Prim Care Clin Office Pract 33 (2006) xi–xii
Preface
William F. Miser, MD, MA John R. McConaghy, MD, FAAFP Guest Editors
In these two issues of the Primary Care Clinics of North America, we are pleased to offer you Evidence-Based Approaches to Common Primary Care Dilemmas. During the past decade, evidence-based medicine (EBM) has had a major impact on health care and the way we practice medicine in the office. As the focus of medical research has evolved from disease-based outcomes to the important, patient-oriented outcomes of mortality and quality of life, we are beginning to have answers that change the way we practice. To address this change, we present an evidence-based approach to common problems encountered by primary care physicians, written by primary care physicians. Where possible, we provide the latest recommendations based on the best available evidence. We also acknowledge when that evidence is lacking and if recommendations are based upon opinion rather than fact. We use the Strength of Recommendation Taxonomy to rate recommendations based upon the evidence within each chosen topic area. In the December 2006 issue, we present an introduction to EBM, with a focus on its definition, major steps, strengths, and challenges. We address ways in which busy clinicians can efficiently answer clinical questions that arise during a normal day at the office, and how they can stay current with the medical literature applicable to primary care. We also address how primary care physicians can critically evaluate an article, when necessary, and assess its validity and applicability as well as whether its findings should be incorporated into their practices. We begin to address common conditions and their diagnostic or treatment dilemmas. We provide the latest evidence for the screening, treatment, and follow-up of the number one cause of 0095-4543/06/$ - see front matter Ó 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.pop.2006.11.001 primarycare.theclinics.com
xii
PREFACE
death in the United States, coronary artery disease. We also address two common cardiac risk factors: essential hypertension (with its comorbid conditions) and hyperlipidemia. We chose depression as a topic because, more than ever, this condition is becoming an issue that primary care physicians are treating. We conclude the issue with two common areas pertinent to women’s health, osteoporosis and hormone replacement therapy. In the March 2007 issue, we begin with the fastest growing chronic disease in America: type-2 diabetes mellitus, with an emphasis on quality care in the office. We then address the most common acute illnesses seen in primary care practice, the upper respiratory infection and acute otitis media. Three common primary care ‘‘pains’’ are discussed: low back pain, headache, and dyspepsia. We conclude the issue with evidence related to health promotion and disease prevention. This category will cover the latest evidence on exercise and weight management, approaches to help our patients quit smoking, and the latest screening recommendations for colorectal, lung, prostate, breast, cervical, uterine, and ovarian cancers. We hope that primary care physicians will find this information helpful in providing high-quality care to their patients. Recognizing that evidence changes, we know that studies will be published that may differ from and update our recommendations. We encourage readers to stay abreast of the literature and to incorporate into their practices the best and latest evidence that will allow their patients to live longer, more satisfying lives. William F. Miser, MD, MA Associate Professor Department of Family Medicine The Ohio State University College of Medicine 2231 North High Street, Room 203 Columbus, OH 43201, USA E-mail address:
[email protected] John R. McConaghy, MD, FAAFP Associate Professor Department of Family Medicine The Ohio State University College of Medicine OSU Family Practice at Upper Arlington 1615 Fishinger Road Columbus, OH 43221, USA E-mail address:
[email protected]
Prim Care Clin Office Pract 33 (2006) 811–829
An Introduction to Evidence-Based Medicine William F. Miser, MD, MA Department of Family Medicine, The Ohio State University College of Medicine, 2231 North High Street, Room 203, Columbus, OH 43201, USA
Sir William Osler once said, ‘‘The practice of medicine is an art, based on science. Medicine is a science of uncertainty and an art of probability’’ [1]. Since this observation nearly 100 years ago, the practice of medicine has witnessed an explosion in medical knowledge with dramatic advancements in diagnostic technologies and therapeutic options. Yet, despite these changes, medicine remains a science of uncertainty for many. During the past decade, evidence-based medicine (EBM) has made a major impact on health care and the way that we practice medicine in the office. Embraced by many as the best way to practice medicine, others scorn its use, calling it arrogant, inflammatory, and misleading [2,3]. It is often misunderstood, as evidenced by the remarks made recently by a former director of the National Institutes of Health [4]. The challenge to health care providers is to provide up-to-date medical care to their patients while incorporating valid new information. The ultimate goal should be to help patients live long, functional, satisfying, and pain- and symptom-free lives. To do this requires a balance of competence and compassion. The advancement of medical practice relies on research and the scientific method. That said, a significant portion of clinical decision making is still based on personal and anecdotal experience. This article addresses the major steps, strengths, and challenges of EBM. Because of space limitations, this material is introductory, and the reader who would like to learn more is encouraged to take one or more of the free online EBM courses that are listed in Box 1.
E-mail address:
[email protected] 0095-4543/06/$ - see front matter Ó 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.pop.2006.10.001 primarycare.theclinics.com
812
MISER
Box 1. Web-based evidence-based medicine courses http://www.poems.msu.edu/infomastery: An Introduction to Information Mastery, created by Dr. Mark Ebell, a family physician, while at Michigan State University. This course introduces the basic concepts of Information Mastery, Evidence-Based Medicine, and critical appraisal of the medical literature. Each module contains goals and tools, topic-specific curriculum, and a quiz. http://www.hsl.unc.edu/services/tutorials/ebm/welcome.htm: This tutorial, developed in 2004 by a collaborative effort of the Duke University and University of North Carolina at Chapel Hill, is intended for any health care practitioner or student who needs a basic introduction to the principles of EBM, and consists of five major units: What is EBM?; The Well-Built Clinical Question; Literature Search; Evaluating the Evidence; and Testing Your Knowledge. http://www.uic.edu/depts/lib/lhsp/resources/ebm.shtml: This guide, developed by the Library of Health Sciences at Peoria, Illinois, is designed to assist health care professionals and students become effective and efficient users of the medical literature. Seven units contain information on formulating the question using the PICO method, levels of evidence, searching MEDLINE, clinical filters, other EBM databases, EBM publications, and EBM Internet references. http://library.ncahec.net/ebm/pages/index.htm: This North Carolina Evidence-Based Medicine Education Center of Excellence Web site provides a collection of EBM resources that is intended for faculty, librarians, students, and health care professionals who are interested in learning about EBM. Its goal is to make this site the preferred entry point when one wants to learn EBM, teach EBM, find current EBM research, or find key EBM resources. http://www.urmc.rochester.edu/hslt/miner/resources/ evidence_based/index.cfm: This University of Rochester Medical Center EBM Clinical Practice Tutorial walks one through the EBM process, from designing an answerable question to finding the evidence. http://library.downstate.edu/EBM2/contents.htm: This State University of New York Downstate Medical Center EBM tutorial provides an overview of EBM and focuses on developing the research question and finding the answer.
INTRODUCTION TO EVIDENCE-BASED MEDICINE
813
http://www.healthsystem.virginia.edu/internet/library/ collections/ebm/index.cfm: The University of Virginia Health Sciences Library ‘‘Navigating the Maze: Obtaining EvidenceBased Medical Information’’ is a resource page for obtaining EBM information. It provides an introduction to computer-based resources that will help to reduce the work needed to find information based upon valid and reliable evidence. http://www.cebm.net/: This Oxford-Centre for EBM site was developed by Dr. Sackett and others who were instrumental in developing EBM. This site is packed with great information! http://www.sheffield.ac.uk/wscharr/ir/netting/: This United Kingdom ‘‘Netting the Evidence – A ScHARR Introduction to Evidence-Based Practice on the Internet’’ Web site is intended to facilitate evidence-based healthcare by providing support and access to helpful organizations and useful learning resources, such as an evidence-based virtual library, software, and journals.
Definition of evidence-based medicine In 1996, Dr. David Sackett and colleagues at McMasters University in Ontario, Canada defined EBM as ‘‘the conscientious, explicit and judicious use of current best evidence in making decisions about the care of individual patients’’ [5]. Sackett and colleagues [6] went further to state that EBM is the ‘‘integration of best research evidence with clinical expertise and patient values.’’ They recognized the tension between evidence and clinical expertise, noting that, ‘‘without clinical expertise, practice risks becoming tyrannized by external evidence, for even excellent external evidence may be inapplicable to or inappropriate for an individual patient. Without current best external evidence, practice risks becoming rapidly out of date, to the detriment of patients.’’ EBM requires the integration, patient by patient, of the physician’s clinical expertise and judgment with the best available, relevant evidence. It is the misunderstanding of this balance that causes tension among clinicians when discussing and implementing EBM. The value of EBM is to enhance, not detract, from clinical expertise. It also is meant to deal scientifically, if possible, with anecdotal evidence and the philosophy of, ‘‘That’s the way I learned it in medical school, so it must be true.’’ EBM allows for healthy skepticism without necessarily discarding anecdotal evidence or personal opinion. EBM integrates best available evidence with physician knowledge, experience and skills, and patient values and preferences to make the best possible clinical decision [7]. Advantages of EBM are summarized in Box 2.
814
MISER
Box 2. Important advantages of evidence-based medicine Has the potential to improve quality of patient care Identifies and promotes practices that are proven scientifically to be effective Identifies practices that are ineffective or harmful Promotes critical thinking Requires clinicians to be open-minded Encourages researchers to focus on evidence and outcomes that are important to clinicians and patients
Evidence-based medicine focuses on patient outcomes One major advantage of EBM is that it has forced researchers to focus on outcomes that are important to clinicians and patients. The number of articles indexed in MEDLINE with the word ‘‘evidence based’’ in the title has grown dramatically over the past decade, increasing from 295 in 1997 to 1162 in 2005. EBM also has caused a shift from studies that are aimed at increasing our understanding of diseases (disease-oriented evidence; eg, etiology, prevalence, pathophysiology) to those that focus on patient-oriented evidence that matters (POEMs) [8,9]. Disease-oriented studies are crucial to medicine, but are limited in that they only deal with intermediate outcomes, such as blood pressure or glucose levels. These studies do not tell us what we really want to knowdwhether our patient will be better off undergoing this diagnostic test or using this therapy. In addition, applying the information obtained from these types of studies may result in potential harm to our patients. For example, previous studies showed that two antiarrhythmics (encainide and flecainide) suppressed ventricular arrhythmias (a disease-oriented outcome). As such, they were believed to be beneficial for patients after myocardial infarction who were at risk for sudden death. The Cardiac Arrhythmia Suppression Trial subsequently showed that subjects who were receiving these medications were nearly four times more likely to die of an arrhythmia than were similar subjects who were not taking them [10]. Obviously, focusing on the disease (arrhythmia) instead of the outcome (overall mortality) resulted in potential harm to the patients. In contrast, studies that deal with POEMs address issues that evaluate the effectiveness of diagnostic studies or interventions that patients care about, and that we, as clinicians, care about for our patients. Outcomes that are studied include issues such as quality of life, improving function, staying independent, overall mortality, and cost-effectiveness. A recent example is the use of hormonal replacement therapy to prevent cardiac disease, dementia, and other chronic diseases. The Women’s Health Initiative, a multicenter, randomized controlled trial (RCT) that involved more than 27,000 women, demonstrated that the often-used combination therapy of estrogen and
INTRODUCTION TO EVIDENCE-BASED MEDICINE
815
progesterone actually harmed more women than it potentially benefited [11]. The results of this study argue for a change in the reason why we prescribe hormones to postmenopausal women, shifting from chronic disease prevention to short-term (%5 years) management of vasomotor symptoms. Using the EBM model, armed with this new evidence, clinicians would discuss these findings with their patients so that they could make a mutual decision about the use of hormonal replacement therapy. Another recent example is the management of children who have acute otitis media (AOM). In the past, it was not uncommon for clinicians to treat children who were diagnosed with AOM with 10 days of antibiotics, because ‘‘we always have done it that way.’’ Over time, however, studies showed that children were being overtreated, which resulted in increasing antibiotic resistance. A meta-analysis questioned the treatment period of 10 days, and advocated for a much shorter period [12]. This was followed by several studies that evaluated symptomatic treatment of AOM (through the use of oral and topical analgesics) without the use of antibiotics. Recovery and potential complications were similar whether or not antibiotics were prescribed [13,14]. Most recently, an RCT of 283 children who presented to the emergency department found that, compared with the standard approach of treating all children who had AOM with antibiotics, the ‘‘wait-and-see prescription’’ approach (do not fill unless the child worsens or is not better in 48 hours) substantially reduced the unnecessary use of antibiotics. There was no difference found in subsequent fever, otalgia, or unscheduled visits for medical care [15]. Five essential steps of the evidence-based medicine model The EBM model requires five essential steps (Box 3) [6]. Step 1. Develop a specific answerable question from a clinical problem In a busy primary care practice, physicians constantly encounter patient care problems for which they have no immediate answer. In one study, Box 3. Five essential steps of the evidence-based medicine model Develop a specific answerable question from a clinical problem Search the best evidence that answers that question Critically assess that evidence for its relevance, validity, and usefulness to the patient and practice Implement the evidence that has been found to be relevant, valid, and useful into everyday practice Evaluate the performance of that practice and revise as new evidence becomes available
816
MISER
family physicians had, on average, 3.2 questions for every 10 patients that they saw, but did not seek an answer 64% of the time [16]. Most often, questions center around treatment (45%), diagnosis or evaluation (22%), etiology (4%), adverse effects of treatments or exposures (4%), epidemiology (4%), screening (3%), prognosis (3%), and prevention (2%) [17]. Studies have suggested that physicians have difficulty in developing an unstructured and complex clinical question into one that is clear, directly focused on the specific problem and person, and answerable by searching the medical literature [7]. Questions should be short and to the point [18]. To assist physicians in this skill, Sackett and colleagues suggested that a good clinical question should focus on the PICO framework, which stands for Patient or Problem; Intervention, test, or exposure of interest; Comparison interventions (if relevant); and Outcomes of interest [6,7,18,19]. To illustrate the PICO model, consider that a clinician is seeing a 2-year old girl in the office who, for the past day, has been fussy, pulling at her left ear, and has had a low-grade fever. The clinician diagnoses her with AOM, and is considering treatment options. To formulate the question, he or she would focus on the following, using the PICO format. PdWho is this particular patient, such as her age and history (eg, Is this her first episode of AOM, or one of many? Is she in otherwise good health? Does she have any other family, social, or health issues that are significant, such as her parents’ ability to follow instructions?), and what is this specific problem (an episode of AOM being seen in the first 24–48 hours of onset)? IdFor the intervention, should the clinician follow the ‘‘usual’’ management and treat her with 10 days (or shorter) of antibiotics? CdOr is the comparison intervention of the wait-and-see approach (provide analgesics, and a prescription for an antibiotic if she does not get better within the next 48 hours) a viable option? OdWhat are the outcomes that are considered important for this patient and problem (eg, reduction in symptoms and use of antibiotics; recurrence of symptoms; or development of complications, such as mastoiditis)? Specificity in formatting the question in such a manner focuses the clinician on the aspects that are important for the search of an answer, if one is available. Step 2. Search the best evidence that answers that question Once the question has been formulated, the next step is to search for relevant evidence that will help to answer that question. Searching for the evidence efficiently is important, because physicians spend an average of less than 2 minutes pursuing an answer before they give up [16]. Until recently, the search for the evidence meant doing a literature search (eg, MEDLINE), that yielded several (or more) articles that needed to be assessed critically for relevancy and validity. Although these skills are important for physicians to
INTRODUCTION TO EVIDENCE-BASED MEDICINE
817
master (see elsewhere in this issue), this approach is impractical for most busy clinicians who are seeking an answer at the point of care. Access to the Internet and availability of good secondary sources that summarize the literature and provide ‘‘useful, actionable bottom-line’’ evidence now allow the clinician to find available evidence more efficiently [20]. A list of excellent, leading EBM resources is found in Table 1 [20,21]. These resources are created and maintained by individuals who do the hard work of sifting through the latest evidence and developing summaries and recommendations for the busy clinician. Readers are advised to explore each of these resources to determine which works best for them. A word of caution is in order. Although outcome-focused research has improved over the years, gaps still exist in evidence that is relevant to the delivery of quality primary care [22]. Many times, the answer to the specific question simply does not exist, and the physician must rely on current opinion or best practices that are not based on good evidence, while waiting for the studies to be done that will answer the question. It is hoped that as EBM continues to gain momentum, more answers to our questions will become available. Step 3. Critically assess that evidence for its relevance, validity, and usefulness to the patient and practice No matter the source, the next step is to assess the validity of the evidence (critical appraisal). Validity refers to what extent the knowledge gained from the study represents ‘‘truth.’’ There are two types of validity: internal (ie, How well was the study done? Do the results reflect the truth?), and external (ie, Are the results generalizable to a larger population?). The clinician is responsible for establishing that the evidence is valid before implementing it into everyday clinical practice. It is not enough to accept evidence at face value simply because it has been published in a well-known journal or comes from a specialist. The clinician should have a working knowledge of common EBM terms (Box 4). Details on how to assess an individual article critically are found elsewhere in this issue. Primary studies represent ‘‘original research’’ and consist of experiments (something is performed on subjects in an artificial or controlled environment), clinical trials (an intervention, such as a new medication, is offered to a group of patients that is followed to see what happens to them), and surveys (something is measured in a group). Depending on the type of question being asked, there is a preferred methodology to answer that question (Table 2), with an established ‘‘hierarchy’’ of evidence. Box 5 provides the hierarchy for research that involves interventions [23,24]. A case report is a description of a single patient who has an outcome of interest, whereas a case series is a series of patients that has an outcome of interest but no control group. These usually form the basis for further studies. In observational studies, the investigator does not alter or manipulate variables and does not assign subjects to groups randomly; one does analyze
818
MISER
Table 1 Current reliable evidence-based medicine resources for the busy clinician Resource
Description
American College of Physicians Journal Club http://www.acpj.org
Published bimonthly, the general purpose is to select from the biomedical literature articles that report original studies and systematic reviews that warrant immediate attention by physicians attempting to keep pace with important advances in internal medicine. More than 170 journals are researched for clinically relevant, methodologically sound studies. Published twice a month, contains POEMs, Cochrane for Clinicians, and Practice Guidelines that are of relevance to primary care physicians. ‘‘Evidence based thinking about health care’’ is the motto for this monthly journal written by Oxford scientists who search PubMed and the Cochrane Library for systematic reviews and meta-analyses that are ‘‘both interesting and make sense.’’ ‘‘The international source of the best available evidence for effective health care’’ that summarizes what is known, and not known, about more than 200 medical conditions and 2000 treatments. A free copy for primary care physicians can be obtained from the United Health Foundation, http://www.unitedhealthfoundation.org/ registration.cfm ‘‘The definitive resource for electronic information’’ in the EBM movement, this database regularly performs meta-analyses of studies and writes detailed, structured reviews that allow clinicians to get fast, ‘‘bottom line’’ answers to their most commonly-asked questions. These reviews are updated regularly. DARE contains summaries of systematic reviews that have met strict quality criteria. Included reviews are about the effects of interventions. Each summary also provides a critical commentary on the quality of the review. The database covers a broad range of health and social care topics and can be used for answering questions about the effects of interventions, as well as for developing guidelines and policy making. Developed and maintained by a family physician, this site contains several links to many sites listed in this table and are organized primarily for primary care. Created by a family physician, a medical information point-of-care reference resource that summarizes latest evidence on nearly 1,800 topics.
