2009;89:162–168. 2 Morris ME. Movement disorders in people with Parkinson disease: a model for physical therapy. Phys Ther. 2000;80:578 –597. 3 Canning CG. The effect of directing attention during walking under dual-task conditions in Parkinson’s disease. Parkinsonism Relat Disord. 2005;11:95–99. 4 Hoehn MM, Yahr M. Parkinsonism: onset, progression, and mortality. Neurology. 1967;17:427– 442. 5 Morris ME, Iansek R, Matyas TA, Summers JJ. Stride length regulation in Parkinson’s disease: normalization strategies and underlying mechanisms. Brain. 1996;119: 551–558. 6 Behrman AL, Teitelbaum P, Cauraugh JH. Verbal instructional sets to normalize the temporal and spatial gait parameters in Parkinson’s disease. J Neurol Neurosurg Psychiatry. 1998;65:580 –582.

Gabriele Wulf, Rebecca Lewthwaite, Merrill Landers, Thomas To ¨ llner

We thank Morris for her thoughtful commentary1 on our study.2 We concur with her in the observation that our study provides further support for the role that instructional sets or pretask instructions can play in enhancing the motor performance of individuals with Parkinson disease (PD).3,4 Given the constellation of cognitive and movement symptoms common in people with PD, as well as the relationship of cognitive processes to movement control,5 it is important to recognize that many individuals with PD retain their capabilities to distinguish between differing instructions and to act upon those distinctions motorically (eg, to increase stride length upon request). Our study provides additional evidence that individuals with mild to moderate disease severity (in the present study, at Hoehn and Yahr levels II and III; in the study by Behrman et al,3 Hoehn and Yahr levels II and III and a few with level IV staging) can distinguish between specific instructions and produce move-

170

insights into the nature of motor control deficits in people with PD. The physical therapy profession now awaits the results of large-scale controlled clinical trials that quantify the effects of therapeutic strategies routinely used by physical therapists, such as fall prevention, movement strategy training, and progressive resistance strength training.

ment nuanced to these instructional distinctions. Furthermore, evidence of instructional set effectiveness is consistent with a recent review that noted stronger support for gaitrelated physical therapy interventions that combined conventional physical therapy with the provision of external visual or auditory cueing than for conventional physical therapy alone in individuals with PD.6 It seems reasonable to suggest that future refinements in physical therapy interventions for people with PD may be enhanced by recognition of the potential of particular forms of instructions to promote more optimal control of movement. A relatively large and remarkably consistent body of research is accumulating to suggest that instructions to individuals to adopt an external focus of attention are more beneficial than are instructions inducing an internal focus. In our within-subject design, we found again that instructions to focus attention (concen-

Number 2

tration) on the intended movement effect (“focus on minimizing movements of the inflatable disk”) produced greater reduction in postural sway than instructions inducing an internal or body-related focus (“focus on minimizing the movement of your feet”) or the control condition with no explicit focus instruction. The finding that control conditions (if they are included) produce very similar performances as internal focus conditions has been consistent across studies. This finding may suggest that participants “naturally” tend to adopt an internal focus, and, more importantly, it indicates that an external focus results in enhanced performance or learning. Interestingly, clinicians have reported that individuals with PD may be aware of their need for external cueing; one individual with PD was observed to remove her belt and place it in a straight line on the floor ahead of her to assist in gait initiation (Carolee J Winstein, PT, PhD, FAPTA; personal communication; December 5, 2008).

February 2009

Parkinson Disease and Focus of Attention task, in addition to the small sample size. Although the findings of this study represent an important step forward, caution needs to be exercised when considering the generalizability of the findings to the population of people with PD as a whole. This is because only a small sample of 14 people with PD was tested and there were no individuals who were mildly affected (Hoehn and Yahr stage I) or, due to the nature of the task, individuals who were severely affected (Hoehn and Yahr stage V). To conclude, this trial on a novel laboratory-based task provides some

Author Response

f

Physical Therapy

Volume 89

ME Morris, PT, PhD, FACP, is Professor of Physiotherapy, The University of Melbourne, 3010, Melbourne, Victoria, Australia, and Director of Allied Health Research, Rehabilitation & Aged Care Program, Southern Health, Kingston Centre, Warrigal Rd, Cheltenham, 3192, Australia. Address all correspondence to Dr Morris at: [email protected]. DOI: 10.2522/ptj.20080045.ic

References 1 Wulf G, Landers M, Lewthwaite R, To ¨ llner T. External focus instructions reduce postural instability in individuals with Parkinson disease. Phys Ther. 2009;89:162–168. 2 Morris ME. Movement disorders in people with Parkinson disease: a model for physical therapy. Phys Ther. 2000;80:578 –597. 3 Canning CG. The effect of directing attention during walking under dual-task conditions in Parkinson’s disease. Parkinsonism Relat Disord. 2005;11:95–99. 4 Hoehn MM, Yahr M. Parkinsonism: onset, progression, and mortality. Neurology. 1967;17:427– 442. 5 Morris ME, Iansek R, Matyas TA, Summers JJ. Stride length regulation in Parkinson’s disease: normalization strategies and underlying mechanisms. Brain. 1996;119: 551–558. 6 Behrman AL, Teitelbaum P, Cauraugh JH. Verbal instructional sets to normalize the temporal and spatial gait parameters in Parkinson’s disease. J Neurol Neurosurg Psychiatry. 1998;65:580 –582.

Gabriele Wulf, Rebecca Lewthwaite, Merrill Landers, Thomas To ¨ llner

We thank Morris for her thoughtful commentary1 on our study.2 We concur with her in the observation that our study provides further support for the role that instructional sets or pretask instructions can play in enhancing the motor performance of individuals with Parkinson disease (PD).3,4 Given the constellation of cognitive and movement symptoms common in people with PD, as well as the relationship of cognitive processes to movement control,5 it is important to recognize that many individuals with PD retain their capabilities to distinguish between differing instructions and to act upon those distinctions motorically (eg, to increase stride length upon request). Our study provides additional evidence that individuals with mild to moderate disease severity (in the present study, at Hoehn and Yahr levels II and III; in the study by Behrman et al,3 Hoehn and Yahr levels II and III and a few with level IV staging) can distinguish between specific instructions and produce move-

170

insights into the nature of motor control deficits in people with PD. The physical therapy profession now awaits the results of large-scale controlled clinical trials that quantify the effects of therapeutic strategies routinely used by physical therapists, such as fall prevention, movement strategy training, and progressive resistance strength training.

ment nuanced to these instructional distinctions. Furthermore, evidence of instructional set effectiveness is consistent with a recent review that noted stronger support for gaitrelated physical therapy interventions that combined conventional physical therapy with the provision of external visual or auditory cueing than for conventional physical therapy alone in individuals with PD.6 It seems reasonable to suggest that future refinements in physical therapy interventions for people with PD may be enhanced by recognition of the potential of particular forms of instructions to promote more optimal control of movement. A relatively large and remarkably consistent body of research is accumulating to suggest that instructions to individuals to adopt an external focus of attention are more beneficial than are instructions inducing an internal focus. In our within-subject design, we found again that instructions to focus attention (concen-

Number 2

tration) on the intended movement effect (“focus on minimizing movements of the inflatable disk”) produced greater reduction in postural sway than instructions inducing an internal or body-related focus (“focus on minimizing the movement of your feet”) or the control condition with no explicit focus instruction. The finding that control conditions (if they are included) produce very similar performances as internal focus conditions has been consistent across studies. This finding may suggest that participants “naturally” tend to adopt an internal focus, and, more importantly, it indicates that an external focus results in enhanced performance or learning. Interestingly, clinicians have reported that individuals with PD may be aware of their need for external cueing; one individual with PD was observed to remove her belt and place it in a straight line on the floor ahead of her to assist in gait initiation (Carolee J Winstein, PT, PhD, FAPTA; personal communication; December 5, 2008).

February 2009

Parkinson Disease and Focus of Attention In the perhaps 50 studies that used appropriate experimental designs to compare the effectiveness of external versus internal foci (or control conditions), we have not come across any exceptions to the superiority of an external focus. Conditions in the present study were counterbalanced for order and each individual encountered all conditions within a relatively short time frame within a single testing session, ensuring that differences in movement effectiveness were not a function of which condition was experienced first, medication-related timing, cognitive capabilities, or other aspects of disease severity within the sample. It should be pointed out that an external focus directed at the intended movement effect facilitates the achievement of that effect or movement goal. This is independent of what that goal might be (eg, a reduction of postural sway during standing, quick movement adjustments to maintain equilibrium during dynamic balance tasks). In fact, evidence for the motor system’s capability to automatically achieve a desired movement outcome also comes from studies examining suprapostural task performance.7–9 These studies have shown that the healthy motor control system is able to optimize the control of posture as a function of the performer’s suprapostural task goal (eg, focusing visually, pointing at a target, holding an object still). Thus, tasks such as carrying a tray10 would not be considered as “competing” with the postural task of walking. Rather, the gait pattern would be expected to adjust as a function of the suprapostural task goal of keeping the tray still. Importantly, the achievement of the suprapostural goal and ensuing regulation of postural control can be enhanced further by inducing an external relative to an internal focus on the suprapostural task.11–13 Regardless of the use of a suprapostural February 2009

task or one with another movement goal, there will be some level of challenge, albeit different across individuals, that would be expected to tax and exceed their available attentional resources, resulting in poor performance on a secondary, and sometimes primary, task. This dualtask breakdown might be forestalled or limited with external attentional focus instructions—which have been shown to reduce attentional demands14— but this effect has yet to be demonstrated directly. We concur with Morris in noting that the generalizability of the attentional focus effect awaits further evaluation. We would think that evidence of sustainable effects and of the potential for portable selfinstruction3 for individuals at different stages in their disease time course would be of particular importance. The external attentional focus advantage has been found across a wide range of non-PD samples (from children to adults and novice learners to more-expert motor performers to individuals following stroke) and with a variety of movement tasks (from discrete sport skills to fundamental movements such as balance, jumping, and ballistic actions) with diverse movement requirements (from aiming accuracy to force production to reductions in postural sway, to productions of large movement amplitudes, to open and closed loop control).15 Therefore, we might speculatively predict that individuals with less impairment than that found in our present sample may find benefit in carefully worded external focus instructions. In addition, they may find it for movement tasks other than the type of balance task used in the present study, although the impact may be most evident for tasks that are of relatively moderate to high challenge for the individual. Clearly, however, extensions to individuals with greater impairment should be undertaken with caution;

this form of intervention invokes cognitive and attentional, as well as motoric, abilities that may all be impaired in people with PD.16 However, the benefits for those individuals at risk for movement disorders may make careful implementation and evaluation of particular value. Furthermore, attentional focus probably is most relevant as an integrated feature of otherwise effective interventions.6,17 Assumptions about the residual and latent capabilities of individuals with diseased neuromotor systems to mitigate impairment and enhance activity and participation will continue to be challenged,16,18 and principled use of available movement science seems important. As Rubinstein and colleagues noted about PD: “One of the most interesting features of PD is that, despite severe motor symptoms, patients are sometimes still able to perform complex movements almost normally under certain conditions. . . . It appears that the underlying problem has more to do with motor control than with actual motor function. The challenge lies in enabling patients to access this ability at will and to use it in normal, daily function.”6(p1149) DOI: 10.2522/ptj.20080045.ar

References 1 Morris ME. Commentary on “External focus instructions reduce postural instability in individuals with Parkinson disease.” Phys Ther. 2009;89:169 –170. 2 Wulf G, Landers M, Lewthwaite R, To ¨ llner T. External focus instructions reduce postural instability in individuals with Parkinson disease. Phys Ther. 2009;89:162–168. 3 Behrman AL, Teitelbaum P, Cauraugh JH. Verbal instructional sets to normalize the temporal and spatial gait parameters in Parkinson’s disease. J Neurol Neurosurg Psychiatry. 1998;65:580 –582. 4 Landers M, Wulf G, Wallmann H, Guadagnoli MA. An external focus of attention attenuates balance impairment in Parkinson’s disease. Physiotherapy. 2005;91: 152–185. 5 Yogev-Seligmann G, Hausdorff JM, Giladi N. The role of executive function and attention in gait. Mov Disord. 2007;23:329 – 342.

Volume 89

Number 2

Physical Therapy f

171

Parkinson Disease and Focus of Attention 6 Rubinstein TC, Giladi N, Hausdorff JM. The power of cueing to circumvent dopamine deficits: a review of physical therapy treatment of gait disturbances in Parkinson’s disease. Mov Disord. 2002;17:1148 – 1160. 7 Balasubramaniam R, Turvey MT. The handedness of postural fluctuations. Hum Mov Sci. 2000;19:667– 684. 8 Riley MA, Stoffregen TA, Grocki MJ, Turvey MT. Postural stabilization for the control of touching. Hum Mov Sci. 1999;18: 795– 817. 9 Stoffregen TA, Pagulayan RJ, Bardy BG, Hettinger LJ. Modulating postural control to facilitate visual performance. Hum Mov Sci. 2000;19:203–220.

172

f

Physical Therapy

Volume 89

10 Canning CG. The effect of directing attention during walking under dual-task conditions in Parkinson’s disease. Parkinsonism Relat Disord. 2005;11:95–99. 11 McNevin NH, Wulf G. Attentional focus on supra-postural tasks affects postural control. Hum Mov Sci. 2002;21:187–202. 12 Wulf G, Mercer J, McNevin NH, Guadagnoli MA. Reciprocal influences of attentional focus on postural and supra-postural task performance. J Mot Behav. 2004;36: 189 –199. 13 Wulf G, Weigelt M, Poulter DR, McNevin NH. Attentional focus on supra-postural tasks affects balance learning. Q J Exp Psychol. 2003;56:1191–1211. 14 Wulf G, McNevin N, Shea CH. The automaticity of complex motor skill learning as a function of attentional focus. Q J Exp Psychol A. 2001;54:1143–1154.

Number 2

15 Wulf G. Attention and Motor Skill Learning. Champaign, IL: Human Kinetics Inc; 2007. 16 Morris ME. Locomotor training in people with Parkinson disease. Phys Ther. 2006; 86:1426 –1435. 17 Farley BG, Koshland GF. Training BIG to move faster: the application of the speedamplitude relation as a rehabilitation strategy for people with Parkinson’s disease. Exp Brain Res. 2005;167:462– 467. 18 Fisher BE, Wu AD, Salem GJ, et al. The effect of exercise training in improving motor performance and corticomotor excitability in people with early Parkinson’s disease. Arch Phys Med Rehabil. 2008;89: 1221–1229.

February 2009

Research Report Harris Infant Neuromotor Test: Comparison of US and Canadian Normative Data and Examination of Concurrent Validity With the Ages and Stages Questionnaire Sarah Westcott McCoy, Alicia Bowman, Jessica Smith-Blockley, Katie Sanders, Antoinette M Megens, Susan R Harris

Background. The Harris Infant Neuromotor Test (HINT) was developed as a screening tool for potential motor and cognitive developmental disorders in infants. Scoring on the HINT has been shown to be reliable, and several studies have supported the validity of the HINT. Normative values for the tool have been developed using Canadian infants.

Objective. The aims of this study were (1) to further evaluate the validity of the HINT by comparing data obtained on US infants who were developing typically with data previously acquired on Canadian infants and (2) to determine the concurrent validity of the HINT with the Ages and Stages Questionnaire (ASQ). Secondary analyses of HINT scores for US white and nonwhite infants and for US infants who had parents with lower levels of education and US infants who had parents with higher levels of education (as a proxy for socioeconomic status [SES]) were conducted.

Design. Cross-sectional exploratory and quasi-experimental comparative research designs were used to evaluate the validity of the HINT. Methods. Sixty-seven infants from the United States who were developing typically and who were aged 2.5 to 12.5 months were recruited via convenience sampling. Sixty-four of these infants were compared with Canadian infants matched for age, sex, ethnicity or race, and parental education. The HINT was administered by raters who had been trained to attain acceptable levels of interrater reliability, and parents completed the ASQ. The HINT scores for US white versus nonwhite infants (n⫽46) and infants who had parents with lower SES versus a higher SES (n⫽52) were compared.

Results. There were no significant differences between HINT total scores for US and Canadian infants or for US racial or ethnic groups and SES groups. There were high correlations (r⫽⫺.82 to ⫺.84) between HINT and ASQ scores. Limitations. The study used a small US sample with limited geographical diversity. Small sample numbers also did not allow for comparisons of specific racial or ethnic groups. The SES groups were created primarily using parental education as a proxy for SES.

Conclusions. The results suggest that HINT screening in the United States is

S Westcott McCoy, PT, PhD, is Associate Professor, Department of Rehabilitation Medicine, University of Washington, 1959 Pacific NE St, Box 356490, Seattle, WA 98195-6490 (USA). Address all correspondence to Dr McCoy at: [email protected]. A Bowman, PT, DPT, is Physical Therapist, Clinic of Orthopaedic and Sports Physical Therapy, Tacoma, Washington. J Smith-Blockley, PT, DPT, is Physical Therapist, Therapeutic Associates, Boise, Idaho. K Sanders, PT, DPT, is Physical Therapist, Kindering Center, Bellevue, Washington. AM Megens, MSc, BScPT, is Physical Therapist, Ottawa Children’s Treatment Centre, Ottawa, Ontario, Canada. SR Harris, PT, PhD, FAPTA, is Professor Emerita, Department of Physical Therapy, Faculty of Medicine, University of British Columbia, Vancouver, British Columbia, Canada. [Westcott McCoy S, Bowman A, Smith-Blockley J, et al. Harris Infant Neuromotor Test: comparison of US and Canadian normative data and examination of concurrent validity with the Ages and Stages Questionnaire. Phys Ther. 2009;89:173–180.] © 2009 American Physical Therapy Association

supported on the basis of Canadian norms and the validity of the HINT in screening for motor and cognitive delays. Although there is preliminary support for the HINT as an appropriate screening tool for US infants who are nonwhite or who have parents with a lower SES, more research is warranted.

February 2009

Volume 89

Number 2

Post a Rapid Response or find The Bottom Line: www.ptjournal.org Physical Therapy f

173

Validity of HINT as Screening Tool for Motor and Cognitive Delays

R

esearch has suggested that the earlier services are provided for at-risk infants and their families, the better the overall developmental and family outcomes.1 Therefore, accurate screening and diagnostic tools are crucial for prompt identification of infants at risk for developmental delays or disabilities.1 Improvements in technology over the last few decades have led to increased survival of infants born very early and at very low birth weights.2,3 This decrease in infant mortality has led to an increase in the absolute number of children with neurodevelopmental disabilities.2 It has been suggested that current tests designed to screen for developmental disabilities have one or more problems that limit their clinical use.1 In response to this limitation, the Harris Infant Neuromotor Test (HINT) was developed for use in clinical and research settings to provide early screening for potential developmental disorders in infants.1,4 The HINT is a noninvasive tool designed for use with infants from 2.5 to 12.5 months of age.4,5 The HINT

also aims to validate the importance of parental opinions in making screening decisions.4 The HINT contains 4 general areas: background information on the child and the caregiver; questions assessing the caregiver’s level of concern about the infant’s movement and play; a 21-item observational or testing section that is scored to assess the infant’s movement against gravity, muscle tone, behavior and cooperation, stereotypical behaviors, and head circumference (Tab. 1); and 1 item in which the clinician notes his or her overall clinical impression of the infant’s development.6 Items aimed specifically at the identification of early cognitive or behavioral deficits include observation of the infant’s behavior and cooperation, the presence of stereotypical behaviors, and measurement of head circumference as an indication of brain volume. Overall, the test takes approximately 15 to 30 minutes to administer and score, depending on the infant’s behavior and the skill of the administrator, and is minimally stressful for the infant. Physical ther-

apists, occupational therapists, physicians, pediatric nurses, and other health care providers who have had appropriate training can administer the HINT. The interrater reliability (intraclass correlation coefficient [ICC]⫽.99), intrarater reliability (ICC⫽.98 –.99), and test-retest reliability (ICC⫽.98) of the HINT have been shown to be high.5,7 Through expert review, the content validity of the HINT was tested, and the reviewers agreed that most of the items were valid and that the items were not culturally biased.1 With regard to concurrent validity, comparison of the HINT with the Bayley Scales of Infant Development revealed good to excellent correlations (r⫽⫺.73 and ⫺.89 for the Mental Scale and the Motor Scale, respectively).5,6 The strong concurrent correlation of the HINT and the Bayley Mental Scale supports the validity of the HINT in detecting early cognitive deficits. Preliminary studies suggested that the HINT has stronger predictive validity than the Bayley-II Motor Scale,4 but more re-

Table 1. Harris Infant Neuromotor Test (HINT) Items Cognitive or Behavioral Development Items

Test Method

174

Motor Development Items

Observation (infant is observed when placed in or allowed to move independently into supine, prone, sitting, and standing positions)

● Behavior and cooperation ● Presence of stereotypical behaviors

● ● ● ● ● ● ● ● ● ● ●

Testing (infant is provided stimulation or is handled by the examiner to determine scores)

● Head circumference

● ● ● ●

f

Physical Therapy

Volume 89

Number 2

Mobility, supine Neck retraction, supine Eye muscle control Head position, prone Upper-extremity position, prone Head position, sitting Trunk position, sitting Locomotion and transition skills Posture of hands Posture of feet Frequency and variety of movements

Visual following Asymmetrical tonic neck reflex Reaching from supine position Passive range of motion in supine position ● Head righting in transition from supine to prone to supine positions ● Trunk mobility in transition from supine to prone to supine positions ● Passive range of motion in prone position