American Family Physiciana http://www.aafp.org/afp Bandoliera http://www.rj2.ox.ac.uk/ bandolier
Clinical Evidence http://www.clinicalevidence.com
Cochrane Database of Systematic Reviewsa http://www.cochrane.org/ reviews/en/
Database of Abstracts of Reviews of Effects (DARE)a http://www.york.ac.uk/ inst/crd/crddatabases.htm
Dr. Alper’s Useful Linksa http://www.myhq.com/public/a/ l/alper DynaMed http://www.dynamicmedical.com
(continued on next page)
INTRODUCTION TO EVIDENCE-BASED MEDICINE
819
Table 1 (continued ) Resource
Description
Family Practitioners Inquiries Network (FPIN) Clinical Inquiriesa http://www.fpin.org
Clinical inquiries are a series of answers based on the best possible evidence to practicing physicians’ questions, researched by primary care physicians as part of FPIN. These inquiries also are published in the Journal of Family Practice. Evidence from the Cochrane Database, National Guidelines Clearinghouse, Clinical Evidence, MD Consult, and Consensus Statements are reviewed by practicing clinicians and regional and national experts to develop reviews on a variety of common problems. Developed by family physicians, this system consists of two components. DailyInfoPOEMs provides valid, relevant POEMs sent by way of daily e-mail. InfoRetriever allows one to simultaneously search complete POEMs database along with 6 additional evidence-based databases. An independent, nonprofit organization, ICSI is a collaboration of nearly 50 health organizations that review quality of care guidelines for preventive services and disease management. Monthly journal that contains POEMs, clinical inquiries, and other evidence-based topics that are of relevance of primary care physicians. A unique search engine that gathers clinical information from DARE, MEDLINE, and the National Guideline Clearinghouse, with links to appropriate articles, all collated into one page. Turning Research into Practice (TRIP) is a search engine of more than 55 evidence-based databases covering material relevant to primary care. A comprehensive evidence-based clinical information resource available to physicians on the Internet, CD-ROM, and Pocket PC. It is designed to provide physicians with concise, practical answers at the point of care to a multitude of common problems encountered in primary care. Extensive database of evidence-based clinical practice guidelinesdgood first source to locate guidelines. Sponsored by the Agency for Healthcare Research and Quality (AHRQ), an independent panel of experts in primary care and prevention that systematically reviews the latest evidence of effectiveness and develops recommendations for clinical preventive services.
FIRSTConsult http://www.firstconsult.com
InfoPOEMs – The Clinical Awareness System http://www.infopoems.com
Institute for Clinical Systems Improvement (ISCI)a http://www.icsi.org/knowledge Journal of Family Practicea http://www.jfponline.org SUMSearcha http://sumsearch.uthscsa.edu
TRIP Databasea http://www.tripdatabase.com UpToDate http://www.uptodate.com
US National Guideline Clearinghousea http://www.guidelines.gov U.S. Preventive Services Task Force (USPSTF) Recommendationsa http://www.ahrq.gov/clinic/ uspstfix.htm a
Free online access.
820
MISER
Box 4. Common terminology used in evidence-based medicine Absolute risk and absolute risk reduction: Percentage of subjects in a group that experiences a discrete bad outcome, such as admission to the hospital or death. An effective therapy serves to reduce that risk. For example, if 20% of the group that received placebo had a myocardial infarction compared with 12% in the group that received active treatment, the absolute risk reduction (ARR) for myocardial infarction from the therapy is 8%. Bias: Any factor that might change the results of a study from what they would have been if that factor were NOT present. Types of bias include selection (ways in which subjects are selected for the study), attrition (withdrawals or exclusions of people entered into the study), and detection (how outcomes are assessed). Blinding: Also known as masking; the process in which subjects, investigators, or assessors remain ignorant concerning the treatments that subjects are receiving. A single-blind study occurs when the subjects or those who are making assessments are blind to the allocation, but not both. A double-blind study occurs when subjects and assessors are blinded to the allocation. Concealment of allocation: The process used to prevent foreknowledge of group assignment in an RCT. 95% confidence interval: An estimate of precision. If a study is repeated 100 times, the results will fall within this range 95 times. Likelihood ratio (LR): The percentage of patients that is positive by the ‘‘gold standard’’ for a particular disease, condition, or injury who have a particular test result, divided by the percentage of patients that does not have a problem but who have that same test result. LR of greater than 1 indicates an increased likelihood of disease; an LR of less than 1 indicates a decreased likelihood of disease. Generally, the most helpful tests have a ratio of less than 0.2 or greater than 6. Negative predictive value: Percentage of patients with a negative test for a disease who do not have the disease in question. Number needed to treat (NNT): The number of patients that needs to receive an intervention, instead of the alternative, for one additional patient to benefit. The NNT is calculated as 1/ARR. For example, if the ARR is 10%, the NNT = 1/10% = 1/0.1 = 10.
INTRODUCTION TO EVIDENCE-BASED MEDICINE
821
Pretest probability: Probability of disease before a test is performed. Positive predictive value: Percentage of patients with a positive test for a disease who have the disease in question. Posttest probability: Probability of disease after a test is performed. Relative risk reduction (RRR): The percentage difference in risk or outcomes between treatment and control groups. Example, if death from a myocardial infarction is 25% in controls, and 20% in those who receive treatment, the RRR is (25 20)/25 = 20%. Sensitivity: Percentage of patients with disease who have a positive test for the disease in question. Specificity: Percentage of patients without disease who have a negative test for the disease in question.
the data using statistical tests. The three major types of observational studies include cross-sectional, case-control, and longitudinal cohort studies. In a cross-sectional (prevalence) study, measurements are made all at once (one point in time), looking at disease status and exposure factors. An example would be studying the prevalence of diabetes in a practice, and determining the factors that are associated with the disease. The advantages of these types of studies are that they yield prevalence rates, one may study several outcomes at one time, and they are inexpensive and quick to Table 2 Preferred study design based on type of question being asked Clinical question being asked
Preferred study design
Therapy - Tests the effectiveness of a treatment such as a drug, surgical procedure, or other intervention
Randomized, double-blinded, placebo-controlled trial
Diagnosis and screening - Measures the validity (is it dependable?) and reliability (will the same results be obtained every time?) of a diagnostic test, or evaluates the effectiveness of a test in detecting disease at a presymptomatic stage when applied to a large population.
Cross-sectional survey (comparing the new test with a ‘‘gold standard’’)
Prognosis - Determines what is likely to happen to someone whose disease is picked up at an early stage
Longitudinal cohort study
Causation - Determines whether a harmful agent is related to the development of an illness
Cohort or case-control study, depending on how rare the disease is, but case reports also may provide crucial information.
Adapted from Greenhalgh T. How to read a paper–getting your bearings (deciding what the paper is about). BMJ 1997;315:243–6; with permission.
822
MISER
Box 5. The hierarchy of evidence for interventions (from highest to lowest) N-of-1 randomized trial Systematic reviews and meta-analyses RCTs involving multicenter sites RCTs at one location Cohort studies Case-control studies Cross-sectional studies Case reports and case series Expert opinion not based on research
complete. Major disadvantages are that one cannot establish causality (one does not know if the exposure preceded the outcome), they do not yield incidence of disease, and they are not feasible for rare conditions. Case-control studies usually are retrospective in nature. The investigator identifies groups of subjects who do and do not have disease, and then looks backward in time to identify the presence or absence of risk factors. A classic example is vaginal cancer and maternal exposure to diethylstilbestrol. These studies are useful for rare conditions and are inexpensive and quick to complete. Disadvantages include potential bias from sampling two populations, they do not yield prevalence or incidence rates, selection of controls may be a problem, and exposure data may be subject to biased recall. A cohort is a group of subjects who have something in common and are followed over time. There are two basic types of cohort studies. In a prospective cohort study, the investigator defines the sample and measures variables before any disease has occurred. In a retrospective cohort study, the investigator defines the sample and measures the variables after the disease has occurred. An example is identifying a cohort of women who were born in the 1970s and who are currently on oral contraceptives; the variable is smoking and the outcome is myocardial infarction. Strengths and weaknesses of cohort studies are found in Box 6. Often, RCTs are considered the ‘‘gold standard’’ for questions that deal with treatment. In this type of study design, participants are allocated randomly by a process that is equivalent to the flip of a coin to one intervention (eg, a new medication) or another (eg, placebo). Both groups are followed for a specified period, and results are analyzed in terms of outcomes defined at the outset (eg, death, heart attack, serum cholesterol). Some trials that compare an intervention group with a control group are not randomized trials. Random allocation may be impossible, impractical, or unethical (eg, one cannot make someone start smoking). Strengths and weaknesses of RCTs are found in Box 7.
INTRODUCTION TO EVIDENCE-BASED MEDICINE
823
Box 6. Strengths and weaknesses of cohort studies Strengths Time sequence of exposure and outcome is known Can measure the risk for a bad outcome directly Can study many outcomes of a single variable Good for occupational diseases Weaknesses Need large numbers, especially for rare diseases Potentially expensive Time consuming Attrition may be a problem
A special type of RCT is an N-of-1 randomized trial, in which the patient undergoes pairs of treatment periods. These periods are organized so that one period involves the experimental treatment (eg, a new drug) and the other involves an alternate or placebo therapy. In this manner, the patient acts as his/her own control. Preferably, the physician and patient are blinded to the allocation; the order of the periods is randomized; and specific outcomes are monitored, usually through a patient diary. Treatment periods are replicated and the trial is continued until the patient and clinician conclude that the patient is, or is not, obtaining benefit from the experimental treatment [23]. Integrative studies attempt to summarize and draw conclusions from primary studies. There are various types of integrative studies. The simplest, and least stringent, is a nonsystematic review that is written as a continuing medical education article by authors who are ‘‘expert’’ in a subject matter. Typically, these articles are filled with opinions and clinical experiences; a cursory, if any, literature review is performed. References used for support usually are review articles themselves. The authors may choose the studies that support their argument, while ignoring those that disagree with their point of view. Research has found that the greater the expertise of the reviewer, the lower the quality of the review [25]. For example, a review article on screening for prostate cancer that is written by a urologist may reach a completely different conclusion than one that is written by a primary care physician, although the literature available to both is the same. This type of subjective review article is fraught with bias and may be misleading [26,27]. A systematic review is a more precise integrative study. This type of review article provides an overview of original research using a precise protocol with a statement of objectives and a literature review that is conducted according to specific and reproducible methodology [28]. Investigators
824
MISER
Box 7. Strengths and weaknesses of randomized controlled trials Strengths Allow rigorous evaluation of a single variable in a precisely defined patient group Data are collected on events that occur after the study is started (prospective) Use hypotheticodeductive reasoning (seeks to falsify, rather than confirm, its own hypothesis - considered the ‘‘pure’’ scientific method) Potentially eradicate bias by comparing two otherwise identical groups Allow for meta-analysis (combining the numerical results of several similar trials) Weaknesses Expensive and time consuming Most are funded by large research bodies or drug companies that ultimately dictate the research agenda Surrogate end points are used often in preference to clinical outcome measures, which may introduce ‘‘hidden bias,’’ especially through: imperfect randomization: failure to randomize all eligible patients (clinicians only offer participation in the trial to patients whom they consider will respond well to the therapy) failure to blind evaluators to the randomization status of patients
provide a detailed description of how the articles were obtained, and the methods by which articles were included and excluded for consideration. Much like conducting original research, investigators follow a protocol in researching primary studies. When assessing such a review, the reader can judge the quality of science that went into the writing of the paper. A meta-analysis is a type of systematic review of the literature that involves combining and analyzing data from individual studies, most often RCTs [29,30]. A meta-analysis should have a detailed written protocol prepared in advance that includes a narrowly focused question to be answered. Original studies are found that address this question, their data are combined, statistical tests are applied to these combined results, and the results are reported. In addition to reviewing the primary research critically, the investigators also combine the results statistically. It literally is a ‘‘study of studies’’ [31].
INTRODUCTION TO EVIDENCE-BASED MEDICINE
825
Other types of integrative studies that use the results of primary studies include decision analyses, which generate probability trees that health professionals and patients use in making choices about clinical management; economic analyses, which determine whether a particular course of action is a good use of resources; and clinical practice guidelines (CPGs), which are ‘‘systematically developed statements to assist practitioner and patient decisions about appropriate health care for specific clinical circumstances’’ [32]. Potential benefits of evidence-based CPGs are to improve the quality of care received by patients, quality of clinical decisions, health outcomes, consistency of care, and efficiency of care [33,34]. They also can clarify to the clinician which interventions are of proved benefit and document the quality of supporting data [33]. Clinical practice guidelines have improved over the years. Before EBM, most of them were ‘‘BOGSATs’’ (created by a ‘‘Bunch of Old Guys Sitting Around Talking’’) [20]. Often, they were flawed by conflict of interest, specialty turf battles, and opinions instead of scientific evidence [34]. Having the words ‘‘evidence based’’ in the title of CPGs does not necessarily mean they truly are evidence based [23]. It is important for clinicians to assess critically any guidelines that they are contemplating using in their practice, focusing on the methods by which the guidelines were developed. Evaluate the composition of the panel that developed these guidelines [35]. Does the panel consist of expert clinicians from a variety of specialties, including primary care, and health services researchers with expertise in epidemiology and methodologic research? Did the panel select an area or question that is defined clearly? Did it perform a thorough search for the evidence, and how did it determine whether to include the evidence? How did it categorize the evidence? How did it summarize the evidence and come to its recommendations? Well-written, evidence-based CPGs should provide answers to the above questions. They should also make clear recommendations, with a grade attached as to the strength of those recommendations. A panel of family practice educators developed a Strength of Recommendation Taxonomy that addresses the quality, quantity, and consistency of evidence (Table 3) [36].