February 2009

Validity of HINT as Screening Tool for Motor and Cognitive Delays search is needed to confirm the actual predictive validity of the HINT. Most recently, the known-groups validity of the HINT was analyzed by comparing HINT total scores from a group of 54 high-risk infants with those from the current Canadian normative sample.8 For infants at ages 4, 5, 7, and 8 months, the HINT appeared to discriminate effectively between low-risk and high-risk infants. The HINT has been used primarily to screen for the development of Canadian infants. Normative data have been collected on 412 Canadian infants from 5 provinces. No data on infants from the United States have been published to determine whether HINT scores, reflective of infant developmental trajectories, are similar for infants in the 2 nations. Although the Canadian and US cultures related to caregiving for infants can be considered similar overall, there are environmental (colder climate in Canada) and health care (universal health care in Canada) differences that might affect infants’ development. Although no studies comparing the effects of US and Canadian early child-rearing practices on infant development could be found, it has been shown that subtle cultural preferences for sleep9 and play positions and the use of equipment10 might affect rates of motor development. Research with the HINT should be expanded to include infants from the United States, so that the use of the test can be generalized to this population. Because the authors of the HINT suggested that parental perceptions are important and that the HINT screens for early cognitive as well as motor delays, an assessment of concurrent validity with a parental report of cognitive and motor development could strengthen an evaluation of the validity of the HINT.11 The Ages and Stages Questionnaire (ASQ) was designed to provide an economical and February 2009

convenient alternative to professionally administered early infant or child assessments.6,12 Typically requiring 10 to 20 minutes to complete, the ASQ contains 30 questions covering 5 developmental domains— communication, gross motor, fine motor, problem solving, and personalsocial—at various ages (including 4, 6, 8, and 12 months). Scores below established cutoff scores in one or more domains suggest developmental delays in the child being assessed.12 Previous research showed that the ASQ is reliable when completed by parents or primary caregivers.6 Compared with established gold standards, the ASQ had a sensitivity of 90%, a specificity of 77%, a positive predictive value of 40%, and a negative predictive value of 98%.13 We concluded that the high negative predictive value of the ASQ supported its use as a screening tool for cognitive and motor delays.13 The ASQ should be comparable to the HINT, especially in the motor and problem-solving domains. No comparison of HINT scores for infants of different ethnicities or races and socioeconomic status (SES) has been published. As of 2005, it was estimated that 33% of the US population was nonwhite, with the projection that by 2010, approximately 35% of the US population will be of minority ethnicities or races.14,15 Previous research suggested that infants from different ethnic or racial and socioeconomic backgrounds may have different developmental patterns.16 –22 For example, a study of full-term Asian American infants found that 40% were identified as having scores on the Movement Assessment of Infants indicating high risk for movement disorders or delays, even though they were not suspected of having any developmental concerns and had unremarkable birth histories.17 Further,

both the cognitive development and the motor development of infants from African American and Hispanic backgrounds were shown to be correlated with the home environment, mother-infant interaction, and mother’s education.18,19 However, there are conflicting data on the lack of effect of a mother’s education on infant development as indicated by HINT scores.23 This information suggests that ethnicity or race and SES may together influence the rate of development of infants. Therefore, it is important to examine scores for infants who are from different ethnic or racial and socioeconomic backgrounds and who are developing typically to determine whether screening with the HINT will be valid for these populations. Two primary research questions were addressed in this study. First, is there a difference in HINT total scores between matched US and Canadian samples of infants who are developing typically? Second, what is the concurrent validity of the HINT with the ASQ in US infants? Our hypotheses were that HINT scores for US and Canadian infants would be comparable and that the HINT would correlate most highly with the problem-solving and motor domains of the ASQ. Secondary analyses were conducted for the US sample of infants to determine whether there were differences in HINT total scores for white versus nonwhite infants and for infants who had parents with lower SES versus higher SES. We hypothesized that there would be differences in HINT scores between these groups.

Method Participants Sixty-seven infants who were healthy, developing typically, and distributed across the age span for the HINT (2.5–12.5 months of age) were recruited from Washington, Idaho, and Hawaii via convenience

Volume 89

Number 2

Physical Therapy f

175

Validity of HINT as Screening Tool for Motor and Cognitive Delays Table 2. Distribution of US Sample No. of:

Boys

Girls

White Infantsa

Nonwhite Infantsb

Infants With High Socioeconomic Statusc

2–4 (19)

11

8

14

5

12

7

5–6 (19)

12

7

11

8

11

8

Age of Infant, mo (No. of Infants)

7–9 (13)

Infants With Low Socioeconomic Statusd

9

4

8

5

8

5

10–12 (16)

10

6

11

5

6

10

Total (67)

42

25

44

23

37

30

a

Both parents were white. One or both parents were not white. Parental education beyond high school. d Parental education of high school or less or participation in programs for families of low socioeconomic status. b c

sampling. On the basis of the assumption of a medium effect size and a power of .80, 64 subjects per group were needed to demonstrate a significant difference when a t test was used to examine differences between US and Canadian infants.24(p720) On the basis of the assumption of a correlation of at least .40, 46 subjects were needed to demonstrate significance at a power of .80.24(p724) Infants who were born full term (37– 42 weeks), weighed more than 2,500 g (5.5 lb), and had no history of major prenatal, perinatal, or postnatal medical complications or maternal complications were included in the study. Infants who were born preterm (⬍37 weeks of gestation), were small for their gestational age (⬍2,500 g), had a history of maternal alcohol or drug use during pregnancy, or had any other high-risk condition, such as a chromosomal abnormality or a congenital heart defect, were excluded from the study. Attempts were made to select a sample with variability in ethnic or racial and socioeconomic backgrounds. Two facilities were used to help recruit subjects: the Good Samaritan Women, Infants and Children (WIC) clinic in Puyallup, Washington, and the Cottesmore Child Development Center in Gig Harbor, Washington. 176

f

Physical Therapy

Volume 89

Table 2 provides details for the sample.

with Spanish, Hispanic, or Latino ethnic backgrounds.25

To compare US and Canadian infants, data for 64 Canadian infants were provided by the authors (SRH and AMM) of the HINT. These data were matched for age, sex, parental ethnicity or race, and parental education with data collected from US infants. (There were no appropriate Canadian matches for 3 of the 67 US infants.)

Two groups were established for SES, primarily on the basis of parental level of education, which has been correlated with income level26 and which is a common proxy for SES. Infants were placed in the highSES group if their parents were educated beyond high school. Infants were placed in the low-SES group if their parents had a level of education of high school or less or if they were recruited from the Good Samaritan WIC clinic or the Cottesmore Child Development Center, because low SES is a prerequisite for participation in these programs.

For further data analysis of the US sample, infants were classified according to ethnicity or race and SES. Two groups were established for ethnicity or race: white and nonwhite. Infants were placed in the nonwhite group if one or both parents were of nonwhite descent. Infants in the nonwhite group were of the following descent: Asian or Pacific Islander (n⫽3), black or African American (n⫽3), Spanish, Hispanic, or Latino (n⫽4), and mixed (n⫽13). Overall, the race breakdown for our sample— 66% white and 34% nonwhite—was similar to that in the US 2000 census (75% white and 25% nonwhite).25 However, our sample overrepresented Asian infants and infants of 2 or more races and underrepresented black or African American infants and infants

Number 2

Procedure Data were collected by 3 of the investigators (AB, JSB, and KS). Interrater reliability was established for all raters during pilot testing by comparison with a rater (SWM) trained and verified as reliable by one of the HINT authors (SRH). The 3 raters achieved an ICC (3,1) of .98 for HINT total scores for 5 infants, with values ranging from .72 to .98 (P⬍.05) on individual test items. Testing took place in the infant’s home; in the University of Puget Sound Physical Therapy clinic in TaFebruary 2009

Validity of HINT as Screening Tool for Motor and Cognitive Delays Table 3. Descriptive Statistics for Harris Infant Neuromotor Test Data X (SD) [n] for the Following Infants: Age of Infant, mo

US White

US High Socioeconomic Status

US Nonwhite

US Low Socioeconomic Status

Canadian

US

2–4

34.78 (6.44) [18]

38.21 (6.84) [18]

39.82 (5.73) [5]

33.70 (8.33) [5]

39.13 (6.27) [7]

36.64 (7.99) [7]

5–6

25.66 (7.09) [17]

22.76 (6.02) [17]

23.23 (3.84) [8]

22.13 (8.44) [8]

20.64 (4.88) [8]

25.69 (6.50) [8]

7–9

10.92 (6.99) [13]

11.62 (8.15) [13]

13.88 (7.75) [5]

8.00 (7.91) [5]

14.13 (8.92) [5]

7.60 (4.72) [5]

10–12

4.72 (2.49) [16]

7.56 (6.61) [16]

5.09 (4.99) [5]

13.00 (6.93) [5]

5.00 (4.29) [6]

9.10 (7.46) [6]

Total

19.71 (13.28) [64]

21.49 (14.10) [64]

19.65 (11.42) [23]

20.12 (13.36) [26]

20.65 (14.24) [26]

coma, Washington; at the Good Samaritan WIC clinic; or at the Cottesmore Child Development Center. Parents signed consent forms approved by the University of Puget Sound Institutional Review Board. Descriptive data, including parental ethnicity or race and parental education, were collected from parental report as part of the background information section of the HINT. Parents then were asked to complete the ASQ while the investigators administered the HINT to the infants. Total time to administer and score the HINT was approximately 30 minutes, including breaks for the infants as needed and discussion of results with the parents. Most parents completed the ASQ in approximately 20 minutes. Total testing time with the families was approximately 45 to 50 minutes because parents tended to participate in HINT testing rather than complete the ASQ during HINT testing. Data Analysis All data were analyzed with the Statistical Package for the Social Sciences, version 14.0 (SPSS 14.0),* and an alpha value set at .05. The HINT total scores were used for all analyses. Data for all comparisons explained below were verified as being normally distributed across groups.

* SPSS Inc, 233 S Wacker Dr, Chicago, IL 60606.

February 2009

19.59 (11.97) [23]

For comparison of Canadian and US infant HINT total scores, infants were matched for age, sex, parental ethnicity or race, and parental education. An independent sample t test was used to compare Canadian and US data, with 64 infants in each group.

points for that domain. Pearson correlation coefficients were used to examine the relationship between HINT and ASQ total scores and between the HINT and each ASQ domain (communication, gross motor, fine motor, problem solving, and personal-social).

To correlate HINT and ASQ scores, we had to exclude some data because of improper completion (multiple answers for questions) of the ASQ by the parent; therefore, the correlation was calculated for 53 infants. The ASQ data were adjusted to reflect changes in scores across development to allow for determination of the level of correlation with the HINT data. The HINT scores become progressively lower as the infant gains skills across the first year of life. The ASQ scores are calculated for a specific age; therefore, no matter how old the infant, these scores range from 0 to 60. To adjust the ASQ scores so that they progressively increased with the age of the children, we gave each child credit for the 60 points in all age levels below the one in which he or she was tested. For instance, a child tested at the 6-month level and receiving a score of 30 points at his or her age level in one domain was awarded an additional 60 points for all items at the 4-month level (the one level below the 6-month level) in that domain and, therefore, achieved an adjusted score of 90

For comparison of white and nonwhite US infant HINT scores, data were age matched. There was no significant difference in sex across ages, as determined with a chi-square test of independence (␹2⫽8.77, df⫽8, P⬎.05). An independent sample t test was used to compare means for 23 infants in each group, because these data represented the infants for whom age matching was possible. For comparison of low-SES and high-SES infant HINT scores, data were age matched. There was no significant difference in sex across ages, as verified with a chi-square test of independence (␹2⫽3.79, df⫽8, P⬎.05). An independent sample t test was used to compare means for 26 infants in each group, because these data represented the infants for whom age matching was possible.

Results Tables 3 and 4 present descriptive data on HINT and ASQ scores across age bands for the US and Canadian infants as well as the ethnicity or

Volume 89

Number 2

Physical Therapy f

177

Validity of HINT as Screening Tool for Motor and Cognitive Delays Table 4. Descriptive Statistics for Ages and Stages Questionnaire Data X (SD) Score Age of Infant, mo (No. of Infants)

Total

Communication Domain

Gross Motor Domain

Fine Motor Domain

Problem-Solving Domain

Personal-Social Domain

2–4 (10)

239.00 (43.70)

48.64 (11.64)

54.09 (7.01)

42.00 (20.17)

47.50 (10.34)

45.91 (15.62)

5–6 (19)

235.79 (37.91)

52.89 (10.18)

41.32 (12.89)

45.79 (12.50)

50.79 (8.04)

45.00 (12.13)

7–9 (12)

265.83 (37.47)

54.17 (5.97)

50.00 (11.28)

51.67 (13.54)

55.00 (9.77)

55.00 (7.07)

10–12 (12)

221.25 (49.92)

42.31 (16.66)

38.08 (20.47)

48.21 (9.73)

45.00 (10.74)

41.92 (12.67)

race and SES subgroups. The data were divided into these age bands to demonstrate the decrease in HINT scores across ages and to provide the reader with a more detailed description of the data analyzed. There were no significant between-group differences in HINT total scores for US and Canadian infants (N⫽128, t⫽0.74, P⫽.46). Table 5 presents correlations for HINT and ASQ scores. The HINT showed good to excellent correlations with the ASQ total score and with each ASQ domain (n⫽52; correlations ranged from ⫺.82 to ⫺.84; P⬍.05). No differences were seen in the correlations for individual domains. Correlations were negative because superior performance is indicated by lower scores on the HINT versus higher scores on the ASQ. The subgroup analyses for the US infants who were classified as white and nonwhite were not significantly different (n⫽46, t⫽0.02, P⫽.99),

nor were there differences between low-SES infants and high-SES infants (n⫽52, t⫽⫺0.14, P⫽.89). Sample sizes were small, but total score means and variances were very similar across groups (Tab. 3).

Discussion Our results reflect no significant difference in HINT total scores for infants who were healthy in the United States and in Canada and, therefore, support our first hypothesis. From a preliminary perspective, these results suggest that norms developed from data previously collected from Canadian infants can be applied to US infants. Further normative testing of US infants with larger numbers of infants and with infants from different geographical regions would strengthen this conclusion. Additional geographical diversity in Canadian data would also strengthen this conclusion because the majority of normative testing took place in British Columbia and Ontario.

Table 5. Correlations for Harris Infant Neuromotor Test (HINT) and Ages and Stages Questionnaire (ASQ) Scores Pearson Correlation Coefficienta for HINT (nⴝ52)

ASQ

a

Total score

⫺.83

Communication subscale

⫺.83

Gross motor subscale

⫺.82

Fine motor subscale

⫺.82

Problem-solving subscale

⫺.82

Personal-social subscale

⫺.84

All values were significant at P⬍.05.

178

f

Physical Therapy

Volume 89

Number 2

Strong correlations between the HINT and the ASQ across all individual ASQ domains were found. Because the ASQ has performed well in tests against gold standards for infant developmental screening, these data provide support for the validity of the HINT. On the basis of our second hypothesis, we expected to find stronger correlations between the motor and problem-solving domains of the ASQ and the HINT, but all domains of the ASQ were equally correlated. This result may have occurred because of the screening nature of both tools. On the basis of a comparison of developmental scores on the ASQ and the HINT, we suggest that there may be differences in the abilities of these tools to screen for mild to moderate developmental delays. Discrete items are provided at various ages in the ASQ; therefore, the assumption is that an infant in a particular age group will have mastered all items in the younger age groups. This may not always be the case; therefore, scores on the ASQ may underrepresent children at risk for delays. Screenings carried out with the HINT require that movement be scored for all test items across all ages to avoid this potential problem. Evaluation of the relationships between these tools would be strengthened by further testing with larger samples that include children at risk for developmental delays. We do not yet know whether the correlations found between the HINT and the February 2009

Validity of HINT as Screening Tool for Motor and Cognitive Delays ASQ for infants who are developing typically are generalizable to infants at risk for delays or infants with developmental disabilities. No significant differences were found in HINT total scores for infants who were healthy in the white versus nonwhite groups and in the lowSES versus high-SES groups, thus negating our third hypothesis. However, this finding provides some support for the notion that the HINT can be appropriately applied to infants from various backgrounds. We expected to find differences and perhaps did not find them because our study was limited by a small sample size. Consequently, we were not able to examine the data to determine whether the scores were similar across all age levels of the HINT. The mean scores for 7- to 9-monthold and 10- to 12-month-old infants appeared to be different for both white versus nonwhite groups and low-SES versus high-SES groups (Tab. 3). However, the differences were opposite and were canceled out in the total group comparison analyses. A larger sample would allow the examination of differences at different ages. Ideally, infants also should be divided into several different groups of ethnicity or race and multiple levels of SES to fully understand the effects of ethnicity or race and SES on the rates of typical development. Further testing is needed with larger numbers of subjects and the ability to stratify samples to support or refute the preliminary results found in the present study. Follow-up studies should aim to collect a sample of sufficient size and diversity to represent US demographic variability. Also, although level of education has been noted as an acceptable proxy measure for SES,26 it is not the most comprehensive SES measure. For instance, we encountered several families whose level of education would February 2009

place them in a high SES class but who qualified for low-income assistance programs. Unlike in the United States, which has Part C (for early intervention) of the Individuals With Disabilities Education Act (IDEA, Public Law 108446),27 Canada has no federal law mandating early intervention services for infants at risk for developmental delays or those with known delays or disabilities. Health care is provided on a provincial basis. In British Columbia, for example, there is a province-wide Infant Development Program that serves children from birth to 3 years of age and their families.28 According to the program’s Web site,28 the rationale for this early intervention program is “in common with similar services in the USA and elsewhere in Canada.” That is, interventions may be most effective if begun early in life; infancy is an important period of life, and delays during that period may be long lasting; and the family is the most crucial source of learning, emotional support, and developmental encouragement for a child. Major referral sources in both countries include public health nurses, physicians, and parents. The HINT was developed to be used by a variety of health care providers; therefore, it should serve in a similar manner in each country as a possible tool for early screening of infant development.

Conclusion The present study demonstrated support for the premise that HINT norms developed from data previously collected from Canadian infants may be appropriate for application to US infants. Additionally, the HINT was shown to be correlated with the ASQ, an established developmental screening tool with acceptable validity and sensitivity. The latter finding supports the notion that the HINT may be an appropriate tool for pediatric physical therapists

to use in screening for cognitive and motor delays in various health care settings, such as offices of general medical practitioners, pediatric nurse practitioners, or within the home. The results also provide preliminary support for the notion that HINT scores for white versus nonwhite infants and for high-SES versus low-SES infants in the United States are similar; however, future research with larger samples is necessary to verify these preliminary results. Dr Westcott McCoy, Dr Bowman, Dr SmithBlockley, and Dr Sanders provided concept/ idea/research design, writing, and fund procurement. Dr Bowman, Dr Smith-Blockley, and Dr Sanders provided data collection. Dr Westcott McCoy and Dr Smith-Blockley provided data analysis. Dr Westcott McCoy and Dr Bowman provided project management. Dr Bowman provided subjects. Ms Megens and Dr Harris provided consultation (including review of manuscript before submission). The authors acknowledge the families and children who participated in this study. This research was completed as part of Dr Bowman’s, Dr Smith-Blockley’s, and Dr Sanders’s DPT degree requirements at the University of Puget Sound, Tacoma, Washington. Dr Harris is the primary author of the Harris Infant Neuromotor Test (HINT), to be published by PRO-ED Inc, Austin, Texas. Ms Megens is the second author of the HINT. The study was approved by the University of Puget Sound Institutional Review Board. Grant support for this study was provided by the University of Puget Sound Enrichment Committee (grant SR0543). Role of the Funding Source: Funding for the purchase of the equipment for the test kits and forms, and a small present for the families for their participation in the research was provided by the grant. This research was presented as a poster at the Combined Sections Meeting of the American Physical Therapy Association; February 14 –18, 2007; Boston, Massachusetts; and at the 15th International Congress of the World Confederation for Physical Therapy; June 2– 6, 2007; Vancouver, British Columbia, Canada.

Volume 89

Number 2

Physical Therapy f

179

Validity of HINT as Screening Tool for Motor and Cognitive Delays This article was received June 18, 2008, and was accepted November 7, 2008. DOI: 10.2522/ptj.20080189

References 1 Harris SR, Daniels LE. Content validity of the Harris Infant Neuromotor Test. Phys Ther. 1996;76:727–737. 2 Platt MJ, Cans C, Johnson A, et al. Trends in cerebral palsy among infants of very low birthweight (⬍1,500 g) or born prematurely (⬍32 weeks) in 16 European centres: a database study. Lancet. 2007;369:43–50. 3 Hediger ML, Overpeck MD, Ruan WJ, Troendie JF. Birthweight and gestational age effects on motor and social development. Paediatr Perinat Epidemiol. 2002;16; 33– 46. 4 Harris SR. Harris Infant Neuromotor Test (Development Edition 3). Austin, TX: Pro-Ed Inc; 2003. 5 Harris SR, Megens AM, Backman CL, Hayes V. Development and standardization of the Harris Infant Neuromotor Test. Infant Young Child. 2003;16:143–151. 6 Lee LLS, Harris SR. Psychometric properties and standardization of four screening tests for infants and young children: a review. Pediatr Phys Ther. 2005;17: 140 –147. 7 Harris SR, Daniels LE. Reliability and validity of the Harris Infant Neuromotor Test. J Pediatr. 2001;139:249 –253. 8 Megens AM, Harris SR, Backman CL, Hayes VE. Known-groups analysis of the Harris Infant Neuromotor Test. Phys Ther. 2007; 87:164 –169. 9 Majnemer A, Barr RG. Association between sleep position and early motor development. J Pediatr. 2006;149:623– 629.

180

f

Physical Therapy

Volume 89

10 Bartlett DJ, Kneale Fanning JE. Relationships of equipment use and play positions to motor development at eight months corrected age of infants born preterm. Pediatr Phys Ther. 2003;15:8 –15. 11 Harris SR. Parents’ and caregivers’ perceptions of their children’s development. Dev Med Child Neurol. 1994;36:918 –923. 12 Squires J, Potter L, Bricker D. The ASQ User’s Guide for the Ages & Stages Questionnaires®: A Parent-Completed, Child Monitoring System. 2nd ed. Baltimore, MD: Paul H Brookes Publishing Co; 1999. 13 Skellern CY, Rogers Y, O’Callaghan MJ. A parent-completed developmental questionnaire: follow up of ex-premature infants. J Paediatr Child Health. 2001;37: 125–129. 14 Status and trends in the education of racial and ethnic minorities. Available at: http:// nces.ed.gov/pubs2007/minoritytrends/ind_ 1_1.asp. Accessed August 22, 2008. 15 US Census Bureau. USA statistics in brief— race and hispanic origin. Available at: http://www.census.gov/compendia/statab/ files/racehisp.html. Accessed August 22, 2008. 16 Kolobe TH. Childrearing practices and developmental expectations for MexicanAmerican mothers and the developmental status of their infants. Phys Ther. 2004;84: 439 – 453. 17 Toy CC, Deitz J, Engel JM, Wendel S. Performance of 6-month-old Asian American infants on the Movement Assessment of Infants: a descriptive study. Phys Occup Ther Pediatr. 2000;19:5–23. 18 Zahr LK. Predictors of development in premature infants from low-income families: African Americans and Hispanics. J Perinatol. 1999;19:284 –289. 19 Zahr LK. Quantitative and qualitative predictors of development for low-birthweight infants of Latino background. Appl Nurs Res. 2001;14:125–135.

Number 2

20 Kelly Y, Sacker A, Schoon I, Nazroo J. Ethnic differences in achievement of developmental milestones by 9 months of age: the Millennium Cohort Study. Dev Med Child Neurol. 2006;48:825– 830. 21 Nelson EA, Yu LM, Wong D, et al. Rolling over in infants: age, ethnicity, and cultural differences. Dev Med Child Neurol. 2004; 46:706 –709. 22 Santos DC, Gabbard C, Goncalves VM. Motor development during the first year: a comparative study. J Genet Psychol. 2001; 162:143–153. 23 Ravenscroft EF, Harris SR. Is maternal education related to infant motor development? Pediatr Phys Ther. 2007;19:56 – 61. 24 Portney L, Watkins M. Foundations of Clinical Research: Applications to Practice. 2nd ed. Upper Saddle River, NJ: Prentice-Hall Inc; 2000. 25 Grieco EM, Cassidy RC. Overview of race and Hispanic origin: Census 2000 brief. US Department of Commerce; 2001. Available at: http://www.census.gov/ prod/2001pubs/cenbr01–1.pdf. Accessed March 19, 2007. 26 Parents’ low education leads to low income despite full time employment. 2007. Available at: http://nccp.org/publications/ pub_786.html. Accessed December 13, 2008. 27 Public Law 108-446. December 3, 2004. Available at: http://www.copyright.gov/ legislation/pl108-446.pdf. Accessed December 13, 2008. 28 Infant Development Program of BC. Available at: http://www.idpofbc.ca/parents. html. Accessed November 17, 2008.