Table 3 Strength of recommendation taxonomy for evidence-based clinical practice guidelines Strength of recommendation
Definition
A
Recommendation based on consistent and good quality patient-oriented evidence Recommendation based on inconsistent or limited quality patient-oriented evidence Recommendation based on consensus, usual practice, opinion, disease-oriented evidence or case series for studies of diagnosis, treatment, prevention, or screening.
B C
826
MISER
Step 4. Implement the evidence that has been found to be relevant, valid, and useful into everyday practice This step involves integrating the evidence with clinical expertise and applying it to the patient who is sitting in the office. This is a time when one takes into account the patient’s values and circumstances, and then discusses with him or her the efficacy and risks so that an informed decision can be made. Unfortunately, some clinicians find this step difficult to implement. Studies have shown that it takes, on average, about 17 years for new knowledge that is generated by RCTs to be incorporated into practice, and even then the application is variable [37]. Change is always difficult, especially if the practice has been well established. For example, despite the overwhelming evidence that patients who have acute bronchitis do not need antibiotics, most primary care physicians still prescribe them. Old ways die hard. Shaughnessy and Slawson [38] humorously outlined ways that physicians resist this change, such as simply not paying attention to the evidence, attacking the evidence, following the pack, deferring to experts, blaming patients, and blaming the fear of malpractice. In a systematic review, Cabana and colleagues [39] studied why clinical practice guidelines are not adopted by clinicians. They found that the major barriers to implementation were lack of awareness that the guidelines existed, lack of familiarity with the guidelines, lack of agreement, lack of self-efficacy, lack of outcome expectancy, and inertia of previous practice. They also found that there were guideline barriers (eg, some not being easy to use or not convenient), patient-related barriers (eg, resistance to follow recommendations), and environmental-related barriers (eg, not having the resources or facilities needed to follow the guidelines). Despite these barriers, using the steps that are outlined in this article will facilitate the adoption of well-researched and well-written guidelines. Periodic review of the resources that are outlined in Table 1, such as the US National Guideline Clearing House, the US Preventive Services Task Force, and InfoPoems, will allow one to stay current on what guidelines are valid and relevant to primary care. Step 5. Evaluate the performance of that practice and revise as new evidence becomes available Once new evidence is incorporated into the practice, the clinician should constantly assess its effectiveness. This is also a time to assess how well one is doing in incorporating EBM into the practice, and to improve one’s ability to follow the previously mentioned four steps. Regular use of the resources listed in Table 4 will improve one’s performance. Also, having a good reading plan will allow one to find POEMs as they are reported. The ‘‘high impact journals’’ are those that are most often
INTRODUCTION TO EVIDENCE-BASED MEDICINE
827
Box 8. Journals with highest frequency of articles that contain patient oriented evidence that matters (POEMs) High-impact journals (those cited most frequently by others) Annals of Internal Medicine British Medical Journal Journal of the American Medical Association Lancet New England Journal of Medicine Primary care journals Annals of Family Medicine (not available in 1999) Archives of Internal Medicine Journal of the American Board of Family Practice Journal of Family Practice Journal of General Internal Medicine Other journals American Journal of Emergency Medicine Arthritis and Rheumatology Journal of the American College of Cardiology Medical Decision Making Obstetrics and Gynecology Pediatrics
referenced by others and include the Annals of Internal Medicine, the British Medical Journal, the Journal of the American Medical Association (JAMA), Lancet, and the New England Journal of Medicine. Although these journals frequently are filled with disease-oriented evidence, they also include those POEMs that are most likely to change the way primary care physicians practice medicine. In their review of the medical literature, Ebell and colleagues [40] evaluated the frequency of POEMs in each of the major journals (Box 8).
Summary Incorporating EBM into one’s practice will not only make one a better clinician, it also allows one to provide the best possible quality of medical care to his or her patients. Becoming a medical information master is a task that all can learn [8,9]. In a primary care specialty that, by definition, is broad in scope, and with the seemingly overwhelming amount of medical literature that is produced on a daily basis, this task is essential. One constantly should be on the look out for validated evidence that is relevant to
828
MISER
everyday practice, focusing on those POEMs that address issues that are common to primary care practice. Using the tools and steps that are outlined in this article, along with taking the Web-based courses that are mentioned in Box 1, will allow the primary care physician to develop the essential skills that are required in today’s practice of medicine.
References [1] Silverman ME, Murray TJ, Bryan CS, editors. The quotable Osler. Philadelphia: American College of Physicians; 2003. [2] Wilson CF, Callaway CH. Evidence-based medicine: ready for prime time? N C Med J 2004; 65:285–7. [3] Krist A. Evidence-based medicine: how it becomes a 4-letter word. J Fam Pract 2005;54(7): 604–6. [4] Healy B. Who says what’s best? US News World Rep 2006;(September):3. [5] Sackett DL, Rosenberg WM, Gray JA, et al. Evidence based medicine: what it is and what it isn’t [editorial]. BMJ 1996;312(7023):71–2. [6] Sackett DL, Straus SE, Richardson WS, et al, editors. Evidence-based medicinedhow to practice and teach EBM. New York: Churchill Livingstone; 2000. [7] Akobeng AK. Principles of evidence based medicine. Arch Dis Child 2005;90:837–40. [8] Slawson DC, Shaughnessy AF, Bennett JH. Becoming a medical information master: feeling good about not knowing everything. J Fam Pract 1994;38:505–13. [9] Shaughnessy AF, Slawson DC, Bennett JH. Becoming an information master: a guidebook to the medical information jungle. J Fam Pract 1994;39:489–99. [10] The Cardiac Arrhythmia Suppression Trial (CAST) Investigators. Preliminary report: effect of encainide and flecainide on mortality in a randomized trial of arrhythmia suppression after myocardial infarction. N Engl J Med 1989;321:406–12. [11] Rossouw JE, Anderson GL, Prentice RL, et al. Risks and benefits of estrogen plus progestin in healthy postmenopausal women: principal results from the Women’s Health Initiative Randomized Controlled Trial. JAMA 2002;288(3):321–33. [12] Kozyrskuh AL, Hildes-Ripstein GE, Longstaffe SEA, et al. Treatment of acute otitis media with a shortened-course of antibiotics: a meta-analysis. JAMA 1998;279:1736–42. [13] Little P, Gould C, Williamson I, et al. Pragmatic randomised controlled trial of two prescribing strategies for childhood acute otitis media. BMJ 2001;322:336–42. [14] McCormick DP, Chonmaltree T, Pittman C, et al. Nonsevere acute otitis media: a clinical trial comparing outcomes of watchful waiting versus immediate antibiotic treatmentPediatrics 2005;115:1455–65. [15] Spiro DM, Tay KY, Arnold DH, et al. Wait-and-see prescription for the treatment of acute otitis mediada randomized controlled trial. JAMA 2006;296:1235–41. [16] Ely JW, Osheroff JA, Ebell MH, et al. Analysis of questions asked by family doctors regarding patient care. BMJ 1999;319:358–61. [17] Alper BS, White DS, Ge B. Physicians answer more clinical questions and change clinical decisions more often with synthesized evidence: a randomized trial in primary care. Ann Fam Med 2005;3:507–13. [18] Coumou HCH, Meijman FJ. How do primary care physicians seek answers to clinical questions? A literature review. J Med Libr Assoc 2006;94:55–60. [19] Weinfeld JM, Finkelstein K. How to answer your clinical questions more efficiently. Fam Pract Manag 2005;12(7):37–41. [20] White B. Making evidence-based medicine doable in everyday practice. Fam Pract Manag 2004;11(2):51–8. [21] Alper BS. Practical evidence-based internet resources. Fam Pract Manag 2003;10(7):49–52.
INTRODUCTION TO EVIDENCE-BASED MEDICINE
829
[22] Mickan S, Askew D. What sort of evidence do we need in primary care? General practitioners need evidence from and about the patients they see. BMJ 2006;332:619–20. [23] Guyatt GH, Haynes RB, Jaeschke RZ, et al. Users’ Guides to the Medical Literature. XXV. Evidence-based medicine: principles for applying the users’ guides to patient care. JAMA 2000;284:1290–6. [24] Evans D. Hierarchy of evidence: a framework for ranking evidence evaluating healthcare interventions. J Clin Nurs 2003;12:77–84. [25] Strite S, Stuart ME. Evaluating clinical literature. Fam Pract Manag 2004;11(9):14, 16. [26] Shaughnessy A, Slawson D. Getting the most from review articles: a guide for readers and writers. Am Fam Physician 1997;55:2155–60. [27] Oxman A, Guyatt G. Guidelines for reading literature reviews. CMAJ 1988;138:697–703. [28] Greenlaugh T. How to read a paperdpapers that summarise other papers (systematic reviews and meta-analyses). BMJ 1997;315:672–5. [29] Miser WF. Applying a meta-analysis to daily clinical practice. J Am Board Fam Pract 2000; 13(3):201–10. [30] Egger M, Smith G, Phillips A. Meta-analysis: principles and procedures. BMJ 1997;315: 1533–7. [31] Kassirer J. Clinical trials and meta-analysis: what do they do for us? N Engl J Med 1992;327: 273–4. [32] Field MJ, Lohr MJ, editors. Clinical practice guidelines: directions for a new program. Washington, DC: National Academy Press; 1990. [33] Woolf SH, Grol R, Hutchinson A, et al. Clinical guidelinesdpotential benefits, limitations, and harms of clinical guidelines. BMJ 1999;318:527–30. [34] O’Connor PJ. Adding value to evidence-based clinical guidelines. JAMA 2005;294:741–3. [35] Shekelle PG, Woolf SH, Eccles M, et al. Clinical guidelinesddeveloping guidelines. BMJ 1999;318:593–6. [36] Ebell MH, Siwek J, Weiss BD, et al. Strength of Recommendation Taxonomy (SORT): a patient-centered approach to grading evidence in the medical literature. Am Fam Phys 2004;69: 548–56. [37] Balas EA, Suzanne AB. Managing clinical knowledge for health care improvement. Bethesda (MD): Yearbook of Medical Informatics National Library of Medicine; 2000. [38] Shaughnessy AF, Slawson DC. Easy ways to resist change in medicine. BMJ 2004;329: 1473–4. [39] Cabana MD, Rand CS, Powe NR, et al. Why don’t physicians follow clinical practice guidelines? A framework for improvement. JAMA 1999;282:1458–65. [40] Ebell MH, Barry HC, Slawson DC, et al. Finding POEMs in the medical literature. J Fam Pract 1999;48:350–5.
Prim Care Clin Office Pract 33 (2006) 831–837
Evolving Medical Knowledge: Moving Toward Efficiently Answering Questions and Keeping Current John R. McConaghy, MD, FAAFP Department of Family Medicine, The Ohio State University College of Medicine, OSU Family Practice at Upper Arlington, 1615 Fishinger Road, Columbus, OH 43221, USA
One of the most difficult tasks for any physician is to keep his/her fund of knowledge up to date. This can be a challenge, because it is estimated that medical information doubles approximately every 5 years [1]. The number of sources of information, as well as their format and diversity, are growing at a rapid pace [2]. In addition, there has been an explosion of Internet sites that contain medical information. Therefore, it is not surprising that clinicians often feel paralyzed by the amount of information that is available. Physicians seek information for two primary reasons. The first is to stay up to date with new and updated clinical information that is relevant to their practice. The second is to find answers to specific patient questions [3]. To provide the best possible patient care, clinicians must have access to resources that will assist them in making clinical decisions. The tools and methods to retrieve relevant information need to match these varied information needs. An active clinical practice, although with time constraints, generates many specific patient questions. Clinicians generate 1 to 3 questions for every three patient visits [4,5]. Of these, physicians may seek answers to 4 of 10 questions generated and find answers to only 3 of them. In other words, 7 of every 10 clinical questions go unanswered [6]. This likely is due to the amount of time and effort that it takes to find the needed answers. To answer these questions, one basically needs to conduct a Medline search to answer each clinical question. Yet, primary care physicians rarely seek original research in this manner to answer clinical questions during practice. During an observation of 1101 questions during practice, this type of research occurred only twice [6]. Even family medicine residents who are trained in
E-mail address:
[email protected] 0095-4543/06/$ - see front matter Ó 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.pop.2006.10.002 primarycare.theclinics.com
832
MCCONAGHY
evidence-based medicine select rapid and convenient sources over evidencebased searches during clinical practice [7]. Physicians, like other adult learners, seek and retain new knowledge when they have the need to know. In addition, they also require validated evidence that has direct benefit to their patients [8]. This information, however, is not readily available in the typical primary literature. Of more than 8000 articles that were published in 85 journals over a 6-month period, only 2.6% of these articles met the criteria of being valid and relevant new information [9]. There are other barriers to finding useful information, such as a lack of time to search and review new medical information [10–12], and the lack of skills to find new information efficiently and to appraise it appropriately [11–14]. A new barrier to finding useful information is the increasing prevalence of lower-quality information on the Internet. Low-quality electronic information is easy to produce and distribute because it costs much less to produce than does the print version [15]. Several commercial products are available that condense large amounts of information into easily digestible reviews; they frequently rely on overviews by experts and may not be current. The focus in the recent past has been to educate physicians and physicians-in-training in the techniques of critical appraisal of the literature (eg, evidence-based medicine). It has been suggested, however, that it is more important that clinicians learn how to find, evaluate, and incorporate available information into their clinical practice (ie, become proficient in information management). These information management skills include identifying relevant and valid information for keeping up to date, locating needed (‘‘just in time’’) information quickly at the point of care, and combining the identified evidence with patient preferences to provide bestevidence, patient-centered care [16]. Some observations suggest that the primary medical literature is the predominant method that health professionals use to stay current [17]. Medical librarians believe that the medical literature can be useful in clinical practice. Evidence suggests, however, that practicing physicians do not agree [18]. It is useful, then, to define the type of information that physicians seek more precisely. Clinicians seek information to help them fulfill the primary goal of medical practice: help patients live long, healthy, and symptom-free lives [19]. Stated a different way, physicians desire information that evaluates the effectiveness of medical care on outcomes that are important: morbidity, mortality, and quality of life. This type of information has been named ‘‘patient-oriented evidence that matters’’ (POEMs) [20]. Only a small portion of the published medical literature is relevant if we use this definition. One survey revealed that physicians can avoid having to review a large number of journals on a regular basis because few POEMs are published, even by prestigious and widely-read journals [9]. In 1960, the National Library of Medicine created Medline, which now contains more than 13 million citations from nearly 5000 journals [21]. To be more efficient in searching these citations, however, one needs to
KEEPING CURRENT
833