February 2009

Perspective

Electrical Stimulation Using KilohertzFrequency Alternating Current Alex R Ward Transcutaneous electrical stimulation using kilohertz-frequency alternating current (AC) became popular in the 1950s with the introduction of “interferential currents,” promoted as a means of producing depth-efficient stimulation of nerve and muscle. Later, “Russian current” was adopted as a means of muscle strengthening. This article reviews some clinically relevant, laboratory-based studies that offer an insight into the mechanism of action of kilohertz-frequency AC. It provides some answers to the question: “What are the optimal stimulus parameters for eliciting forceful, yet comfortable, electrically induced muscle contractions?” It is concluded that the stimulation parameters commonly used clinically (Russian and interferential currents) are suboptimal for achieving their stated goals and that greater benefit would be obtained using short-duration (2– 4 millisecond), rectangular bursts of kilohertz-frequency AC with a frequency chosen to maximize the desired outcome.

AR Ward, PhD, is Associate Professor, Musculoskeletal Research Centre, Faculty of Health Sciences, La Trobe University, Victoria 3086, Australia. Address for correspondence: School of Human Biosciences, Faculty of Health Sciences, La Trobe University, Victoria 3086, Australia. Address all correspondence to Dr Ward at: a.ward@ latrobe.edu.au. [Ward AR. Electrical stimulation using kilohertz-frequency alternating current. Phys Ther. 2009; 89:181–190.] © 2009 American Physical Therapy Association

Post a Rapid Response or find The Bottom Line: www.ptjournal.org February 2009

Volume 89

Number 2

Physical Therapy f

181

Electrical Stimulation Using Kilohertz-Frequency Alternating Current

T

wo forms of electrical stimulation are commonly used clinically: pulsed current (PC) and burst-modulated alternating current (BMAC). Examples of BMAC are “Russian current” and “interferential current.” Burst-modulated alternating current stimulation is claimed to be more comfortable than PC and capable of eliciting greater muscle torque.1–3

The response of nerve and muscle to PC electrical stimulation has been studied by physiologists since the late 19th century.4,5 Consequently, our present understanding of the effects of PC is relatively good. The physiological response to BMAC stimulation is less-well understood. This article reviews the known physiology and clinically relevant, laboratory-based studies of electrical stimulation, which offer some insight into the mechanism of action of BMAC and provide some answers to the questions “Does BMAC stimulation have an advantage over PC?” and “What are the optimum treatment parameters for BMAC stimulation?”

BMAC Stimulus Parameters Alternating current (AC) used clinically is normally kilohertz-frequency AC, delivered in bursts, with the burst frequency in the “physiological” range (up to 100 Hz or so). It, therefore, is called “burst-modulated alternating current.” Figure 1 illustrates, for comparison, unmodulated AC, monophasic PC, and 2 examples of BMAC. The currents illustrated in Figures 1A, 1C, and 1D are defined as AC because the waveforms have alternating positive and negative phases with no gap between them. The current shown in Figure 1B is defined as PC because successive phases (the pulses) are separated by an appreciable gap.6 182

f

Physical Therapy

Volume 89

Pulsed current is easily described by specifying 3 things: (1) the waveform (eg, rectangular and monophasic, as in Fig. 1B), (2) the pulse duration (normally in the range of 50 microseconds to 1 millisecond), and (3) the pulse frequency (normally in the range of 1 Hz to about 100 Hz). The description of AC is more complex. Alternating current, by definition, is biphasic, and the biphasic waveform can be sinusoidal or rectangular. The current also can be delivered continuously (Fig. 1A), in rectangular bursts (Fig. 1C), or in sinusoidally modulated bursts (Fig. 1D). Thus, when describing the stimulus, there is the potential for confusion because several parameters must be specified to completely describe the waveform. Figure 2 shows an example of BMAC, with particular parameters identified. In Figure 2, the burst duration is 4 milliseconds, and because the interval between bursts is 16 milliseconds, the period (the burst repetition time) is 20 milliseconds, or 1/50th of a second. Therefore, the burst repetition frequency is 50 times per second in this example (ie, the burst frequency is 50 Hz). Each burst consists of a number of AC cycles. In this example, each 4-millisecond burst consists of 4 AC sine waves. Each sine wave has a duration of 1 millisecond, or 1/1,000th of a second, so the sinewave frequency is 1,000 times per second (ie, 1 kHz). The sine-wave frequency is sometimes referred to as the “carrier frequency.”1,2,7 Each 1-millisecond sine wave comprises 2 phases: one positive phase followed by one negative phase, so each phase has a duration of 0.5 milliseconds, or 500 microseconds. The greater the number of parameters, the greater the number of possible permutations and combinations. This raises the question of

Number 2

whether AC stimulators used clinically have the best combination of parameters for achieving the desired clinical outcome.

BMAC Stimulation Types Used Clinically Russian Current Russian current is 2.5-kHz AC, applied in 50-Hz rectangular bursts with a burst duty cycle of 50%. The stimulus waveform is shown in Figure 1C. The burst duration is 10 milliseconds at 50 Hz. Russian current is claimed to be beneficial for muscle strengthening (increasing force-generating capacity). The choice of a 2.5-kHz frequency for Russian current appears to be based on measurements of maximum electrically induced torque (MEIT) by Kots and co-workers8 using not bursts but a continuous AC stimulus (Fig. 1A) in the frequency range of 100 Hz to 5 kHz.8,9 The choice of a burst-modulated, 50% duty cycle (Fig. 1C) is based on the observation that there was little difference in MEIT between continuous AC and rectangular bursts with a 50% duty cycle but that with a 50% duty cycle, half as much electrical energy is delivered, so there is less risk for tissue damage.8,9 Russian currents became popular despite an equivocal evidence base due to the limited number of studies and their different findings.3,9 The balance of evidence supports the notion that strengthening can be produced, but at one extreme there is the single-case study reported by Delitto et al,10 which demonstrated a substantial strength gain, whereas at the other extreme there is the study by St Pierre et al,11 which demonstrated no strength gain. Other than the original Russian study,8,9 only 2 subsequent studies have addressed whether 2.5 kHz is the best AC frequency for muscle torque production.12,13 These 2 studies used 50-Hz bursts of kilohertz-frequency AC, February 2009

Electrical Stimulation Using Kilohertz-Frequency Alternating Current

Figure 1. (A) Steady, unmodulated alternating current; (B) monophasic pulsed current; (C) burst-modulated alternating current with rectangular burst modulation; and (D) burst-modulated alternating current with sinusoidal modulation.

and both studies showed that maximum torque was elicited at a frequency of 1 kHz. It is noteworthy that Andrianova et al8 reported that 2.5 kHz is optimum if stimulation is applied directly (over the muscle) but that if stimulation is applied indirectly (over the nerve trunk), the optimum frequency is 1 kHz. Thus, it might be concluded that “optimal stimulus parameters” may well depend on electrode positioning and that the popular frequency (2.5 kHz) could be suboptimal for commonly used electrode placements. Interferential Currents Interferential currents are reported to be the most popular form of electrical stimulation used in clinical practice in the United Kingdom and other European countries and in Australia.1 Interferential stimulators produce 2 independent kilohertz-frequency AC currents of constant intensity (Fig. 1A) applied by 2 separate pairs of electrodes, which are positioned diagonally opposed to produce an “interference” effect (Fig. 1D) in the central region of intersection of the currents (Fig. 3).1,2,7

February 2009

The currents are applied continuously at constant intensity (Fig. 1A), but they have different frequencies (eg, 4,000 and 4,050 Hz), and in the tissue between the electrodes, the 2 currents interfere. It is stated1,2,7 that the currents reinforce in the central region of intersection (Fig. 3A) to produce a stimulus waveform that is

sinusoidally modulated at a frequency equal to the difference between the 2 AC frequencies (Fig. 3B, top). The stimulation waveform, therefore, resembles that illustrated in Figure 1D and would have a modulation frequency of 50 Hz in this example. This argument is misleading because it ignores the effect of

Figure 2. An example of burst-modulated alternating current. A minimum of 5 parameters must be specified in order to describe the waveform. In this example, the waves are sinusoidal, the alternating current (AC) frequency is 1 kHz, the bursts are rectangular, the burst frequency is 50 Hz, and the burst duration is 4 milliseconds.

Volume 89

Number 2

Physical Therapy f

183

Electrical Stimulation Using Kilohertz-Frequency Alternating Current

Figure 3. (A) Interferential currents are claimed to produce maximum stimulation in the region of intersection of the 2 diagonally opposed currents, as shown. (B) The actual stimulation intensity experienced by nerve fibers has maximum modulation if the fibers are oriented optimally and zero modulation when fibers are oriented along one of the current pathways.

tissue inhomogeneity and, perhaps more importantly, nerve fiber orientation.14,15 Nerve fibers oriented along an axis directly between one pair of electrodes will experience continuous, unmodulated AC, while only those angled optimally between the 2 axes will experience fully modulated AC (Fig. 3). The optimum angle depends on the relative intensities of the current. If the current intensities are equal, the optimum angle is 45 degrees to the current paths (ie, horizontally or vertically in Fig. 3A), but in practice the currents will not be equal due to the variation with position (relative to the electrodes) and the variation in electrical impedance of different tissues (fat, muscle, connective tissue, and bone) in the current pathway.15 Unless the orientation of the nerve fibers is optimal, the stimulus modulation will be partial. Thus, with interferential currents, the actual stimulus waveform applied to the nerve fibers is not known and can vary between unmodulated and fully modulated AC (Fig. 3B), depending on the nerve fiber orientation and location relative to the electrode placement.

184

f

Physical Therapy

Volume 89

Premodulated Interferential Current Most interferential stimulators also offer premodulated interferential current. The term “premodulated interferential” is something of a misnomer because it refers to current that is fully modulated (as in Fig. 1D) and applied between one pair of electrodes. Thus, by definition, this current is no longer interferential (ie, no longer produced by the interference of 2 currents). The current is simply kilohertz-frequency AC, modulated at a low frequency, typically in the range of 1 to 120 Hz.1,2,7 Unlike “true” interferential current, the amount of modulation of the stimulation waveform does not depend on the nerve fiber orientation relative to the electrodes. The stimulus waveform is simply that provided by the stimulator and, therefore, is predictable. If the “premodulated” current is sinusoidally modulated (as produced by traditional interferential stimulators and shown in Fig. 1D), some parts of the burst will be below threshold while other parts of the burst will be above threshold. Thus, the effective burst duration for any

Number 2

given nerve fiber is uncertain and will vary with stimulation intensity, which varies with proximity to the electrodes. Nerve fibers close to the electrodes will be stimulated suprathreshold for a larger part of each burst than those further away; thus, the effective burst duration will vary. Some modern interferential stimulators use rectangular burst modulation (Fig. 1C), so there is no uncertainty as to the effective duration: the burst is either fully “on” or “off.”

Importance of Modulation Effect of Burst Duration on Thresholds As noted earlier, Russian current is burst modulated with a rectangular envelope (Fig. 1C). Premodulated interferential current may be either rectangular burst modulated (Fig. 1C) or, more commonly, sinusoidally modulated (Fig. 1D), whereas with “true” interferential currents, the stimulus experienced by a nerve fiber may be continuous (unmodulated), fully modulated, or partially modulated, depending on the fiber location and orientation relative to the electrodes.

February 2009

Electrical Stimulation Using Kilohertz-Frequency Alternating Current The first published report of the effect of modulation appears to be the report by Soloviev published in 1963.16 Soloviev used AC stimulation over the frequency range of 2 to 8 kHz and found that there was little difference in motor threshold regardless of whether the current applied was continuous or burst modulated at 50 Hz with a 50% duty cycle. A 2001 study of motor thresholds by Ward and Robertson17 again showed little difference, this time over the frequency range of 1 to 25 kHz. It should be noted, however, that only continuous AC and 50% duty cycle, 50-Hz bursts were compared, so the comparison was between 10-millisecond bursts and continuous AC. In 2007, Ward and LucasToumbourou18 reported a study of sensory, motor, and pain thresholds using AC frequencies of 1 kHz and 4 kHz applied as 50-Hz bursts. They used burst durations in the range of 0.25 to 20 milliseconds and found that thresholds decreased to a plateau with increasing burst duration. An interesting finding of this study was that the plateau in threshold with burst duration depended on the response evoked (ie, sensory, motor, or pain threshold, with values of 5 to 7, 10, and ⬎20 milliseconds, respectively). Thus, motor thresholds decrease with increasing burst duration, but at burst durations above about 10 milliseconds, there is no further decrease. These findings explain the lack of differences found in the earlier studies, when only 10millisecond bursts and continuous AC were compared. The most intriguing finding of the study, however, was that the burst duration plateaus were different for sensory, motor, and pain thresholds. This means that there will be optimal burst durations where the pain/ sensory threshold and pain/motor threshold ratios are maximum. Ward and Lucas-Toumbourou estimated an optimal burst duration for both senFebruary 2009

sory and motor stimulation as 2 to 3 milliseconds. This is appreciably shorter than the burst durations commonly used clinically (typically 10 milliseconds for Russian current and greater or similar for interferential currents). Effect of Burst Duration on MEIT and Discomfort Andrianova et al8 used different AC frequencies in the range of 100 Hz to 5 kHz and compared not thresholds but maximum torque production using continuous (unmodulated) AC and AC bursts (modulated at 50 Hz with a 50% duty cycle [ie, 10millisecond burst duration]). They concluded that there was little difference in MEIT with burst-mode or continuous stimulation, but they did not make any statistical comparisons. Their published data (reproduced in Tab. 3 of the perspective article by Ward and Shkuratova9) however, show that across the frequency range, the torques produced by burst-modulated currents were, on average, 14% higher (SD⫽12%). Ward and Shkuratova9 conducted a paired t-test comparison across frequencies, using Andrianova and colleagues’ published data,8 and found that this difference is significant (P⫽.03) (ie, torques are significantly higher when a rectangular burstmodulated stimulus of 10 milliseconds’ duration is used rather than a continuous AC stimulus). Bankov,19 in 1980, compared 5-kHz AC, modulated at 60 Hz, using stimulation intensities that produced just enough contraction of the biceps brachii muscle to maintain the elbow at 90 degrees of flexion with the upper arm vertical (an antigravity flexion level of muscle activity). He compared rectangular bursts of 1, 2, and 5 milliseconds’ duration and reported that the 1-millisecond burst was the most comfortable. Another study reported by Bankov in the same year20 compared 60-Hz sinusoi-

dally modulated bursts of AC, which varied in their modulation depth from 0% (steady, continuous AC; Fig. 1A) to 100% (fully modulated; Fig. 1D), and hypermodulated bursts of AC (gaps between bursts). He reported that force increased with the degree of modulation but that the associated discomfort showed little variation. A conclusion is that shorter burst durations produce more force at the same level of discomfort. In 1981, Bankov and Daskalov21 compared 5-kHz AC applied in 2-millisecond bursts with PC of varying pulse widths. Each was applied 3 seconds on and 3 seconds off at an intensity that produced antigravity flexion of the biceps muscle. The 5-kHz stimulus was found to be more comfortable. These early studies, thus, had 2 major findings: (1) that for a given level of force production, burst-modulated AC is preferable to continuous AC or PC, and (2) a short AC burst duration (1 or 2 milliseconds) is optimal for least discomfort. A recent study13 measured MEIT and relative discomfort using 50-Hz bursts of AC in the frequency range of 0.5 to 20 kHz. Burst durations ranging from the shortest possible (1 cycle) to the longest (continuous AC) were used. Maximum torque was produced at a frequency of 1 kHz and a burst duration of 2 milliseconds (10% duty cycle). Minimum discomfort occurred at a frequency of 4 kHz and a burst duration of 4 milliseconds (20% duty cycle). Continuous AC produced the least torque and the greatest discomfort at all frequencies. Single cycles (biphasic PC) produced significantly less torque than 2-millisecond bursts and were more uncomfortable. A later study22 compared Russian current (2.5-kHz AC applied in 10millisecond bursts) and “Aussie current” (1-kHz AC applied in 4-millisecond bursts) with PC of the same phase duration (200 and 500

Volume 89

Number 2

Physical Therapy f

185

Electrical Stimulation Using Kilohertz-Frequency Alternating Current microseconds, respectively) in terms of discomfort and torque production. The AC bursts (Fig. 1C) were more comfortable than their PC counterparts. Both Aussie current and the 2 forms of PC produced similarly high torques, but, perhaps surprisingly, Russian current evoked less. Thus, it seems reasonable to conclude that a stimulus waveform that consists of kilohertz-frequency AC in short-duration bursts (2– 4 milliseconds) is more comfortable and elicits greater MEIT than PC, continuous AC, Russian current, or interferential current stimulation.

The “Conventional Wisdom” Historical Claims Concerning Interferential Currents Nemec23–26 promoted the therapeutic use of interferential currents and advocated the use of sinusoidal AC at frequencies around 5 kHz. He argued that the 2 currents of slightly different frequency “interfere” in tissue, producing maximum stimulation in the region of intersection of the 2 current paths, and that the resulting (endogenous) current at depth would be modulated at the “beat” frequency, which is the difference in frequency of the 2 currents (Fig. 3).1,7 Nemec23–26 gave 3 arguments for the use of interferential current rather than PC: 1. Skin impedance is lower at high AC frequencies; therefore, less electrical energy is dissipated in the skin and, consequently, there is less sensory stimulation and discomfort than with low-frequency PC. 2. When the constant-intensity currents intersect and interfere, the resulting current will be modulated in intensity at the beat 186

f

Physical Therapy

Volume 89

frequency (the difference between the 2 AC frequencies) and will produce endogenous lowfrequency stimulation (ie, at depth, rather than superficially). 3. Currents interfere in tissue, producing maximum stimulation at the region of intersection of the 2 current paths, where a “cloverlike” pattern of stimulation is produced.1,7 The first point is incorrect for 2 reasons. First, the skin impedance to PC depends on the phase duration, not the pulse frequency.1,2,5,27–29 The skin impedance to low-frequency AC is much higher than to kilohertzfrequency AC because the phase duration is much longer. If the PC has the same phase duration as the kilohertz-frequency AC, the skin impedance is the same even if the pulse frequency is low.1,2,5,27–29 Conventional PC typically has a phase duration similar to that of interferential current. Thus, the argument that interferential current would meet with a lower impedance is without any basis. Second, a lower skin impedance does not mean less stimulation of sensory and pain fibers in the skin and, therefore, less discomfort. The high skin impedance with long phase durations (eg, with lowfrequency AC) is due to the skin capacitance, which is due almost entirely to the stratum corneum: the dead, scaly, relatively dry, outermost layer.1,2,5,27,28 The stratum corneum has no sensory, pain, or other kind of nerve fibers.30,31 These fibers are located beneath, in the dermis, which is well hydrated and of similar conductivity to the deeper tissues.5,30,31 The second and third points are oversimplifications. There are 3 important things to consider with interferential stimulation: 1. An interference pattern of stimulation is produced everywhere,

Number 2

not just at the predicted region of intersection of the currents, and the extent of modulation of the resulting current will depend on the location and orientation of the nerve fibers relative to the electrodes.1,2,14,15 This means that throughout the tissue volume, fibers orientated at an optimum angle will experience a fully modulated current, whereas those at other angles (the majority) will be subject to a partially modulated or unmodulated stimulus.1,2,14,15 2. Current spreading means that there will not be a region at the center of intersection of the currents where maximum stimulation occurs. Although the stimulation at depth might be expected to be greater, current spreading would be expected to significantly reduce the value of any reinforcement effect.1,14,15 3. It might be expected that the current intensity at depth would be greater with quadripolar stimulation than with bipolar stimulation because of interference and reinforcement. Lambert et al15 demonstrated that this is not true. When currents are applied using conventional interferential stimulation, the pattern of stimulation is not focused centrally. It is more diffuse due to current flow between adjacent electrodes because of the shorter-distance, lower-resistance pathways. Thus, the depth efficiency claims for interferential current are not substantiated. This, together with the uncertain degree of modulation of the stimulus, calls into question whether the “interference” effect of interferential current is of any value. Ozcan et al32 addressed this question when they assessed the relative discomfort of true and premodulated interferential currents (delivered in 50-Hz bursts, 10 milliseconds on and February 2009

Electrical Stimulation Using Kilohertz-Frequency Alternating Current 10 milliseconds off). Premodulated interferential current was found to be significantly more comfortable than true interferential current and more effective for muscle contraction. Historical Claims Concerning Russian Current A talk given by Kots,33 of the Central Institute of Physical Culture, Moscow, at a conference hosted by Concordia University, Montreal, in 1977 laid the foundation for what became known in the Western world as Russian current electrical stimulation.9 Kots reported strength gains of up to 40% in elite Russian athletes stimulated with 2.5-kHz AC applied in 10millisecond rectangular bursts at a frequency of 50 Hz. His protocol used currents with a 10-second on period followed by a 50-second rest period, applied 10 times in each stimulation session (ie, 10-minute treatment sessions). Treatment was applied daily over a period of weeks. As noted previously, Russian currents became popular despite an equivocal evidence base due to the limited number of studies and their different findings.3,9 The choice, by Kots’ group, of 10-millisecond bursts (50% duty cycle) was because of their observation that it evoked just as much muscle torque as continuous AC but, because of the burst modulation, the average current applied to tissue was halved. The effect of different burst durations was not explored. Bankov19,20 and Bankov and Daskalov,21 in the 1980s, examined the effect of burst duration and found that, for the same level of force production, short-duration bursts are more comfortable. An inference is that greater levels of force would be produced at the same level of discomfort if shortduration bursts were used. This is supported by the recent work of Ward et al,13 who measured torque at the pain tolerance limit and found that the greatest MEIT is produced February 2009

using 2-millisecond bursts of AC with a frequency of 1 kHz. Thus, the rationale for the clinical use of Russian current is called into question. The evidence is that stimulation with short-duration bursts of kilohertz-frequency AC would be preferable and that a burst duration of 2 milliseconds appears to be optimal for torque production.