understand Medical Subject Headings as well as how to increase the sensitivity of one’s search or adjust its specificity [22,23]. More than half a million new articles are published yearly; however, not all medical articles are indexed on Medline and many that are published are misclassified [20]. With regards to the primary care literature, few family medicine articles are published [24]. One study showed that if the information gathered from searches can be generalized, and if the time and cost of performing searches could be reduced, then the usefulness of the medical literature in answering daily clinical questions would improve [18]. Subsequently, PubMed was developed to try to fulfill these needs by searching the Medline database using a more intuitive approach [1]. It was refined in 1996 by applying clinical filters and by introducing Clinical Queries to make literature searches more sensitive or more specific [25]. Still, clinicians have varying information needs, and the information in the medical literature has varying levels of relevance to them. Clinicians often have specific clinical questions that are answered best with a single article or systematic review, rather than a broad topic overview or broad literature search. Despite the development of PubMed, original research, as typically published, still is not useful in the care of patients until it has been transformed into a form that is useful [2]. Clinicians, then, need resources to be able to identify (and use) high-quality information and resources to retrieve the needed information rapidly [26]. A growing number of sources of highly filtered, highly relevant information is available. These sources attempt to fulfill the requirement for information that has been evaluated for validity and scientific rigor as well as relevance to clinical practice. More importantly, it should be summarized in a form that is useful for keeping abreast of new developments or for answering clinical questions [2], thus bridging the gap between research and practice. Finding this information should require the least amount of time and effort. Electronic media and the Internet play key roles in providing quick accessibility to this information. All of these factorsdrelevance, validity, and minimizing workdfacilitate the incorporation of evidence into patient care and medical practices [20,27]. Clinicians rarely take the time or have the resources or skill to evaluate the literature and find the high-quality information that they seek; thus it is imperative that comprehensive and systematic literature surveillance mechanisms exist [20,28]. Many physicians rely on article abstracts to find the information that they need. Subsequently, the construct of article abstracts has shifted from unstructured to more highly structured [29]. Although one might predict that more highly structured abstracts would be more useful, the construct of the abstract itself does not influence physician decision making [30]. How the results are presenteddwhether in the body of the manuscript or in the abstractdhas more influence on clinical decision making: relative risk and relative risk reduction are more influential than are absolute risk difference or numbers needed to treat [31,32]. Highly structured abstracts
834
MCCONAGHY
do not influence clinical decision making; having valid information presented in the context of a clinical scenario does [28]. Many clinicians turn to summary reviews to answer their questions or to keep current. These sources of information have been in existence for a long time as books, reviews articles, expert statements, consensus reports, and the like [2]. This type of information attempts to synthesize information into a format that is more useful in clinical decision making by making such information easier to read. One must use caution, however, in selecting which reviews or recommendations to incorporate. Although review articles are prominent in the primary care literature, fewer than 25% were constructed using a systematic and scientific approach [33]. The conclusions and recommendations of poorly constructed reviews should be used with extreme caution because incorporating poor information may be more harmful than useful. The editors of a prominent primary care journal recently set forth recommendations for constructing a quality review article [34]. Quality reviews should provide an outline of how the literature search was conducted; identify the included evidenced-based sources of information (eg, Cochrane Collaboration, the British Medical Journal’s [BMJ’s] Clinical Evidence), present and incorporate evidence based on clinical outcomes relating to morbidity, mortality, and quality of life; and include studies of primary care populations. Quality review articles will rate the level of evidence obtained and the strength of the recommendations using the ‘‘strength of recommendation taxonomy’’ or a similar scheme [35]. The Cochrane Library is a tool for physicians and providers who seek highquality, systematic, and rigorous synthesis of the literature [36–38]. The Cochrane Collaboration was created in 1993 as an international and independent clearinghouse for providing and disseminating up-to-date, clinically relevant systematic reviews of health care interventions [39]. These reviews are published quarterly in The Cochrane Library as the Cochrane Database of Systematic Reviews [40]. The Library is a valuable, online, authoritative reference that can help physicians with everyday treatment decisions. A complimentary database is The Database of Abstracts of Reviews of Effectiveness (DARE) [41]. DARE, prepared by the National Health Centre for Reviews and Dissemination at the University of York, England, United Kingdom, complements the Cochrane Database by offering a database of structured abstracts of quality-assessed reviews, many of which do not have a Cochrane Review. Clinical Evidence is another online collection of evidence on the benefits and harms of therapeutic interventions. The evidence is retrieved, appraised, and synthesized on topics that focus on outcomes that matter to patients [42]. One of the most comprehensive databases of evidence-based information is InfoRetriever [43]. This tool was developed by family physicians to meet the information needs of busy clinicians. InfoRetriever searches POEMs, evidence-based guidelines, the Cochrane Database, Family Practice Inquiries Network Answers, and other evidence-based sources of information. The
KEEPING CURRENT
835
editors review more than 100 journals for relevant and valid information. This database also includes calculators and decision support tools. Valid and relevant evidence are synopsized, which makes the information easier to locate. The information is accessible on the Internet as well as in desktop and handheld computer versions. The aim of InfoRetriever is to provide ‘‘just-in-time’’ information that can be retrieved by clinicians while they are practicing, which provides the opportunity to incorporate evidence into practice [44]. In summary, the information needs of clinicians are many and varied; yet, these information needs remain largely unmet [45]. This likely diminishes the quality of patient care. Although it is important that physicians be proficient in evaluating the medical literature critically, it is more important that they become proficient in the ‘‘applied science of information management’’ [16]. Clinicians must learn the techniques and skills to focus on finding, evaluating, and using relevant and valid information at the point of care. Clinicians also need sources for rapid retrieval of this information to integrate it into their daily practice and their careers of lifelong learning. Selected resources Information databases InfoRetriever: http://www.infopoems.com/ (subscription required) The Cochrane Collaboration: http://www.cochrane.org/index.htm (subscription required) Database of Abstracts of Reviews of Effectiveness (DARE): http://www. york.ac.uk/inst/crd/crddatabases.htm Clinical Evidence: http://www.clinicalevidence.com/ (subscription required) Agency for Healthcare Research and Quality (AHRQ): http://www.ahrq. gov/clinic/ Bandolier on the Web: http://www.jr2.ox.ac.uk/bandolier/index.html DynaMed: http://www.dynamicmedical.com/ (subscription required) Educational resources Evidence-Based Medicine; Michigan State University: http://www.poems. msu.edu/EBM/ The evidence-based medicine toolkit: http://www.med.ualberta.ca/ebm/ ebm.htm Centre for Evidence-Based Medicine; Oxford-Centre for Evidence Based Medicine: http://www.cebm.net/ References [1] Ebbert JO, Dupras DM, Erwin PJ. Searching the medical literature using PubMed: a tutorial. Mayo Clin Proc 2003;78:87–91.
836
MCCONAGHY
[2] Grandage KK, Slawson DC, Shaughnessy AF. When less is more: a practical approach to searching for evidence-based answers. J Med Libr Assoc 2002;90(3):298–304. [3] Thompson ML. Characteristics of information resources preferred by primary care physicians. Bull Med Libr Assoc 1997;85(2):187–92. [4] Covell DG, Uman GC, Manning PR. Information needs in practice: are they being met? Ann Intern Med 1985;103(4):596–9. [5] Osheroff JA, Forsythe DE, Buchanan BG, et al. Physicians’ information needs: analysis of questions posed during clinical teaching. Ann Intern Med 1991;114(7):576–81. [6] Ely JW, Osheroff JA, Ebell MH, et al. Analysis of questions asked by family doctors regarding patient care. BMJ 1999;319(7206):358–61. [7] Ramos K, Linscheid R, Schafer S. Real-time information-seeking behavior of residency physicians. Fam Med 2003;35(4):257–60. [8] Slawson DC, Shaughnessy AF, Barry H. Which should come first: rigor or relevance? J Fam Pract 2001;50(3):209–10. [9] Ebell MH, Barry HC, Slawson DC, et al. Finding POEMs in the medical literature. J Fam Pract 1999;48:350–5. [10] McColl A, Smith H, White P, et al. General practitioner’s perceptions of the route to evidence based medicine: a questionnaire survey. BMJ 1998;316:361–5. [11] Young JM, Ward JE. Evidence-based medicine in general practice: beliefs and barriers among Australian GPs. J Eval Clin Pract 2001;7:201–10. [12] Ely JW, Osheroff JA, Ebell MH, et al. Obstacles to answering doctors’ questions about patient care with evidence: qualitative study. BMJ 2002;324:1–7. [13] McAlister FA, Graham I, Karr GW, et al. Evidence-based medicine and the practicing clinician. J Gen Intern Med 1999;14:236–42. [14] Putnam W, Twohig PL, Burge FI, et al. A qualitative study of evidence in primary care: what the practitioners are saying. CMAJ 2002;166:1525–30. [15] Coiera E. Information economics and the Internet. J Am Med Inform Assoc 2000;7:215–21. [16] Slawson DC, Shaughnessy AF. Teaching evidence-based medicine: should we be teaching information management instead? Acad Med 2005;80:685–9. [17] Stinson ER, Mueller DA. Survey of health professionals information habits and needs conducted through personal interviews. JAMA 1980;243:140–3. [18] Gorman PN, Ash J, Wykoff L. Can primary care physicians’ questions be answered using the medical journal literature? Bull Med Libr Assoc 1994;82(2):140–6. [19] Slawson DC, Shaughnessy AF, Bennett JH. Becoming a medical information master: feeling good about not knowing everything. J Fam Pract 1994;38(5):505–13. [20] Shaughnessy AF, Slawson DC, Bennett JH. Becoming an information master: a guidebook to the medical information jungle. J Fam Pract 1994;39(5):489–99. [21] Lindberg DAB. The National Library of Medicine and its role. Bull Med Libr Assoc 1993; 81:71–3. [22] Lowe HJ, Barnett GO. Understanding and using the medical subject headings (MeSH) vocabulary to perform literature searches. JAMA 1994;271:1103–8. [23] Greenhalgh T. How to read a paper: the Medline database. BMJ 1997;315:180–3. [24] Mendis K, Solangaarachchi I. PubMed perspective of family medicine research: where does it stand? Fam Pract 2005;22:570–5. [25] Haynes RB, Wilczynski N, McKibbon KA, et al. Developing optimal search strategies for detecting clinically sound studies in MEDLINE. J Am Med Inform Assoc 1994;1(6): 447–58. [26] Jacobson LD, Edwards AG, Granier SK, et al. Evidence-based medicine and general practice. Br J Gen Pract 1997;47:449–52. [27] Coomarasamy A, Gee H, Publicover M, et al. Medical journals and effective dissemination of health research. Health Info Libr J 2001;18:183–91. [28] Alper BS, Hand JA, Elliott SG, et al. How much effort is needed to keep up with the literature relevant for primary care? J Med Libr Assoc 2004;92(4):429–37.
KEEPING CURRENT
837
[29] Harbourt AM, Knecht LS, Humphreys BL. Structured abstracts in MEDLINE, 1989–1991. Bull Med Libr Assoc 1995;83:190–5. [30] Barry HC, Ebell MH, Shaughnessy AF, et al. Family physicians’ use of medical abstracts to guide decision making: style or substance? J Am Board Fam Pract 2001;14:437–42. [31] Cranney M, Walley T. Same information, different decisions: the influence of evidence on the management of hypertension in the elderly. Br J Gen Pract 1996;46:661–3. [32] Naylor CD, Chen E, Strauss B. Measured enthusiasm: does the method of reporting trial results alter perceptions of therapeutic effectiveness? Ann Intern Med 1992;117:916–21. [33] Silagy CA. An analysis of review articles published in primary care journals. Fam Pract 1993; 10(3):337–41. [34] Siwek J, Gourlay ML, Slawson DC, et al. How to write an evidence-based clinical review article. Am Fam Physician 2002;65:251–8. [35] Ebell MH, Siwek J, Weiss BD, et al. Strength of Recommendation Taxonomy (SORT): a patient-centered approach to grading evidence in the medical literature. Am Fam Physician 2004;69:548–56. [36] Ghosh A, Robbins K, Kelly J. The Cochrane Library: a resource for current reviews of clinical evidence. Minn Med 2000;83(7):43–5. [37] Volmink J, Siegfried N, Robertson K, et al. Research synthesis and dissemination as a bridge to knowledge management: the Cochrane Collaboration. Bull World Health Organ 2004;82: 778–83. [38] Haines A, Kuruvilla S, Borchert M. Bridging the implementation gap between knowledge and action for health. Bull World Health Organ 2004;82:724–32. [39] The Cochrane Collaboration. The Cochrane Manual Issue 4, 2006 [updated 23 August 2006]. Available at: http://www.cochrane.org/admin/manual.htm. Accessed September 20, 2006. [40] The Cochrane Collaboration. The Cochrane Library Issue 4, 2006 [updated 23 August 2006]. Available at: http://cochrane.org/reviews/index.htm. Accessed September 20, 2006. [41] Centre for Reviews and Dissemination. National Health Service. York, UK: University of York. Available at: http://www.york.ac.uk/inst/crd/crddatabases.htm#DARE. Accessed September 20, 2006. [42] Clinical evidence. BMJ Publishing Group. Available at: http://www.clinicalevidence.com. Accessed September 20, 2006. [43] InfoPOEMs-The Clinical Awareness System. Available at: http://www.infopoems.com/. Accessed September 20, 2006. [44] Chueh H, Barnett GO. ‘‘Just-in-time’’ clinical information. Acad Med 1997;72(6):512–7. [45] Smith R. What clinical information do doctors need? BMJ 1996;313:1062–8.
Prim Care Clin Office Pract 33 (2006) 839–862
Finding Truth from the Medical Literature: How to Critically Evaluate an Article William F. Miser, MD, MA Department of Family Medicine, The Ohio State University College of Medicine, 2231 North High Street, Room 203, Columbus, OH 43201, USA
With Internet access available to all, patients are increasingly gaining access to medical information, and then looking to their primary care physician for its interpretation. Gone are the days when what the physician says goes unchallenged by a patient. Our society is inundated with medical advice and contrary views from the newspaper, radio, television, popular lay journals, and the Internet, and physicians are faced with the task of ‘‘damage control.’’ Patients are searching for answers even before they come to the office, and are bringing with them articles they have downloaded from the Internet for interpretation. Primary care physicians also encounter an ‘‘information jungle’’ when it comes to the medical literature [1,2]. The amount of information available can be overwhelming [3]. There were 682,121 articles recorded in Pub MED in 2005. If clinicians, trying to keep up with the medical literature, were to read two articles per day, in just 1 year they would be over nine centuries behind in their reading! Despite the volume of medical literature, fewer than 15% of all articles published on a particular topic are useful for clinical practice [4]. Most articles are not peer-reviewed, are sponsored by those with commercial interests, or arrive free in the mail (the so-called ‘‘throwaways’’). Even articles published in the most prestigious journals are far from perfect. Analyses of clinical trials published in a wide variety of journals have identified large deficiencies in design, analysis, and reporting; although improving over time, the average quality score of clinical trials over the past 2 decades is less than 50% [5–7]. This has resulted in diagnostic tests and therapies becoming established as a routine part of practice before being rigorously
E-mail address:
[email protected] 0095-4543/06/$ - see front matter Ó 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.pop.2006.09.012 primarycare.theclinics.com
840
MISER
evaluated; which has led to the widespread use of tests with uncertain efficacy, and treatments that are either ineffective or that may do more harm than good [8]. A good recent example is the widespread use of hormonal replacement therapy to prevent cardiovascular disease, dementia, and other chronic diseases; the Women’s Health Initiative studies showed that this practice did more harm than good [9]. Although several excellent services are available to physicians that sift through and critically assess the medical literature, they are not helpful when a patient brings in the latest article that is ‘‘hot off the presses.’’ Thus, physicians must have basic skills in judging the validity and clinical importance of these articles. The two major types of articles (Fig. 1) found in the medical literature are those that (1) report original research (analytic, primary studies), and (2) those that summarize or draw conclusions from original research (integrative, secondary studies). Primary studies can be either experimental (an intervention is made) or observational (no intervention is made). This article provides an overview of a systematic, efficient, and effective approach to the critical review of original research. This information is pertinent to physicians no matter the clinical setting. Because of space limitations, this article cannot cover everything in exhaustive detail, and the reader is encouraged to refer to the suggested readings in Appendix 1 for further assistance.
Medical Literature
Primary (Analytic) Studies those that report original research
Secondary (Integrative) Studies those that draw conclusions from original research meta-analysis systematic review non-systematic review editorial, commentary practice guideline decision analysis economic analysis
Experimental
Observational
an intervention is made or variables are manipulated
no intervention is made and no variables are manipulated
experiment randomized controlled trial non-randomized controlled trial
cohort case-control cross-sectional descriptive, surveys case reports
Fig. 1. The major types of studies found in the medical literature.