Discussion—The Known Electrophysiology The available laboratory-based evidence indicates that short-duration bursts of kilohertz-frequency AC have advantages over Russian current, interferential current, and PC and that there are optimal frequencies and burst durations for achieving the desired outcome. There are 4 interrelated electrophysiological factors that could help explain the empirical findings: summation, multiple firing, high-frequency fatigue, and neural block. Summation With kilohertz-frequency AC stimulation, there is the possibility of summation, a phenomenon first described by Gildemeister.34,35 Gildemeister reported that when bursts of kilohertz-frequency AC are applied transcutaneously, the threshold voltage for sensory nerve excitation decreases as the burst duration is increased. This phenomenon, later called the “Gildemeister Effect,” occurs because, with each successive pulse in the AC wave-train, the nerve fiber membrane is pushed closer to threshold. Membrane threshold is reached when successive pulses result in sufficient depolarization to produce an action potential. Gildemeister observed a limit to the summation effect. As the number of cycles per burst was increased, the threshold decreased, but only up to a point. Beyond a certain burst duration, no further decrease in threshold was observed. He called this

maximum burst duration (ie, time over which pulses could summate) the “Nutzzeit” or “utilization time.” As noted previously, a recent study by Ward and Lucas-Toumbourou18 showed that the apparent utilization time was different for sensory, motor, and pain thresholds and, consequently, that relative thresholds (pain/motor and pain/sensory) vary with burst duration. These authors found that optimum discrimination (biggest separation between thresholds [ie, maximum relative thresholds]) occurred at burst durations of 2 to 4 milliseconds. High-Frequency Fatigue When electrical stimulation is applied to elicit a motor response using PC frequencies higher than physiological or at the high end of the physiological range (ie, greater than about 50 Hz), it is possible to produce a blockage of muscle activity due to propagation failure or neurotransmitter depletion.36 –38 This is responsible for the phenomenon of “high-frequency fatigue,”38,39 which is characterized by its associated rapid recovery. If a stimulus frequency of 80 Hz, for instance, is used to elicit muscle contraction, the resulting muscle force declines rapidly, but if a brief rest period (a few seconds) is allowed, marked recovery occurs.38,39 This is quite different from “low-frequency fatigue,” which is much more akin to normal physiological fatigue, where the force decline is much slower and the recovery time is much longer. One form of high-frequency fatigue, propagation failure, can occur when action potentials are induced in motoneurons at sufficiently high frequency. This can result in action potential failure at branch points where a motor nerve divides to innervate individual muscle fibers. Failure also can occur at the neuromuscular junction because neurotransmitter

Volume 89

Number 2

Physical Therapy f

187

Electrical Stimulation Using Kilohertz-Frequency Alternating Current depletion is possible at relatively high stimulation frequencies.40 Beyond the neuromuscular junction, transmission failure can occur at the level of the t-tubule system. Normally, the wave of depolarization of a muscle fiber action potential is transmitted over the muscle fiber membrane and throughout the t-tubule system, activating the contractile elements. When sufficiently high frequency action potentials are induced, the t-tubule membranes do not have time to recover between action potentials and muscle fiber contraction ceases.39,40 Whichever the mechanism, whether propagation failure or neurotransmitter depletion, a blockage of muscle contraction at stimulation frequencies around and above about 50 Hz is the result, and the effect is described as high-frequency fatigue. Summation and Multiple Firing When the stimulus is PC applied at low frequency (less than 100 Hz), it can be confidently concluded that, provided that the pulse intensity is sufficiently above threshold, the nerve fiber firing frequency will equal the pulse frequency. The firing frequency could be less if successive pulses occur within the relative refractory period and the stimulus intensity is not sufficiently high, but the firing frequency could never be higher than the PC frequency. With bursts of AC, however, there is the possibility that a single burst will result in multiple action potentials as a result of summation41– 44; therefore, the firing frequency could be some multiple of the burst frequency. If the first few pulses in a burst summate, the nerve fiber could fire, go through a brief period of refractoriness, and then fire again. If this process happens rapidly and, therefore, is repeated during the burst, the nerve fiber firing frequency will be a multiple of the burst frequency. There is sound experimental evidence for this effect.41– 45 188

f

Physical Therapy

Volume 89

A problem with multiple firing is that it could detract from the desired outcome. For example, a motoneuron firing frequency of 50 Hz might elicit an optimally forceful muscle contraction, so 50-Hz PC would be a good option. If long-duration 50-Hz bursts are used, however, the induced firing frequency could be a multiple of 50 Hz. This would initially result in a slightly greater muscle force, but the rate of fatigue would be higher. There also would be a greater amount of high-frequency fatigue. A recent study by Laufer and Elboim44 compared fatigue rates using 50-Hz bursts of 2.5-kHz AC with a burst duration of 10 milliseconds (Russian current), 50-Hz biphasic PC with the same phase duration (200 microseconds), 50-Hz bursts with a burst duration of 4 milliseconds, and 20-Hz bursts with a burst duration of 10 milliseconds. They reported that Russian current was the most rapidly fatiguing, PC was the least rapidly fatiguing, and the 2 currents of shorter burst duration were intermediate and equally fatiguing. A conclusion is that for motor stimulation using kilohertz-frequency AC bursts, if the duration is greater than 2 milliseconds, multiple firing is likely to occur and the fatigue rate will be compromised. Neural Block With kilohertz-frequency AC stimulation, another effect can be produced: direct conduction block of the nerve fiber. A direct observation of neural block was reported by Tanner,46 who measured compound action potentials produced in exposed sciatic nerve in response to direct, repetitive stimuli from a lowfrequency pulse generator and found that neural activity could be blocked using a 20-kHz AC stimulus applied to the nerve trunk between the pulse generator and the recording electrodes. As the AC stimulus intensity was progressively increased, first the fast (large-diameter) fiber re-

Number 2

sponses disappeared, followed by the slower (intermediate-diameter) fiber responses and then the slowest (small-diameter) fiber responses. Bowman and McNeal45 examined the ␣-motoneuron response to blocking signals in the frequency range of 100 Hz to 10 kHz. With high-intensity 2-kHz AC stimulation, they observed that following a brief period of firing at a very high rate (about 1 kHz), there was a progressive decrease in firing frequency, which occurred over a time frame of tens of seconds, after which activity ceased and complete conduction block occurred. At higher AC frequencies (4 kHz or more), the rate of decrease in activity was higher, with the firing frequency dropping to zero in less than a second and with stimulus intensities of 5 times the threshold. Bowman and McNeal concluded that neural block occurs more readily at multiples of threshold stimulation intensities and that the effects occur more rapidly at higher kilohertz frequencies. Direct studies of neural block with AC stimulation, to date, have all used continuous AC. There do not appear to be any reported studies of the blocking effectiveness of burstmodulated AC, so it is not known to what extent neural block contributes to the effects observed. Indirect evidence for neural block was found by Ward and Robertson,17 who measured motor thresholds using continuous kilohertz-frequency AC, 50-Hz bursts, and single sine waves in the range of 1 to 25 kHz. Irregularities in the graphs of force versus stimulus intensity were consistent with multiple firing followed by nerve block. The effects were more pronounced at higher kHz frequencies and were greater with continuous stimulation than with 50-Hz bursts. Whether neural block is of practical significance with electrical stimulaFebruary 2009

Electrical Stimulation Using Kilohertz-Frequency Alternating Current tion as used clinically thus remains uncertain, but it would affect MEIT, as ␣-motoneurons are more susceptible to neural block than pain (A-␦ and C) fibers because of their larger diameter. This means that muscle force could be diminished without any diminution of pain sensation.

Conclusion In assessing the relative merits of different forms of motor electrical stimulation, 2 factors are highly relevant: relative discomfort of stimulation and the ability to elicit maximum muscle torque. These factors, in turn, depend on the neurophysiological responses of different nerve fiber types to electrical stimulation. With kilohertz-frequency AC stimulation, summation and multiple firing, high-frequency fatigue, and neural block can potentially affect the neurophysiological response. The effects will vary, depending on the AC frequency and burst duration. Both the historical evidence and more recent findings indicate that the stimulation parameters commonly used clinically (Russian and interferential currents) are suboptimal for achieving their stated goals and that greater benefit would be obtained using short-duration (2- to 4-millisecond) bursts of kilohertzfrequency AC, with a frequency chosen to maximize the desired outcome. For maximum muscle torque production, a frequency of 1 to 2.5 kHz is indicated, with a burst duration of 2 milliseconds or so. For minimal discomfort (but less muscle torque), a frequency of 4 kHz is indicated, with a burst duration of 4 milliseconds. This article was received February 24, 2008, and was accepted November 8, 2008. DOI: 10.2522/ptj.20080060

February 2009

References 1 Robertson VJ, Ward AR, Low J, Reed A. Electrotherapy Explained: Principles and Practice. 4th ed. Oxford, United Kingdom: Butterworth Heinemann; 2006. 2 Kloth LC. Interference current. In: Nelson RM, Currier DP, eds. Clinical Electrotherapy. 2nd ed. East Norwalk, CT: Appleton & Lange; 1991:221–260. 3 Selkowitz DM. High-frequency electrical stimulation in muscle strengthening. Am J Sports Med. 1989;17:103–111. 4 Geddes LA. A short history of the electrical stimulation of excitable tissue including therapeutic applications. Physiologist. 1984;27(suppl):s1–s47. 5 Reilly JP. Electrical Stimulation and Electropathology. Cambridge, United Kingdom: Cambridge University Press; 1992. 6 Electrotherapeutic Terminology in Physical Therapy. Alexandria, VA: American Physical Therapy Association; 2000. 7 Palmer S, Martin D. Interferential current. In: Watson T, ed. Electrotherapy: Evidence-Based Practice. 12th ed. London, United Kingdom: Churchill Livingstone; 2008:297–315. 8 Andrianova GG, Kots YM, Marmyanov VA, Xvilon VA. Primenenie elektrostimuliatsii dlia trenirovki mishechnoj sili. Novosti Meditsinskogo Priborostroeniia. 1971;3: 40 – 47. 9 Ward AR, Shkuratova N. Russian electrical stimulation: the early experiments. Phys Ther. 2002;82:1019 –1030. 10 Delitto A, Brown M, Strube MJ, et al. Electrical stimulation of quadriceps femoris in an elite weight lifter: a single-subject experiment. Int J Sports Med. 1989;10: 187–191. 11 St Pierre D, Taylor AW, Lavoie M, et al. Effects of 2,500-Hz sinusoidal current on fibre area and strength of the quadriceps femoris. J Sports Med. 1986;26:60 – 66. 12 Ward AR, Robertson VJ. The variation in torque production with frequency using medium-frequency alternating current. Arch Phys Med Rehabil. 1998;79: 1399 –1404. 13 Ward AR, Robertson VJ, Ioannou H. The effect of duty cycle and frequency on muscle torque production using kHz frequency range alternating current. Med Eng Phys. 2004;26:569 –579. 14 Treffene RJ. Interferential fields in a fluid medium. Aust J Physiother. 1983;29: 209 –216. 15 Lambert HL, Vanderstraeten GG, De Cuyper HJ, et al. Electric current distribution during interferential therapy. Eur J Phys Med Rehabil. 1993;3:6 –10. 16 Soloviev NA. Nyekogorii osobynyenostii elektrostimulatsii na povishchennik chastotak, 1: optimalnii chastoti elektrostimulatsii i primenenie chastotnoy i amplitudnoy modulatsii. Trudy Instytut M Vniimio. 1963;3:162–170. 17 Ward AR, Robertson VJ. The variation in motor threshold with frequency using kHz-frequency alternating current. Muscle Nerve. 2001;24:1303–1311.