CRITICAL EVALUATION OF MEDICAL LITERATURE
841
Critical assessment of an original research article It is important for clinicians to master the ability to critically assess an original research article if they are to apply ‘‘evidence-based medicine’’ to the daily clinical problems they encounter. Most busy clinicians, however, do not have the hours required to fully critique an article; they need a brief and efficient screening method that allows them to know if the information is valid and applicable to their practice. By applying the techniques offered here, one can approach the literature confidently and base clinical decisions on ‘‘evidence rather than hope’’ [10]. This approach is modified and adapted from several excellent sources. The Department of Clinical Epidemiology and Biostatistics at McMaster University in Hamilton, Ontario, Canada in 1981 published a series of useful guides to help the busy clinician critically read clinical articles about diagnosis, prognosis, etiology, and therapy [11–15]. These guides have subsequently been updated and expanded to focus more on the practical issues of first finding pertinent articles and then validating (believing) and applying the information to patient care (see Appendix 1) [10]. The recommendations from these users’ guides form the foundation upon which techniques developed by Slawson and colleagues are modified and added [1,2]. With an article in hand, the process involves three steps: (1) conduct an initial validity and relevance screen, (2) determine the intent of the article, and (3) evaluate the validity of the article based on its intent. Step one: conduct an initial validity and relevance screen The first step when looking at an article is to ask, ‘‘Is this article worth taking the time to review in depth?’’ This can be answered within a few seconds by asking six simple questions (Appendix 2). A ‘‘stop’’ or ‘‘pause’’ answer to any of these questions should prompt one to seriously consider whether time should be spent to critically assess the article. Is the article from a peer-reviewed journal? Most national and specialty journals published in the United States are peer-reviewed; if in doubt, this answer can be found in the journal’s ‘‘Instructions for Authors’’ section. Typically, journals sent to clinicians unsolicited and free of charge are known as ‘‘throwaway’’ journals. These journals, although attractive in appearance, are not peer-reviewed, but instead are often geared toward generating income from advertising, and consist of ‘‘expert opinions’’ [3,10]. Articles published in the major peer-reviewed journals have already undergone an extensive process to sift out flawed studies and to improve the quality of the ones subsequently accepted for publication. When an investigator submits a manuscript to a peer-reviewed journal, the editor first establishes whether the manuscript is suitable for that journal, and then, if
842
MISER
acceptable, sends it to several reviewers for assessment. Peer reviewers are not part of the editorial staff, but usually are volunteers who have expertise in both the subject matter and research design. This peer review process acts as a sieve by detecting those studies that are flawed by poor design, are trivial, or are uninterpretable. This process, along with subsequent revisions and editing, improves the quality of the paper and its statistical analyses [16–19]. The Annals of Internal Medicine, for example, receives more than 1200 original research manuscript submissions each year. The editorial staff reject half after an internal review, and the remaining half are sent to at least two peers for review. Of the original 1200 submissions, only 15% are subsequently published [20]. Because of these strengths, peer review has become the accepted method for improving the quality of the science reported in the medical literature [21]; however, this mechanism is far from perfect, and it does not guarantee that the published article is without flaw or bias [4]. Publication biases are inherent in the process, despite an adequate peer review process. Studies showing statistically significant (‘‘positive’’) results and having larger sample sizes are more likely to be written and submitted by authors, and subsequently accepted and published, than are nonsignificant (‘‘negative’’) studies [22–25]. Also, the speed of publication depends on the direction and strength of the trial results; trials with negative results may take twice as long to be published as do positive trials [26]. Finally, no matter how good the peer review system, fraudulent research, although rare, is extremely hard to identify [27]. Is the location of the study similar to mine, so that the results, if valid, would apply to my practice? This question can be answered by reviewing information about the authors on the first page of an article (typically at the bottom of the page). If one is in a rural general practice and the study was performed in a university subspecialty clinic, one may want to pause and consider the potential biases that may be present. This is a ‘‘soft’’ area, and rarely will one want to reject an article outright at this juncture; however, large differences in types of populations should raise caution in accepting the final results. Is the study sponsored by an organization that may influence the study design or results? This question considers the potential bias that may occur from outside funding. In most journals, investigators are required to identify sources of funding for their study. Clinicians need to be wary of published symposiums sponsored by pharmaceutical companies. Although found in peer-reviewed journals, they tend to be promotional in nature, to have misleading titles, to use brand names, and are less likely to be peer-reviewed in the same manner as other articles in the parent journal [28]. Also, randomized clinical trials
CRITICAL EVALUATION OF MEDICAL LITERATURE
843
(RCTs) published in journal supplements are generally of inferior quality compared with articles published in the parent journal [29]. This is not to say that all studies sponsored by commercial interests are biased; on the contrary, numerous well-designed studies published in the literature are sponsored by the pharmaceutical industry. If, however, a pharmaceutical company or other commercial organization funded the study, look for assurances from investigators that this association did not influence the design and results. The answers to the next three questions deal with clinical relevance to one’s practice, and can be obtained by reading the conclusion and selected portions of the abstract. Clinical relevance is important to not only physicians, but to patients. Rarely is it worthwhile to read an article about an uncommon condition one never encounters in practice, or about a treatment or diagnostic test that is not, and never will be, available because of cost or patient preference. Reading these types of articles may satisfy one’s intellectual curiosity, but will not impact significantly on the practice. Slawson and colleagues [1,30] have emphasized that for a busy clinician, articles concerned with ‘‘patient-oriented-evidence-that-matters’’ (POEMs) are far more useful than those articles that report ‘‘disease-oriented-evidence’’ (DOE). So, given a choice between reading an article that describes the sensitivity and specificity of a screening test in detecting cancer (a DOE) and one that shows that those undergo this screening enjoy an improved quality and length of life (a POEM), one would probably want to choose the latter. Will this information, if true, have a direct impact on the health of my patients, and is it something they will care about? Typically the abstract will contain this information. Outcomes such as quality of life, overall mortality, and cost are ones that physicians and patients often consider important. Is the problem addressed one that is common to my practice, and is the intervention or test feasible and available to me? Problems addressed should be something commonly encountered in practice, tests should be feasible, and therapy should be easily available. Will this information, if true, require me to change my current practice? If one’s practice already includes this diagnostic test or therapeutic intervention, this article reinforces what is being done; if not, however, then time should be spent on determining whether or not the results are valid before making any changes. In only a few seconds, one can quickly answer six pertinent questions that allow one to decide if more time is needed to critically assess the article. This ‘‘weeding’’ tool allows one to discard those articles that are not relevant to practice, thus allowing more time to examine the validity of those few articles that may have a direct impact on the care of one’s patients.
844
MISER
Step two: determine the intent of the article If the physician decides to continue with the article after completing step one, the next task is to determine why the study was performed, and what clinical questions the investigators were addressing [31]. The four major clinical categories found in articles of primary (original) research are: (1) therapy, (2) diagnosis and screening, (3) causation, and (4) prognosis (Table 1). The answer to this step can usually be found by reading the abstract, and if needed, by skimming the introduction (usually found in the last paragraph), to determine the purpose of the study. Step three: evaluate the validity of the article based on its intent After an article has successfully passed the first two steps, it is now time to critically assess its validity and applicability to one’s practice setting. Each of the four clinical categories found in Table 1 has a preferred study design and critical items to ensure its validity. The users’ guides published by the Department of Clinical Epidemiology and Biostatistics at McMaster University provide a useful list of questions to help you with this assessment. Modifications of these lists of questions are found in Appendices 3–6. To get started on this step, read the entire abstract, survey the boldface headings, review the tables, graphs, and illustrations, and then skim-read the first sentence of each paragraph to quickly grasp the organization of Table 1 Major clinical categories of primary research and preferred study designs Clinical category
Preferred study design
TherapydTests the effectiveness of a treatment such as a drug, surgical procedure, or other intervention
Randomized, double-blinded, placebocontrolled trial (see Fig. 2)
Diagnosis and screeningdMeasures the validity (Is it dependable?) and reliability (Will the same results be obtained every time?) of a diagnostic test, or evaluates the effectiveness of a test in detecting disease at a presymptomatic stage when applied to a large population
Cross-sectional survey (comparing the new test with a ‘‘gold standard’’) (Fig. 3)
CausationdDetermines whether an agent is related to the development of an illness
Cohort or case-control study, depending on how the rarity of disease; case reports may also provide crucial information (Figs. 4, 5)
PrognosisdDetermines what is likely to happen to someone whose disease is detected at an early stage.
Longitudinal cohort study (see Fig. 4)
Adapted from Greenhalgh T. How to read a paperdgetting your bearings (deciding what the paper is about). BMJ 1997;315:243–6; with permission.
845
CRITICAL EVALUATION OF MEDICAL LITERATURE
the article. One then needs to focus on the methods section, answering a specific list of questions based on the intent of the article. Is the study a randomized controlled trial? Randomized controlled trials (RCTs) (Fig. 2) are considered the ‘‘gold standard’’ design to determine the effectiveness of treatment. The power of RCTs lies in their use of randomization. At the start of a trial, participants are randomly allocated by a process equivalent to the flip of a coin to either one intervention (eg, a new diabetic medication) or another (eg, an established diabetic medication or placebo). Both groups are then followed for a specified period, and defined outcomes (eg, glucose control, quality of life, death) are measured and analyzed at the conclusion. Randomization diminishes the potential for investigators selecting individuals in a way that would unfairly bias one treatment group over another (selection bias). It is important to determine how the investigators actually
Is the sample similar to your population? How was the sample selected?
The Population
The Sample
Randomization • How were the groups randomized? • Did the investigator(s) account for those who were eligible but were not randomized or entered into the study? • Are the study and control groups similar? • Were the investigator(s) and subjects “blinded” to which group they were assigned?
Study Group
Control Group
Outcome
Outcome
• Were both groups treated exactly the same (except for the actual treatment)? • Was follow-up complete? Was everyone accounted for, including those who dropped out of the study? • Are the outcome(s) clearly defined? • Were subjects analyzed in the groups to which they were randomized (“intention to treat” analysis)?
Fig. 2. The randomized controlled trial, considered the ‘‘gold standard’’ for studies dealing with treatment or other interventions.
846
MISER
The Population The Sample Condition Present Risk Factor Present
Condition Present Risk Factor Absent
Condition Absent Risk Factor Present
Condition Absent Risk Factor Absent
Fig. 3. The cross-sectional (prevalence) study. This design is most often used in studies on diagnostic or screening tests.
performed the randomization. Although infrequently reported in the past, most journals now require a standard format that provides this information [6]. Various techniques can be used for randomization [32]. Investigators may use simple randomization; each participant has an equal chance of being assigned to one group or another, without regard to previous assignments of other participants. Sometimes this type of randomization will result in one treatment group being larger than another, or by chance, one group having important baseline differences that may affect the study. To avoid these problems, investigators may use blocked randomization (groups are equal in size) or stratified randomization (subjects are randomized within groups based on potential confounding factors such as age or gender). To determine the assignment of participants, investigators should use a table of random numbers or a computer that produces a random sequence. The final allocation of participants to the study should be concealed from both investigators and participants. If investigators responsible for assigning subjects are aware of the allocation, they may unwittingly (or otherwise) assign those who have a better prognosis to the treatment group and those who have a worse prognosis to the control group. RCTs that have inadequate allocation concealment will yield an inflated treatment effect that is up to 30% better than those trials with proper concealment [33,34]. Are the subjects in the study similar to mine? To be generalizable (external validity), the subjects in the study should be similar to the patients in one’s practice. A common problem encountered by
847
CRITICAL EVALUATION OF MEDICAL LITERATURE
Prospective Cohort Study The Population - Present
The Sample - Future Disease (a)
Risk Factor Present
Disease (c)
Risk Factor Absent
No Disease (b) No Disease (d)
Retrospective Cohort Study The Population - Past The Sample - Present Disease (a)
Risk Factor Present
Disease (c)
Risk Factor Absent
No Disease (b) No Disease (d)
Relative Risk (RR) is the risk of disease associated with a particular exposure. Condition Present Risk Factor Present Risk Factor Absent
Condition Absent
a
b
c
d
(a)/(a+b) RR =
(c)/(c+d)
Fig. 4. Prospective and retrospective cohort study. These types of studies are often used for determining causation or prognosis. Data are typically analyzed using relative risk.
primary care physicians is interpreting the results of studies done on patients in subspecialty care clinics. For example, the group of men participating in a study on early detection of prostate cancer at a university urology practice may be different from the group of men seen in a typical primary care office. It is important to determine who was included and who was excluded from the study. Are all participants who entered the trial properly accounted for at its conclusion? Another strength of RCTs is that participants are followed prospectively; however, it is important that these participants be accounted for at the end of the trial to avoid a ‘‘loss-of-subjects bias,’’ which can occur through the
848
MISER
Population with Disease (cases)
Risk Factor Not Exposed
Exposed
a
Sample of Cases With Disease
c
Population without Disease (controls) Risk Factor Exposed
Not Exposed
b
Sample of Controls Without Disease
d
Odds Ratio (OR) is the measure of strength of association. It is the odds of exposure among cases to the odds of exposure among the controls Cases
Exposed
a
Controls b
(a/a+c)/(c/a+c) OR =
Not Exposed
c
d
(b/b+d)/(d/b+d)
=
a/c b/d
=
ad bc
Fig. 5. The case-control study, a retrospective study in which the investigator selects a group with disease (cases) and one without disease (controls) and looks back in time at exposure to potential risk factors to determine causation. Data are typically analyzed using the odds ratio.
course of a prospective study as subjects drop out of the investigation for various reasons. Subjects may lose interest, move out of the area, develop intolerable side effects, or die. The subjects who are lost to follow-up may be different from those who remain in the study to the end, and the groups studied may have different rates of dropouts. An attrition rate of greater than 10% for short-term trials and 15% for long-term trials may invalidate the results of the study. At the conclusion of the study, subjects should be analyzed in the group in which they were originally randomized, even if they were noncompliant or switched groups (intention-to-treat analysis). For example, a study wishes to determine the best treatment approach to carotid stenosis, and patients are randomized to either carotid endarterectomy or medical management. Because it would be unethical to perform ‘‘sham’’ surgery, investigators and patients cannot be blinded to their treatment group. If, during the initial evaluation, individuals randomized to endarterectomy were found to be
CRITICAL EVALUATION OF MEDICAL LITERATURE
849
poor surgical candidates, they may instead be treated medically; however, at the conclusion of the study, their outcomes (stroke, death) should be included in the surgical group, even if they didn’t have surgerydto do otherwise would unfairly inflate the benefit of the surgical approach. Most journals now require a specific format for reporting RCTs, which includes a chart that allows you to easily follow the flow of subjects through the study [6]. Was everyone involved in the study (subjects and investigators) ‘‘blind’’ to treatment? Investigator bias may occur when those making the observations may unintentionally ‘‘shade’’ the results to confirm the hypothesis or to influence the subjects. The process of masking, in which neither the investigators nor the subjects are aware of group assignment (ie, double-blinding), prevents this bias. For example, in a study comparing a new diabetic medication to a placebo, neither the investigators nor the subjects should be aware of what the subjects are taking. The study medication should be indistinguishable from the comparison medication or placebo; it should have the same look and taste and be taken at the same frequency. If the study medication has a certain bitter taste or other side effect, and the comparison medication does not, subjects may be able to guess what medicine they are on, which may then influence how they perceive their improvement. Were the intervention and control groups similar at the start of the trial? Through the process of randomization, one would anticipate the groups to be similar at the beginning of a trial. Because this may not always be the case, investigators should provide a group comparison. This information is usually found in the first table of the article. Typically, comparisons will be made for demographic factors, other known risk factors, and disease severity. If differences exist between groups, one must use clinical experience and judgment to determine if small differences are likely to influence outcomes. Were the groups treated equally (aside from the experimental intervention)? To ensure both proper blinding and that other unknown determinants are not a factor, groups should be treated equally except for the therapeutic intervention. Everyone should be seen with the same frequency, and interventions should be similar. One should look for assurances that the groups were treated equally except for the experimental intervention. Are the results clinically as well as statistically significant? Statistics are mathematical techniques of gathering, organizing, describing, analyzing, and interpreting numerical data [35]. By their use,
850
MISER
investigators try to convince readers that the results of their study are valid. Internal validity addresses how well the study was done, and if the results reflect truth and did not occur by chance alone. External validity considers whether the results are generalizable to patients outside of the study. Both types of validity are important. The choice of statistical test depends on the study design, the types of data analyzed, and whether the groups are ‘‘independent’’ or ‘‘paired.’’ The three main types of data are categorical (nominal), ordinal, and continuous (interval). An observation made on more than one individual or group is ‘‘independent’’ (eg, measuring serum cholesterol in two groups of subjects), whereas making more than one observation on an individual is ‘‘paired’’ (eg, measuring serum cholesterol in an individual before and after treatment). Based on this information, one can then select an appropriate statistical test (Table 2). Be suspicious of a study that has a standard set of data collected in a standard way but is analyzed by a test that has an unpronounceable name and is not listed in a standard statistical textbook; the investigators may be attempting to prove something statistically significant that truly has no significance [36]. There are two types of errors that can potentially occur when comparing the results of a study to ‘‘reality.’’ A Type I error occurs when the study finds a difference between groups when in reality, there is no difference. This type of error is similar to a jury finding an innocent person guilty of a crime. The investigators usually indicate the maximum acceptable risk (the ‘‘alpha level’’) they are willing to tolerate in reaching this false-positive conclusion. Usually, the alpha level is arbitrarily set at 0.05 (or lower), which means the investigators are willing to take a 5% risk that any differences found were due to chance. At the completion of the study, the investigators then calculate the probability (known as the ‘‘P value’’) that a Type I error has occurred. When the P value is less than the alpha value (eg, !0.05), the investigators conclude that the results are ‘‘statistically significant.’’ Statistical significance does not always correlate with clinical significance. In a large study, very small differences can be statistically significant. For example, a study comparing two antihypertensives in over 1000 subjects may find a ‘‘statistically significant’’ difference in mean blood pressures of only 3 mmHg, which in the clinical realm is trivial. A P value of less than 0.0001 is no more clinically significant than a value of less than 0.05. The smaller P value only means there is less risk of drawing a false-positive conclusion (less than 1 in 1000). When analyzing an article, beware of being seduced by statistical significance in lieu of clinical significance; both must be considered. Instead of using P values, investigators are increasingly using confidence intervals (CI) to determine the significance of a difference. The problem with P values are they convey no information about the size of differences or associations found in the study [37]. Also, P values provide a dichotomous answerdeither the results are ‘‘significant’’ or ‘‘not significant.’’ In contrast,
Table 2 A practical guide to commonly used statistical tests Types of data
Categorical, 2 samples
Continuous
Student’s t
Tests for association between paired observations McNemar’s
Categorical, R3 samples
Ordinal
Continuous
-
-
-
Chi-square (r r)
-
-
Kruskal-Wallis one-way analysis of variance (ANOVA) ANOVA
Spearman’sr Kendall’s Tau
-
Kendall’s Tau Spearman’sr ANOVA
Pearson correlation Linear regression Multiple regression
Cochran Q
Wilcoxon signed rank Friedman two-way ANOVA
Paired t
CRITICAL EVALUATION OF MEDICAL LITERATURE
Tests for association between two independent variables Categorical, Chi-square 2 samples Fisher’s exact Categorical, Chi-square (r r) R3 samples Ordinal Mann-Whitney U Wilcoxon rank-sum
The test chosen depends on study design, types of variables analyzed, and whether observations are independent or paired. Categorical (nominal) data can be grouped, but not ordered (eg., eye color, gender, race, religion, etc). Ordinal data can be grouped and ordered (eg, sense of well-being: excellent, very good, fair, poor). Continuous data have order and magnitude (eg, age, blood pressure, cholesterol, weight, etc).
851
852
MISER
the CIl provides a range that will, with high probability, contain the true value, and provides more information than P values alone [38–40]. The larger the sample size, the narrower and more precise is the CI. A standard method used is the 95% CI, which provides the boundaries in which we can be 95% certain that the true value falls within that range. For example, a randomized clinical trial demonstrates that 50% of patients treated with drug A are cured, compared with 45% of those treated with drug B. Statistical analysis of this 5% difference shows a P value of less than 0.001 and a 95% CI of 0% to 10%. The investigators conclude this is a statistically significant improvement based on the P value; however, as a reader, you decide that a potential range of 0% to 10% is not clinically significant based on the 95% CI. If a negative trial, was a power analysis done? A negative trial is one in which no differences were found using the intervention between the groups. A Type II error occurs when the study finds no difference between groups when, in reality, there is a difference [41]. This type of error is similar to a jury finding a criminal innocent of a crime. The odds of reaching a false-negative conclusion (known as ‘‘beta’’) is typically set at 0.20 (20% chance). The power of a test (1-beta) is the ability to find a difference when in reality one exists, and depends on: (1) the number of subjects in the study (the more subjects, the greater the power), and (2) the size of the difference (known as ‘‘effect size’’) between groups (the larger the difference, the greater the power). Typically, the effect size investigators choose depends on ethical, economic, and pragmatic issues, and can be categorized into small (10%–25%), medium (26%–50%), and large (O50%) [42]. When looking at the effect size chosen by the investigators, ask whether you consider this difference to be clinically meaningful. Before the start of a study, the investigators should do a ‘‘power analysis’’ to determine how many subjects should be included in the study. Unfortunately, this was often not done in the past. Only 32% of the RCTs with negative results published between 1975 and 1990 in JAMA, Lancet, and New England Journal of Medicine reported sample size calculations; on review, the vast majority of these trials had too few patients, which led to insufficient statistical power to detect a 25% or 50% difference [43]. Other studies have shown similar deficiencies in other journals and disciplines [5,19,44,45]. Whenever one reads an article reporting a negative result, ask whether the sample size was large enough to permit investigators to draw such a conclusion. If a power analysis was done, check to see if the study had the required number of subjects. If a power analysis was not done, view the conclusions with skepticismdit may be that the sample size was not large enough to detect a difference. Were there other factors that might have affected the outcome? At times, an outcome may be caused by factors other than the intervention. For example, the simple act of observation can affect an outcome
CRITICAL EVALUATION OF MEDICAL LITERATURE
853
(Hawthorne effect). This effect occurs when subjects change their normal behavior because they are aware of being observed. To minimize this effect, study groups should be observed equally. Also, randomization and sufficiently large sample size assure that both known and unknown determinants of an outcome are evenly distributed between groups. As one reads through an article, think about potential influences that could impact one group more than another, and thus affect the outcome. Are the treatment benefits worth the potential harms and costs? This final question forces one to consider the cost benefit and potential harm of the therapy. The number needed to treat (NNT) takes into consideration the likelihood of an outcome or side effect [46]. Generally, the less common a potential outcome (eg, death), the greater the number of patients that would require treatment to prevent one outcome. If sudden death is a potential risk of a medication used to treat a benign condition, one must question the actual benefit of that drug. If, based upon a critical review of an article, one decides to implement a new test or therapy, one must also make a commitment to monitor its benefits and risks to patients, and to scan the literature for future articles that may offer additional findings. Consistency of the results in one’s practice, as well as across multiple published studies, is one characteristic of the scientific process that leads to acceptance and implementation.
A final word With some practice and the use of the worksheets, one can quickly (within a few minutes) perform a critical assessment of an article. While performing this appraisal, it is important to keep in mind that few articles will be perfect. A critical assessment is rarely black and white, but often comes in shades of gray [47]. Only you can answer for yourself the exact shade of gray that you are willing to accept when deciding to apply the results of the study to your practice. By applying the knowledge, principles, and techniques presented in this section, however, you can more confidently recognize the various shades of gray, and reject those articles that are seriously flawed.
Appendix 1 Suggested readings on critical reading skills 1. Slawson DC, Shaughnessy AF, Bennett JH. Becoming a medical information master: feeling good about not knowing everything. J Fam Pract 1994;38:505–13. [A superb article that addresses the concepts of POEMs and DOEs.]
854
MISER
2. Shaughnessy AF, Slawson DC, Bennett JH. Becoming an information master: a guidebook to the medical information jungle. J Fam Pract 1994;39:489–99. [An excellent article that reviews how to manage one’s way through the medical information jungle without getting lost or eaten alive.] 3. Shaughnessy AF, Slawson DC: Getting the most from review articles: a guide for readers and writers. Am Fam Phys 1997; 55:2155–60. [Provides useful techniques on reading a review article.] Items 4–8 are from ‘‘How to read clinical journals,’’ original McMaster series from The Canadian Medical Association Journal. [Despite being published in 1981, this series still has some great information!] 4. Why to read them and how to start reading them critically. Can Med Assoc J 1981;124:555–58. 5. To learn about a diagnostic test. Can Med Assoc J 1981;124:703–10. 6. To learn the clinical course and prognosis of disease. Can Med Assoc J 1981;124:869–72. 7. To determine etiology or causation. Can Med Assoc J 1981;124: 985–90, 8. To distinguish useful from useless or even harmful therapy. Can Med Assoc J 1981;124:1156–62. Items 9–14 are from ‘‘How to keep up with the medical literature,’’ in Annals of Internal Medicine. [A good series on the approach to keeping up with the medical literature.] 9. Haynes RB, McKibbon KA, Fitzgerald D, et al. Why try to keep up and how to get started. Ann Intern Med 1986;105:149–53. 10. Haynes RB, McKibbon KA, Fitzgerald D, et al. Deciding which journals to read regularly. Ann Intern Med 1986;105:309–12. 11. Haynes RB, McKibbon KA, Fitzgerald D, et al. Expanding the number of journals you read regularly. Ann Intern Med 1986;105:474–8. 12. Haynes RB, McKibbon KA, Fitzgerald D, et al. Using the literature to solve clinical problems. Ann Intern Med 1986;105:636–40. 13. Haynes RB, McKibbon KA, Fitzgerald D, et al. Access by personal computer to the medical literature. Ann Intern Med 1986;105:810–6. 14. Haynes RB, McKibbon KA, Fitzgerald D, et al. How to store and retrieve articles worth keeping. Ann Intern Med 1986;105:978–84. Items 15–45 are from The McMaster’s seriesd‘‘User’s guide to the medical literature’’ in JAMA: The Journal of the American Medical Association. This material can now be found in an interactive format at http://pubs. ama-assn.org/misc/usersguides.dtl. [The ultimate series written from the perspective of a busy clinician who wants to provide effective medical care but is sharply restricted in time for reading.] 15. Oxman AD, Sackett DL, Guyatt GH. How to get started. JAMA 1993;270:2093–8. 16. Guyatt GH, Sackett DL, Cook DJ. How to use an article about therapy or prevention. A. Are the results of the study valid? JAMA 1993;270:2598–601
CRITICAL EVALUATION OF MEDICAL LITERATURE
855
17. Guyatt GH, Sackett DL, Cook DJ. How to use an article about therapy or prevention. B. What were the results and will they help me in caring for my patients? JAMA 1994;271:59–63. 18. Jaeschke R, Guyatt GH, Sackett DL. How to use an article about a diagnostic test. A. Are the results of the study valid? JAMA 1994;271:389–91. 19. Jaeschke R, Guyatt GH, Sackett DL. How to use an article about a diagnostic test. B. What are the results and will they help me in caring for my patients? JAMA 1994;271:703–07. 20. Levine M, Walter S, Lee H, et al. How to use an article about harm. JAMA 1994;271:1615–9. 21. Levine M, Walter S, Lee H, et al. How to use an article about prognosis. JAMA 1994;272:234–37. 22. Oxman AD, Cook DJ, Guyatt GH. How to use an overview. JAMA 1994; 272:1367–71. 23. Richardson WS, Detsky AS. How to use a clinical decision analysis. A. Are the results of the study valid? JAMA 1995;273:1292–5. 24. Richardson WS, Detsky AS. How to use a clinical decision analysis. B. What are the results and will they help me in caring for my patients? JAMA 1995;273:1610–23. 25. Hayward RS, Wilson MC, Tunis SR, et al. How to use clinical practice guidelines. A. Are the recommendations valid? 1995;274:570–4. 26. Wilson MC, Hayward RS, Tunis SR, et al. How to use clinical practice guidelines. B. What are the recommendations and will they help you in caring for your patients? JAMA 1995;274:1630–62. 27. Guyatt GH, Sackett DL, Sinclair JC. A method for grading health care recommendations. JAMA 1995;274:1800–4. 28. Naylor CD, Guyatt GH. How to use an article reporting variations in the outcomes of health services. JAMA 1996;275:554–8. 29. Naylor CD, Guyatt GH. How to use an article about a clinical utilization review. JAMA 1996;275:1435–9. 30. Guyatt GH, Naylor CD, Juniper E, et al. How to use articles about health-related quality of life. JAMA 1997;277:1232–7. 31. Drummond MF, Richardson WS, O’Brien BJ, et al. How to use an article on economic analysis of clinical practice. A. Are the results of the study valid? JAMA 1997;277:1552–7. 32. O’Brien BJ, Heyland D, Richardson WS, et al. How to use an article on economic analysis of clinical practice B. What are the results and will they help me in caring for my patients? JAMA 1997;277:1802–06. 33. Dans AL, Dans LF, Guyatt GH, et al. How to decide on the applicability of clinical trial results to your patients. JAMA 1998;279:545–9. 34. Richardson WS, WIlson MC, Guyatt GH, et al. How to use an article about disease probability for differential diagnosis. JAMA 1999;281:1214–9. 35. Guyatt GH, Sinclair J, Cook DJ, et al. How to use a treatment recommendation. JAMA 1999;281:1836–43.
856
MISER
36. Randolph AG, Haynes RB, Wyatt JC. How to use an article evaluating the clinical impact of a computer-based clinical decision support system. JAMA 1999;282:67–74. 37. Bucher HC, Guyatt GH, Cook DJ. Applying clinical trial results. A. How to use an article measuring the effect of an intervention on surrogate end points. JAMA 1999;282:771–8. 38. McAlister FA, Laupacis A, Wells GA, et al. Applying clinical trial results. B. Guidelines for determining whether a drug is exerting (more than) a class effect. JAMA 1999;282(9):1371–7. 39. Hunt DL, Jaeschke R, McKibbon KA. Using electronic health information resources in evidence-based practice. JAMA 2000;283:1875–9. 40. McAlister FA, Strauss SE, Guyatt GH, et al. Integrating research evidence with the care of the individual patient. JAMA 2000;282:2829–36. 41. McGinn TG, Guyatt GH, Wyer PC, et al. How to use articles about clinical decision rules. JAMA 2000;284:79–84. 42. Giacomini MK, Cook DJ. Qualitative research in health care. A. Are the results of the study valid? JAMA 2000;284:357–62. 43. Giacomini MK, Cook DJ. Qualitative research in health care. What are the results and will they help me in caring for my patients? JAMA 2000;284:478–82. 44. Richardson WS, WIlson MC, Williams JW, et al. How to use an article on the clinical manifestation of disease. JAMA 2000;284:869–75. 45. Guyatt GH, Haynes RB, Jaeschke RZ, et al. Evidence-based medicine: principles for applying the Users’ Guides to patient care. JAMA 2000;284:1290–6. Items 46–55 are from ‘‘How to read a paper’’ in the British Medical Journal. [A great series that compliments the User’s guide.] 46. Greenhalgh T. The MEDLINE database. Br Med J 1997;315(7101): 180–3. 47. Greenhalgh T. Getting your bearings (deciding what the paper is about). Br Med J 1997;315(7102):24–6. 48. Greenhalgh T. Assessing the methodological quality of published papers. Br Med J 1997;315(7103):305–8. 49. Greenhalgh T. Statistics for the non-statistician. Br Med J 1997;315(7104):364–6. 50. Greenhalgh T. Statistics for the non-statistician. II: ‘‘Significant’’ relations and their pitfalls. Br Med J 1997;315(7105):422–5. 51. Greenhalgh T. Papers that report drug trials. Br Med J 1997;315(7106):480–3. 52. Greenhalgh T. Papers that report diagnostic or screening tests. Br Med J 1997;315(7107):540–3. 53. Greenhalgh T. Papers that tell you what things cost (economic analyses). Br Med J 1997;315(7108):596–9. 54. Greenhalgh T. Papers that summarize other papers (systemic reviews and meta-analyses). Br Med J 1997;315(7109):672–5.
CRITICAL EVALUATION OF MEDICAL LITERATURE
857
55. Greenhalgh T. Papers that go beyond numbers (qualitative research). Br Med J 1997;315(7110):740–3. 56. Hulley SB, Cummings SR. Browner WS, Get al. Designing clinical researchdan epidemiologic approach. Baltimore (MD): Lippincott, Williams & Wilkins; 2000. [An excellent textbook on understanding research methods and statistics.] 57. Fletcher RH, Fletcher SW. Clinical epidemiology: the essentials. 4th edition. Baltimore (MD): Lippincott, Williams and Wilkins; 2005. [A basic textbook written for clinicians and organized by clinical questions: diagnosis, treatment, and so on.] 58. Haynes RB, Sackett DL, Guyatt GH, et al. Clinical epidemiology: how to do clinical practice research. 4th edition. Baltimore (MD): Lippincott, Williams and Wilkins; 2006. [A lively introduction to clinical epidemiology, with special emphasis on diagnosis and treatment, by leading proponents of ‘‘evidence-based medicine.’’] 59. Riegelman RK. Studying a study and testing a test: how to read the medical evidence. 4th edition. Baltimore (MD): Lippincott, Williams & Wilkins; 2000. [A clear description of an approach to studies of diagnosis and treatment.] 60. Gelbach SH. Interpreting the medical literature. 4th edition. New York: McGraw-Hill; 2002. [A basic introduction.]