18 Ward AR, Lucas-Toumbourou S. Lowering of sensory, motor and pain-tolerance thresholds with burst duration using kHzfrequency alternating current electrical stimulation. Arch Phys Med Rehabil. 2007;88:1036 –1041. 19 Bankov S. Medium frequency modulated impulse current for electric stimulation of non-denervated muscles. Acta Medica Bulgarica. 1980;7:12–17. 20 Bankov S. Prouchvaniya vrku elektrostimulyatsiita na nedenervirani muskuli ss sinusoidalni modulirani tokovy s razlichni parametri. Kurortologiia i Fizioterapiia Sofia. 1980;1:28 –33. 21 Bankov S, Daskalov I. Electrostimulation de muscles non de´nerve´s par la courant “dynatron” et un courant d’impulsions `a fre´quence moyenne module´es. Revue ´ lectrodiagnostic Therapie. 1981;18:23–32. E 22 Ward AR, Oliver W, Buccella D. Wrist extensor torque production and discomfort associated with low-frequency and burstmodulated kilohertz-frequency currents. Phys Ther. 2006;86:1360 –1367. 23 Nemec H. Interferential therapy: a new approach in physical medicine. Br J Physiother. 1959;12:9 –12. 24 Nemec H. Reizstromtherapie mit interferenzstro ¨ men. Der Deutsche Badebetrieb. 1960;12:320 –322. 25 Nemec H. Endogene elektrostimulierung durch mittelfrequente und interferenzstro ¨ me. Rehabilitation (Bonn). 1967; 20:1–11. 26 Nemec H. Elektrostimulierung in endogener anwendung: aktionsmechanismus der interferenztherapie. Physikalische Medizin und Rehabilitation. 1968;9:73–75. 27 Burton CE, David RM, Portnoy WM, Akers LA. The application of Bode analysis to skin impedance. Psychophysiology. 1974; 11:517–525. 28 Lykken DT. Square-wave analysis of skin impedance. Psychophysiology. 1971;7: 262–275. 29 Yamamoto T, Yamamoto Y. Analysis for the change of skin impedance. Med Biol Eng Comput. 1977;15:219 –227. 30 Saladin KS. Anatomy and Physiology. Boston, MA: McGraw Hill; 1998:202–206. 31 Grimnes S. Electrovibration, cutaneous sensation of microampere currents. Acta Physiol Scand. 1983;118:19 –25. 32 Ozcan J, Ward AR, Robertson VJ. A comparison of true and premodulated interferential currents. Arch Phys Med Rehabil. 2004;85:409 – 415. 33 Babkin D, Timtsenko N, trans. Electrostimulation (notes from Dr YM Kots’ [USSR] lectures and laboratory periods presented at the Canadian-Soviet exchange symposium on electrostimulation of skeletal muscles, Concordia University, Montreal, Quebec, Canada, December 6 –15, 1977). Available from Dr Ward. 34 Gildemeister M. Zur theorie des elektrischen reizes, V: polarisation durch wechselstro ¨ me. Ber Sachs Ges Wiss. 1930;81:303–313.

Volume 89

Number 2

Physical Therapy f

189

Electrical Stimulation Using Kilohertz-Frequency Alternating Current 35 Gildemeister M. Untersuchungen u ¨ ber die wirkung der mittelfrequenzstro ¨ me auf den menschen. Pflugers Arch. 1944;247: 366 – 404. 36 Solomonow M, Eldred E, Lyman J, Foster J. Fatigue considerations of muscle contractile force during high-frequency stimulation. Am J Phys Med. 1983;62:117–122. 37 Sawan M, Hassouna MM, Li JS, et al. Stimulator design and subsequent stimulation parameter optimization for controlling micturition and reducing urethral resistance. IEEE Trans Rehabil Eng. 1996;4: 39 – 46. 38 Jones DA. High- and low-frequency fatigue revisited. Acta Physiol Scand. 1996;156: 265–270. 39 Jones DA. Muscle fatigue due to changes beyond the neuromuscular junction. In: Porter R, Whelan J, eds. Human Muscle Fatigue: Physiological Mechanisms. London, United Kingdom: Pitman Medical; 1981:178 –196.

190

f

Physical Therapy

Volume 89

40 Otsuka M, Endo M, Nonomura Y. Presynaptic nature of neuromuscular depression. Jpn J Physiol. 1962;12:573–584. 41 Stefanovska A, Vodovnik L. Change in muscle force following electrical stimulation. Scand J Rehabil Med. 1985;17: 141–146. 42 Laufer Y, Ries JD, Leininger PM, Alon G. Quadriceps femoris muscle torques produced and fatigue generated by neuromuscular electrical stimulation with three different waveforms. Phys Ther. 2001;81: 1307–1316. 43 Ward AR, Robertson VJ. The variation in fatigue rate with frequency using kHz frequency alternating current. Med Eng Phys. 2000;22:637– 646.

Number 2

44 Laufer Y, Elboim M. Effect of burst frequency and duration of kilohertzfrequency alternating currents and of low-frequency pulsed currents on strength of contraction, muscle fatigue, and perceived discomfort. Phys Ther. 2008;88: 1167–1176. 45 Bowman BR, McNeal DR. Response of single alpha motoneurons to high-frequency pulse trains. Appl Neurophysiol. 1986;49: 121–138. 46 Tanner JA. Reversible blocking of nerve conduction by alternating-current excitation. Nature. 1962;195:712–713.

February 2009

Scholarships, Fellowships, and Grants News from the Foundation for Physical Therapy First-Generation FoundationFunding Recipients Mentor CORRT Scholars Now in its 30th anniversary year, the Foundation for Physical Therapy will celebrate the more than 600 physical therapist researchers whose careers were launched with Foundation grants, scholarships, and fellowships. The Foundation is proud of their successes. Their findings have increased the evidence that clinicians can use in managing a wide range of patient needs. Their work has strengthened the profession, and their impact reverberates today through the next generation of researchers. The expanding influence of those researchers can be seen through the National Institutes of Health’s new $4.6-million Comprehensive Opportunities in Rehabilitation Research Training program (CORRT). The program aims to build investigative skills by matching 10 young researchers, most of whom have been awarded Foundation grants, with senior researchers. Michael Mueller, PT, PhD, FAPTA, is principal investigator and Anthony Delitto, PT, PhD, FAPTA, and Stuart Binder-Macleod, PT, PhD, FAPTA, are program directors of CORRT. Mueller received a Foundation research grant in 1987 to conduct a clinical trial of wound treatment for people with diabetes. His study set the “gold standard” for treatment with total contact casting. This experience started his 20-year exploration of ways to limit physical stresses on extremities and ultimately avoid amputation. His findings have changed the way the medical community views diabetic foot

February 2009

wounds—from a vascular problem to neuropathic weakness—and, as a result, interventions now include muscle-strengthening exercises. As a full-time professor and researcher at Washington University School of Medicine in St Louis, Missouri, Mueller continues to explore and publish his ideas. He also is passing on his expertise to young researchers. Mueller is mentoring Marcie Harris Hayes, PT, DPT, OCS. “The CORRT program is giving all of us the opportunity to participate in something much bigger than ourselves,” Mueller said.

• Kathleen Sluka, PT, PhD, University of Iowa, mentoring Laura Frey-Law, PT, PhD • Lynn Snyder-Mackler, PT, ScD, ATC, SCS, University of Delaware, mentoring Gregory Hicks, PT, PhD • Richard Shields, PT, PhD, FAPTA, University of Iowa, mentoring Glenn Williams, PT, PhD • Steven Wolf, PT, PhD, Emory University, mentoring Jeanne Charles, PT, PhD

Other CORRT mentors and scholars include (Figure):

• Amy Bastian, PT, PhD, Johns Hopkins University, mentoring Ya-Weng Tseng, PT, PhD, Temple University

• Linda Van Dillen, PT, PhD, Washington University, mentoring Joanne Wagner, PT, PhD, ATC, St Louis University

• James Becker, PhD, mentoring Margo Holmes, PhD, OTR/L, University of Pittsburgh, and Elizabeth Skidmore, PhD, OTR/L.

• G Kelley Fitzgerald, PT, PhD, OCS, University of Pittsburgh, mentoring Fabrisia Ambrosio, PT, PhD

Carolyn Baum, PhD, OTR/L, FAOTA, Director of the Program in Occupational Therapy at Washington University in St Louis, and Joan Rogers, PhD, OTR/L, FAOTA, Chair of the Department of Occupational Therapy at University of Pittsburgh, serve on the Executive Committee of CORRT.

• James (Cole) Galloway, PT, PhD, University of Delaware, mentoring Stacey Dusing, PT, PhD, Virginia Commonwealth University

Figure. Stuart Binder-Macleod, Michael Mueller, Anthony Delitto (seated, left to right), and the rest of the CORRT scholars and mentors.

Volume 89 Number 2 Physical Therapy ■ 191

Scholarships, Fellowships, and Grants “You can see how the Foundation’s funding keeps coming back,” Mueller noted. “The Foundation helped start the careers of most of the CORRT mentors. It also provided seed money for these young researchers to conduct studies that generate a track record. That helped to qualify them for this program and will enable them to seek larger grants afterward. The Foundation fills a niche that can’t be filled by any other institution.” For more about CORRT and the Foundation-funded researchers, see the Foundation Web site at www. FoundationForPhysicalTherapy.org. Click on “Program Information.”

Introducing the Foundation’s 2009 Scientific Review Committee Appointees The Foundation for Physical Therapy Board of Trustees announces the selection of 2 physical therapist researchers to serve as members of the Foundation’s Scientific Review Committee (SRC). The 9 member SRC reviews all doctoral scholarship and research grant applications received by the Foundation for funding. Their terms began January 1. Linda Resnik, PT, PhD, OCS, is an Assistant Professor (Research) in the Department of Community Health at Brown University Linda Resnik, PT, and a Research PhD, OCS Health Scientist at the Providence VA Medical Center. Her research portfolio includes both qualitative and quantitative research studies. Her substantive areas of research include rehabilitation service delivery, rehabilitation outcomes measurement, use

192 ■

of assistive technologies and home modifications, and rehabilitation of active duty service members and veterans with combat wounds. Dr Resnik has received over $2.5 million in research funding as principal investigator on 8 research grants, (3 NIH-funded RO3s, 2 VA HSR&D grants, 2 VA RR&D grants, and an RI foundation award). Additionally, she is or has been co-investigator on 4 project/center grants. She is an integral member of 3 dynamic research teams: the Center for Gerontology and Health Care Research at Brown; Providence VA’s HSR&D Research Enhancement Program on Quality of Care for Chronic Disease and Rehabilitation; and the Center for Restorative and Regenerative Medicine, a VA RR&D Center of Excellence. Dr Resnik serves as Editorial Board Member of PTJ and the Journal of Orthopaedic and Sports Physical Therapy and as Chairperson of the Research Committee of APTA’s Section on Health Policy and Administration. Katherine S Rudolph, PT, PhD, a Foundation doctoral research recipient in 1997, has been reappointed to the Scientific Review Committee Katherine Rudolph, and will begin PT, PhD a second 3-year term. Dr Rudolph is currently Assistant Professor in the Department of Physical Therapy and Interdisciplinary Graduate Program in Biomechanics and Movement Science at the University of Delaware, specializing in knee biomechanics and movement analysis of people with neurologic injuries. She has had experience as both an NIH principal investigator/co-

Physical Therapy Volume 89 Number 2

investigator and an NIH grant reviewer. In addition, she serves as a manuscript reviewer for a number of publications, including the Journal of Electromyography and Kinesiology, Journal of Orthopaedic and Sports Physical Therapy, Osteoarthritis and Cartilage, PTJ, and Arthritis Care and Research.

Nominations Requested for Scientific Review Committee The Foundation is seeking recommendations for individuals to serve on its Scientific Review Committee (SRC). Well-qualified physical therapist researchers will review doctoral, fellowship, and research grant applications received by the Foundation. To be considered, individuals must meet the criteria for SRC membership posted on the Foundation’s Web site (www.Foudationfor PhysicalTherapy.org) under Program Information. Selfnominations are welcome. Please e-mail your recommendations to [email protected]. [DOI: 10.2522/ptj.2009.89.2.191]

February 2009