Appendix 2 Step one in critically assessing an original research article Initial validity and relevance screen: is this article worth taking the time to review in depth? A ‘‘stop’’ or ‘‘pause’’ answer to any of the following should prompt one to seriously question whether one should spend the time to critically review the article.
1. Is the article from a peer-reviewed journal? Articles published in a peer-reviewed journal have already gone through an extensive review and editing process. 2. Is the location of the study similar to mine so the results, if valid, would apply to my practice? 3. Is the study sponsored by an organization that may influence the study design or results? Read the conclusion of the abstract to determine relevance. 4. Will this information, if true, have a direct impact on the health of my patients, and is it something they will care about? 5. Is the problem addressed one that is common to my practice, and is the intervention or test feasible and available to me?
Yes (go on)
No (stop)
Yes (go on)
No (stop)
Yes (pause)
No (stop)
Yes (go on)
No (stop)
Yes (go on)
No (stop)
858
MISER
6. Will this information, if true, require me to change my current practice?
Yes (go on)
No (stop)
Questions 4–6 adapted from Slawson D, Shaughnessy A, Ebell M, et al. Mastering medical information and the role of POEMsdPatient-Oriented Evidence that Matters. J Fam Pract 1997;45:195–6.
Appendix 3 Determining validity of an article about therapy If the article passes the initial screen in Appendix 2, proceed with the following critical assessment by reading the Methods section. A ‘‘stop’’ answer to any of the following should prompt one to seriously question whether the results of the study are valid and whether one should use this therapeutic intervention.
1. Is the study a randomized controlled trial? a. How were patients selected for the trial? b. Were they properly randomized into groups using concealed assignment? 2. Are the subjects in the study similar to mine? 3. Are all participants who entered the trial properly accounted for at its conclusion? a. Was follow-up complete and were few lost to follow-up compared with the number of bad outcomes? b. Were patients analyzed in the groups to which they were initially randomized (intention to treat analysis)? 4. Was everyone involved in the study (subjects and investigators) ‘‘blind’’ to treatment? 5. Were the intervention and control groups similar at the start of the trial? (Check Appendix 1) 6. Were the groups treated equally (aside from the experimental intervention)? 7. Are the results clinically as well as statistically significant? Were the outcomes measured clinically important? 8. If a negative trial, was a power analysis done? 9. Were there other factors that might have affected the outcome? 10. Are the treatment benefits worth the potential harms and costs?
Yes (go on)
No (stop)
Yes (go on) Yes (go on)
No (stop) No (stop)
Yes (go on)
No (stop)
Yes (go on)
No (stop)
Yes (go on)
No (stop)
Yes (go on)
No (stop)
Yes (go on) Yes (go on)
No (stop) No (stop)
Yes (go on)
No (stop)
Adapted from Slawson D, Shaughnessy A, Bennett J. Becoming a medical information master: feeling good about not knowing everything. J Fam Pract 1994;38:505–13, and Guyatt G, Sackett D, Cook D. User’s guides to the medical literature. II. How to use an article about therapy or prevention. A. Are the results of the study valid? The Evidence-Based Medicine Working Group. JAMA 1993;270:2598–601.
859
CRITICAL EVALUATION OF MEDICAL LITERATURE
Appendix 4 Determining validity of an article about a diagnostic test If the article passes the initial screen in Appendix 2, proceed with the following critical assessment by reading the Methods section. A ‘‘stop’’ answer to any of the following should prompt one to seriously question whether the results of the study are valid and whether one should use this diagnostic test. 1. What is the disease being addressed and what is the diagnostic test? _______________________________________________________________ Yes (go on) No (stop) 2. Was the new test compared with an acceptable ‘‘gold standard’’ test and were both tests applied in a uniformly blind manner? 3. Did the patient sample include an appropriate Yes (go on) No (stop) spectrum of patients to whom the diagnostic test will be applied in clinical practice? 4. Is the new test reasonable? What are its limitations? Explain: _________________________________________________________________ 5. In terms of prevalence of disease, are the study Yes (go on) No (stop) subjects similar to my patients? Varying prevalences will affect the predictive value of the test in my practice. 6. Will my patients be better off as a result of this test? Yes (go on) No (stop) 7. What are the sensitivity, specificity, and predictive values of the test?
Sensitivity = (a)/(a + c) = ______
“Gold standard” result
Specificity = (d)/(b + d) =______
Test result
Positive predictive value = (a)/(a + b) = ______
Positive
Negative predictive value = (c)/(c + d) = ______
Negative
Positive
Negative
a
b
c
d
Adapted from Slawson D, Shaughnessy A, Bennett J. Becoming a medical information master: feeling good about not knowing everything. J Fam Pract 1994;38:505–13, and Jaeschke R, Guyatt G, Sackett D. User’s guides to the medical literature. III. How to use an article about a diagnostic test. A. Are the results of the study valid? The Evidence-Based Medicine Working Group. JAMA 1994;271:389–91.
Appendix 5 Determining validity of an article about causation If the article passes the initial screen in Appendix 2, proceed with the following critical assessment by reading the Methods section. A ‘‘stop’’ answer to any of the following should prompt one to seriously question whether the
860
MISER
results of the study are valid and whether the item in question is really a causative factor. 1. Was there a clearly defined comparison group or those at risk for, or having, the outcome of interest? 2. Were the outcomes and exposures measured in the same way in the groups being compared? 3. Were the observers blinded to the exposure of outcome, and to the outcome? 4. Was follow-up sufficiently long and complete? 5. Is the temporal relationship correct? Does the exposure to the agent precede the outcome? 6. Is there a dose-response gradient? As the quantity or the duration of exposure to the agent increases, does the risk of outcome likewise increase? 7. How strong is the association between exposure and outcome? Is the relative risk (RR) or odds ratio (OR) large?
Yes (go on)
No (stop)
Yes (go on)
No (stop)
Yes (go on)
No (stop)
Yes (go on) Yes (go on)
No (stop) No (stop)
Yes (go on)
No (stop)
Yes (go on)
No (stop)
Adapted from Levine M, Walter S, Lee H, et al. User’s guides to the medical literature. IV. How to use an article about harm. The Evidence-Based Medicine Working Group. JAMA 1994;271:1615–9.
Appendix 6 Determining validity of an article about prognosis If the article passes the initial screen in Appendix 2, proceed with the following critical assessment by reading the Methods section. A ‘‘stop’’ answer to any of the following should prompt one to seriously question whether the results of the study are valid. 1. Was an ‘‘inception cohort’’ assembled? Did the investigators identify a specific group of people initially free of the outcome of interest, and follow them forward in time? 2. Were the criteria for entry into the study objective, reasonable and unbiased? 3. Was follow-up of subjects adequatedat least 70%–80%? 4. Were the patients similar to mine, in terms of age, sex, race, severity of disease, and other factors that might influence the course of the disease? 5. Where did the subjects come from? (was the referral pattern specified?) 6. Were outcomes assessed objectively and blindly?
Yes (go on)
No (stop)
Yes (go on)
No (stop)
Yes (go on)
No (stop)
Yes (go on)
No (stop)
Yes (go on)
No (stop)
Yes (go on)
No (stop)
Adapted from Slawson D, Shaughnessy A, Bennett J. Becoming a medical information master: feeling good about not knowing everything. J Fam Pract 1994;38:505–13, and Laupacis A, Wells G, Richardson W, et al. User’s guides to the medical literature. V. How to use an article about prognosis. The Evidence-Based Medicine Working Group. JAMA 1994;272:234–37.
CRITICAL EVALUATION OF MEDICAL LITERATURE
861
References [1] Slawson D, Shaughnessy A, Bennett J. Becoming a medical information master: feeling good about not knowing everything. J Fam Pract 1994;38:505–13. [2] Shaughnessy A, Slawson D, Bennett J. Becoming an information master: a guidebook to the medical information jungle. J Fam Pract 1994;39:489–99. [3] Fletcher R, Fletcher S. Keeping clinically up-to-date. Evidence-based approach to the medical literature. J Gen Intern Med 1997;12:S5–14. [4] Lock S. Does editorial peer review work? [editorial]. Ann Intern Med 1994;121:60–1. [5] Sonis J, Jones J. The quality of clinical trials published in The Journal of Family Practice, 1974–1991. J Fam Pract 1994;39:225–35. [6] Begg C, Cho M, Eastwood S, et al. Improving the quality of reporting of randomized controlled trials. The CONSORT statement. JAMA 1996;276:637–9. [7] Altman D. The scandal of poor medical research: we need less research, better research, and research done for the right reasons. BMJ 1994;308:283–4. [8] Reid M, Lachs M, Feinstein A. Use of methodological standards in diagnostic test research. Getting better but still not good. JAMA 1995;274:645–51. [9] Rossouw JE, Anderson GL, Prentice RL, et al. Risks and benefits of estrogen plus progestin in health postmenopausal women: principal results from the Women’s Health Initiative randomized controlled Trial. JAMA 2002;288(3):321–33. [10] Guyatt G, Rennie D. Users’ guides to the medical literature [editorial]. JAMA 1993;270: 2096–7. [11] Department of Clinical Epidemiology and Biostatistics. McMaster University. How to read clinical journals: I. Why to read them and how to start reading them critically. Can Med Assoc J 1981;124(5):555–8. [12] Department of Clinical Epidemiology and Biostatistics. McMaster University. How to read clinical journals: II. To learn about a diagnostic test. Can Med Assoc J 1981;124: 703–10. [13] Department of Clinical Epidemiology and Biostatistics. McMaster University. How to read clinical journals: III. To learn the clinical course and prognosis of disease. Can Med Assoc J 1981;124:869–72. [14] Department of Clinical Epidemiology and Biostatistics. McMaster University. How to read clinical journals: IV. To determine etiology or causation. Can Med Assoc J 1981;124:985–90. [15] Department of Clinical Epidemiology and Biostatistics. McMaster University. How to read clinical journals: V. To distinguish useful from useless or even harmful therapy. Can Med Assoc J 1981;124:1156–62. [16] Kassirer J, Campion E. Peer reviewdcrude and understudied, but indispensable. JAMA 1994;272:96–7. [17] Abby M, Massey M, Galandiuk S, Polk H. Peer review is an effective screening process to evaluate medical manuscripts. JAMA 1994;272:105–7. [18] Goodman S, Berlin J, Fletcher S, et al. Manuscript quality before and after peer review and editing at Annals of Internal Medicine. Ann Intern Med 1994;121:11–21. [19] Gardner M, Bond J. An exploratory study of statistical assessment of papers published in the British Medical Journal. JAMA 1990;263:1355–7. [20] Justice A, Berlin J, Fletcher S, et al. Do readers and peer reviewers agree on manuscript quality? JAMA 1994;272:117–9. [21] Colaianni L. Peer review in journals indexed in Index Medicus. JAMA 1994;272:156–8. [22] Dickersin K, Min Y, Meinert C. Factors influencing publication of research results. Follow-up of applications submitted to two institutional review boards. JAMA 1992; 267(3):374–8. [23] Jadad A, Rennie D. The randomized controlled trial gets a middle-aged checkup [editorial]. JAMA 1998;279:319–20.
862
MISER
[24] Rennie D, Flanagin A. Publication biasdthe triumph of hope over experience. JAMA 1992; 267:411–2. [25] Scherer R, Dickersin K, Langenberg P. Full publication of results initially presented in abstractsda meta-analysis. JAMA 1994;272:158–62. [26] Ioannidis J. Effect of the statistical significance of results on the time to completion and publication of randomized efficacy trials. JAMA 1998;279:281–6. [27] Whitely W, Rennie D, Hafner A. The scientific community’s response to evidence of fraudulent publication. The Robert Slutsky case. JAMA 1994;272:170–3. [28] Bero L, Galbraith A, Rennie D. The publication of sponsored symposiums in medical journals. N Engl J Med 1992;327:1135–40. [29] Rochon P, Gurwitz J, Cheung M, et al. Evaluating the quality of articles published in journal supplements compared with the quality of those published in the parent journal. JAMA 1994;272:108–13. [30] Slawson D, Shaughnessy A, Ebell M, et al. Mastering medical information and the role of POEMsdPatient-Oriented Evidence that Matters. J Fam Pract 1997;45:195–6. [31] Greenhalgh T. How to read a paperdgetting your bearings (deciding what the paper is about). BMJ 1997;315:243–6. [32] Franks P. Clinical trials. Fam Med 1988;20:443–8. [33] Schulz K, Chalmers I, Grimes D, et al. Assessing the quality of randomization from reports of controlled trials published in oObstetrics and gynecology journals. JAMA 1994;272:125–8. [34] Schulz K, Chalmers I, Hayes R, et al. Empirical evidence of bias. Dimensions of methodological quality associated with estimates of treatment effects in controlled trials. JAMA 1995; 273:408–12. [35] O’Brien P, Shampo M. Statistics for cliniciansd1. Descriptive statistics. Mayo Clin Proc 1981;56:47–9. [36] Greenhalgh T. How to read a paperdstatistics for the non-statistician. BMJ 1997;315:364–6. [37] Grimes D. The case for confidence intervals [editorial]. Obstet Gynecol 1992;80:865–6. [38] Simon R. Confidence intervals for reporting results of clinical trials. Ann Intern Med 1986; 105:429–35. [39] Braitman L. Confidence intervals assess both clinical significance and statistical significance. Ann Intern Med 1991;114:515–7. [40] Gehlbach S. Interpreting the medical literature. 3rd edition. New York: McGraw-Hill; 1993. [41] Detsky A, Sackett D. When was a ‘‘negative’’ clinical trial big enough? How many patients you needed depends on what you found. Arch Intern Med 1985;145:709–12. [42] Raju R, Langenberg P, Sen A, et al. How much ‘‘better’’ is good enough? The magnitude of treatment effect in clinical trials. Am J Dis Child 1992;146:407–11. [43] Moher D, Dulberg C, Wells G. Statistical power, sample size, and their reporting in randomized controlled trials. JAMA 1994;272:122–4. [44] Freiman J, Chalmers T, Smith H, et al. The importance of beta, the Type II error and sample size in the design and interpretation of the randomized control trial: survey of 71 ‘‘negative’’ trials. N Engl J Med 1978;299:690–4. [45] Mengel M, Davis A. The statistical power of family practice research. Fam Pract Res J 1993; 13:105–11. [46] Guyatt G, Sackett D, Cook D. Users’ guides to the medical literature. II. How to use an article about therapy or prevention? B. What were the results and will they help me in caring for my patients? The Evidence-Based Medicine Working Group. JAMA 1994;271:59–63. [47] Oxman A, Sackett D, Guyatt G. Users’ guides to the medical literature. I. How to get started. The Evidence-Based Medicine Working Group. JAMA 1993;270:2093–5.
Prim Care Clin Office Pract 33 (2006) 863–885
Coronary Artery Disease Screening, Treatment, and Follow-up Jennifer L. Junnila, MD, MPHa, Guy P. Runkle, MD, MAb,* a
Department of Medical Science, Army Medical Department Center and School, Fort Sam Houston, TX 78234, USA b Department of Family Medicine, Madigan Army Medical Center, Fort Lewis, WA 98431, USA
According to the National Center for Health Statistics [1] and the American Heart Association [2], heart disease affects more than 71 million Americans and causes more than 30% of the total deaths in this countrydmore than 900,000 deaths in 2003 alone. Advanced heart disease may exist before the onset of clinical symptoms; myocardial infarction or even cardiac death may be the first sign of serious disease. A recent prospective cohort study [3] demonstrated that one in five asymptomatic patients who have diabetes has silent myocardial ischemia. Providing medical therapy and other appropriate interventions for patients in the subclinical stages of disease may allow primary care physicians to improve the prognosis for patients at high risk for cardiac events. Using evidence-based screening methods in the early detection of asymptomatic coronary artery disease (CAD) would give physicians the ability to identify which patients that would benefit most from primary prevention efforts. The initial evaluation and management of patients presenting with acute coronary syndrome, including unstable angina and myocardial infarction (MI), is beyond the scope of this article. Instead, the authors discuss the screening of asymptomatic patients for suspected significant CAD, as well as the treatment and ongoing management of those who have documented CAD or a high suspicion of CAD. Heart disease in children and adolescents is not discussed.
* Corresponding author. E-mail address:
[email protected] (G.P. Runkle). 0095-4543/06/$ - see front matter Ó 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.pop.2006.09.011 primarycare.theclinics.com
864
JUNNILA & RUNKLE
Coronary artery disease screening in asymptomatic patients A provider considering a screening test for an asymptomatic patient should also ask if the screening test will result in further testing or treatment that positively affects outcomes. Depending on the patient’s point of view, improved outcomes may be in terms of either quality or quantity of life. Before performing a screening test, three variables should be considered: (1) the sensitivity of the test, (2) the specificity of the test, and (3) the pretest probability of disease. That is, how likely will true, significant disease actually be detected by the test? What is the probability that a negative test actually reflects true absence of disease? The pretest probability of any disease, or likelihood that disease is present, will greatly affect whether a screening test for that disease is indicated. This is no different in screening for CAD. The American College of Cardiology (ACC), American Heart Association (AHA) and the United States Preventive Services Task Force (USPSTF) recommend against screening for either the presence of severe coronary artery stenosis or for the prediction of CAD events in adults at low risk for these events [4–7]. The potential harms of routine screening for CAD in adults at low risk for disease exceed the potential benefits [6,7]. Tests in this population are generally false-positive, especially in women, and may result in unnecessary invasive and possibly injurious procedures and overtreatment. An appropriate algorithm for estimating risk of CAD should be considered before testing for CAD in the asymptomatic patient.
Estimating probability of significant coronary artery disease in the asymptomatic patient There is inadequate evidence to determine the precise balance of benefits (improved CAD-related health outcomes) and harms (including overtreatment and unnecessary invasive procedures) in screening for significant CAD [7]. The rationale for establishing the diagnosis of CAD in asymptomatic patients falls largely into two categories: (1) identification and further risk-stratification of the patient who has multiple risk factors and who may benefit from extensive primary prevention efforts, and (2) screening for critical disease in a patient at moderate or higher risk who plans to begin an exercise program. A person’s risk for CAD can be estimated based on the presence of multiple risk factors. These include older age, male gender, elevated systolic blood pressure or treatment for hypertension, smoking, elevated total cholesterol, low high-density lipoproteins (HDL), diabetes, obesity, and sedentary lifestyle. Estimating the probability of CAD greatly influences whether to screen for disease. The Diamond-Forrester method for estimating pretest probability of CAD was first described in 1979 [8,9]. This method considers a person’s age, gender, and the quality of cardiac symptoms; however, it is only useful in the symptomatic patient, and was not developed from an unselected
CORONARY ARTERY DISEASE
865
population. In the asymptomatic patient, determination of 10-year (shortterm) risk for developing CAD is performed using Framingham risk scoring [10]. The Adult Treatment Panel III Guidelines (ATP III) of the National Cholesterol Education Project recommends consideration of the Framingham Risk Score in its treatment guidelines [11]. The Framingham score estimates risk of major CAD events, including MI and coronary death (Fig. 1). Based on the Framingham Risk Index, the risk of CAD is estimated to be low (!0% risk of a major CAD event in the next 10 years), intermediate (O10% and !20%), or high (O20%) [10,11].
Fig. 1. Framingham risk scores for men. (Adapted from Wilson PW, D’Agostino R, Levy D, et al. Prediction of coronary heart disease using risk factor categories. Circulation 1998;97: 1837–47; with permission.)
866
JUNNILA & RUNKLE
Screening methods Typical screening tests for CAD include resting ECG, exercise treadmill test (ETT), cardiac stress imaging, and electron-beam computerized tomography (EBCT) scanning for coronary calcium. For adults at low risk for CAD event, the USPSTF recommends against screening for disease. The USPSTF does not have a recommendation for or against routine screening with ECG, ETT, or EBCT for either the presence of severe coronary artery stenosis or the prediction of CAD events in adults at increased risk for CAD events [6,7]. Cardiac stress imaging may be appropriate in select patients. It is important to recognize that a negative test does not rule out a future cardiac event in a patient at highest risk. These tests have limited sensitivity, and a false-negative test is more likely in a patient in the highest risk group. Techniques for performing these tests are not discussed in this article. The ‘‘routine’’ electrocardiogram in screening for coronary artery disease The resting 12-lead ECG provides valuable information about myocardial ischemia in symptomatic patients who have known CAD, and may assist in the evaluation of atypical chest pain; however, there is presently no evidence that the routine ambulatory ECG provides reliable information concerning ischemia in asymptomatic subjects who do not have known CAD. One study showed the approximately 30% of patients who have angiographicallyproven CAD have a normal resting ECG [12], and the ACC/AHA guidelines recommend against use of a routine ECG to screen asymptomatic patients [9]. Exercise treadmill testing No study has directly examined the effect on CAD outcomes following screening asymptomatic patients with exercise treadmill testing [13]. In one meta-analysis [14], the sensitivity of exercise treadmill testing ranged from 23% to 100%, and the specificity ranged from 17% to 100%. A cost-effectiveness study from the late 1980s [15] estimated that screening asymptomatic 60-year-old men who have no other risk factors has a cost per life-year saved of $44,332; for 60 year-old women, the cost was $47,606. Although exercise tolerance testing correctly identifies severe coronary artery obstruction in up to 2.7% of those screened, most positive findings will be false when the risk of coronary events is low [13]. Use of the ETT to screen for CAD in asymptomatic low-to-moderate risk individuals is not indicated [7,9]; however, asymptomatic men older than age 45 and women over age 55 who are at high risk for CAD based on estimates such as the Framingham Index may benefit from screening [9]. Cardiac stress imaging The use of cardiac stress imaging in the asymptomatic person is generally reserved for patients who have abnormal exercise ECG. The ACC/AHA
CORONARY ARTERY DISEASE
867
guidelines recommend that asymptomatic patients felt to be at low risk for CAD following exercise ECG should not have further stress imaging performed. A patient who has a moderate to high risk of CAD and abnormal exercise ECG may benefit from screening with exercise myocardial perfusion imaging. In a study of patients evaluated for the presence of CAD [16], stress echocardiography had better prognostic capabilities than stress electrocardiography. Asymptomatic patients at moderate to high risk who are unable to exercise can have imaging enhanced with direct vasodilators such as adenosine or dipyridamole. In patients who have limited exercise capacity and contraindications to direct vasodilators, myocardial perfusion imaging improves the sensitivity of dobutamine stress echocardiography for detecting CAD [17]. Dobutamine stress nuclear myocardial perfusion imaging (DSMPI) represents an alternative, exercise-independent stress modality for the detection of CAD. DSMPI is more sensitive (88%), but less specific (74%) than dobutamine stress echocardiography, and comparable with direct vasodilator myocardial perfusion imaging. Patients who have a normal DSMPI study have less than 1% annual rate of serious cardiac events [18]. The electron beam computed tomography controversy EBCT assesses atherosclerosis by measuring the extent of vascular calcification. In a meta-analysis of highly selected, symptomatic groups of patients [5], EBCT had a pooled sensitivity of 90.5% and specificity of 49.2%. Similar data for those who have no symptoms are lacking [6]. One study demonstrated that EBCT predicted silent ischemia as demonstrated by abnormal single photon emission computed tomography (SPECT) scan in asymptomatic moderate-to-high risk patients [19]. Increased coronary artery calcium scores predict subsequent development of heart disease events in the following 3.5 years in asymptomatic patients, though the increase is not directly proportional to scores [20]. In a cost-effectiveness analysis of EBCT, the marginal cost of using EBCT to identify an additional patient ‘‘at risk’’ that had been missed by the Framingham Risk Index is $9789. The study found that the cost per quality-adjusted life year saved was $86,752 when used to screen a population considered to be at low risk for CAD [21]. No study has examined the effect of EBCT data on clinical decision making [6,22]. The ACC/AHA Writing Group does not recommend EBCT to diagnose obstructive CAD in asymptomatic patients. Invasive testing Coronary angiography can be used to establish the diagnosis of CAD, and may be appropriately used to evaluate patients who have typical anginal symptoms; however, of all patients undergoing outpatient coronary angiography, an estimated 0.08% will die as a result of the procedure and 1.8%
868
JUNNILA & RUNKLE
will experience a potentially serious complication [23]. Invasive testing may be appropriate for patients who have a high probability of CAD, and is not recommended for patients who have a low probability of disease [9]. Biochemical markers of cardiovascular disease Several biochemical markers of cardiovascular disease are being explored for use in screening asymptomatic patients. Plasma natriuretic peptide levels predicted the risk of death and cardiovascular events in a community-based sample of asymptomatic patients who were followed for over 5 years [24]. Although this study did not specifically recommend using plasma natriuretic peptide levels as a trigger for further diagnostic tests in asymptomatic persons, it does raise the possibility that increased levels may aid in the early detection of cardiovascular disease. C-reactive protein is an inflammatory marker that is also being evaluated. Several well-designed studies have shown that elevated C-reactive protein levels are associated with development of nonfatal and fatal CAD in both men and women [25,26]. Other inflammatory markers, such as tumor necrosis factor alpha and interleukin-6, did not show a strong association with development of disease when adjusted for lipid levels. These studies concluded that the increased level of C-reactive protein is a significant contributor to the prediction of coronary artery disease. This finding was not confirmed in diabetic patients [27]. For the primary care physician, the clinical utility of these markers remains to be proven. Screening diabetic patients Diabetic patients who have no symptoms of CAD may present only after a significant CAD event, because silent ischemia is common in diabetic patients [3]. The American Diabetes Association (ADA) consensus guidelines suggest screening diabetic patients with stress testing when two or more additional CAD risk factors, including glomerular filtration rate (GFR) less than 90 mL/min, are present [3,27]. A newly-proposed diabetic cardiac risk score (DCRS) considers the same factors as in the Framingham Risk Index, as well as GFR less than 90 ml/min, presence of peripheral vascular disease, and need for insulin. A recent study compared the DCRS to the Framingham score as well as the ADA scoring system in diabetic patients undergoing CAD screening with exercise echocardiography [27]. The DCRS was found to be slightly more effective than the Framingham Risk Index in predicting which patients had positive findings of CAD on further screening, identifying those asymptomatic diabetic patients at highest risk for a significant CAD event. A diabetic patient who has no other cardiac risk factors may benefit from screening with a stress echocardiogram, because testing provides incremental data for risk stratification of diabetics who have suspected CAD. Diabetic patients who have normal exercise
CORONARY ARTERY DISEASE
869
echocardiography have significantly fewer coronary events when compared with those who have abnormal studies [28]. Abnormal results on stress echocardiography are an independent predictor of cardiac death in diabetic patients who have known or suspected CAD [29]. Screening those who impact public safety Although there are insufficient data to justify screening, considerations for public safety may influence the decision to screen for CAD. The sudden incapacitation or sudden death of people in certain occupations such as airline pilots, truckers, and heavy equipment operators may endanger the safety of others. For the evaluation of asymptomatic men older than 45 and women older than 55 who are involved in these occupations, it may be appropriate to recommend screening with exercise testing; however, its usefulness is not well-established [4,30]. Preparticipation testing It is estimated that the risk of sudden death ranges from 1:15,000 joggers per year to 1:50,000 in marathon participants [31]. There are conflicting data about the use of exercise testing to screen low-risk participants before starting an exercise program. The USPSTF position on preparticipation screening of asymptomatic patients notes that there is not enough evidence to determine the balance of benefits and harms of this practice [6,7]. The American Academy of Family Physicians (AAFP) does not recommend use of routine ECG as part of a periodic health or a preparticipation physical examination in asymptomatic patients [32]. The ACC/AHA guidelines find inadequate evidence to recommend exercise testing low-risk asymptomatic men older than 45 and women older than 55 who plan to start vigorous exercise [9]. It seems reasonable to screen older adults at moderate to high risk for CAD before starting an exercise program [31]; however, in a prospective cohort study of hypercholesterolemic men who were beginning an exercise program, sensitivity of exercise testing for predicting coronary events was 18% [33]. For the evaluation of asymptomatic persons who have diabetes and who plan to start vigorous exercise, the ACC/AHA is in favor of screening with exercise testing [6,7]. Preoperative evaluation Increased sympathetic drive combines with the cardiodepressant effects of anesthesia to create an increased risk of coronary events in the early postoperative period. There are few randomized controlled data regarding the optimum choice for preoperative evaluation of patients at risk for CAD. Consideration of the patient’s cardiovascular risk factors, the type and timing of surgery, and an estimate of functional capacity best determine the need for further investigation [34]. On the basis of available data, exercise
870
JUNNILA & RUNKLE
testing alone or with an imaging study remains the preferred test for preoperative evaluation [35]. Exercise or dobutamine stress echocardiography provide the best validated investigations [34].
The evidence for coronary artery disease therapy Therapy for CAD revolves around two goals: (1) secondary prevention of further morbidity and mortality from CAD and associated conditions; and (2) management of symptoms with the goal of improvement in quality of life. The former goal, for obvious reasons, is the main priority. Cornerstones in the secondary prevention of CAD include identification of high-risk patients who will obtain a mortality benefit from coronary artery bypass grafting (CABG), antiplatelet therapy or anticoagulation, use of beta blockers and angiotensin-converting enzyme (ACE) inhibitors, exercise, smoking cessation, lipid management, and blood pressure management. Key issues in symptom management include revascularization and medical therapy for angina. Antiplatelet therapy and anticoagulation Aspirin therapy has been convincingly demonstrated to decrease the risk of adverse cardiovascular events in patients who had previous MI ([myocardial infarction] number needed to treat [NNT] ¼ 29), acute MI (NNT ¼ 26), and other high-risk patients (those who have stable angina, atrial fibrillation, peripheral arterial disease, or diabetes, (NNT ¼ 45), although the risk of major hemorrhage is increased in patients at high risk of a vascular event (number needed to harm [NNH] ¼ 111). Interestingly, if there is a dose response to aspirin in the prevention of vascular events, it appears to be in favor of lower doses. A meta-analysis of over 200 trials [36] demonstrated benefit for daily aspirin doses of 75 to 325 mg, with higher doses having similar efficacy to lower doses, though doses less than 75 mg may be less effective (there are limited data with these doses). Lower doses are not significantly less likely to cause major hemorrhage [36]. A recently published meta-analysis of aspirin use [37] confirmed a small benefit in the primary prevention of cardiovascular events in both men and women (NNT ¼ 333 women, 270 men), and risk of bleeding was low (NNH ¼ 400 for women, 303 for men) [37]. The USPSTF strongly recommends discussion of aspirin chemoprophylaxis risks and benefits with patients at risk for CAD [7]. The use of thienopyridines (clopidogrel or ticlodipine) has been shown to have a small benefit when compared with aspirin in the prevention of the combined risk of MI, vascular death, or ischemic stroke (NNT ¼ 100), with no evident difference in the risk of hemorrhage, though they are obviously far more expensive [38]. One trial [39] showed a benefit of the addition of clopidogrel to aspirin in reducing death from cardiovascular causes, nonfatal MI, or stroke, as well as death from cardiovascular causes, nonfatal MI, stroke, or refractory ischemia.
CORONARY ARTERY DISEASE
871
High intensity anticoagulation therapy (international normalized ratio [INR] 2.8–4.8) with warfarin has also been shown to be effective in reducing the risk of MI and stroke in patients who have CAD (NNT ¼ 10), but carries a significant increased risk of bleeding (NNH ¼ .26) [40]. Warfarin therapy alone does not appear to be more effective than aspirin alone [40]. Trials of combination therapy with moderate- to high-intensity warfarin combined with aspirin have demonstrated improved cardiovascular outcomes compared with aspirin alone (NNT ¼ 18), but again possibly at an increased risk of hemorrhage (NNH ¼ 53) [40]. Despite the utility of glycoprotein (GP) IIb/IIIa inhibitors in decreasing the likelihood ischemic events associated with angioplasty, long term trials of oral GP IIb/IIIa inhibitors have been disappointing [41]. Lipid-lowering therapy Serum cholesterol levels are a strong predictor of development of symptomatic CAD, and lipid-lowering therapy has been shown to lower the risk of ischemia in patients who have CAD in many clinical trials. A metaanalysis [42] concluded that a cholesterol reduction of only 1% could be expected to lower coronary heart disease mortality by 1.7% (NNT ¼ 85), and all-cause mortality by 1.1% (NNT ¼ 91). Lipid-lowering therapy in patients who had modest elevation of cholesterol following MI significantly decreased the risk of recurrent events (NNT ¼ 33), need for CABG (NNT ¼ 50), and angioplasty (NNT ¼ 45) [43]. Lipid-lowering therapy in patients who have a history of MI or unstable angina results in a decrease in all-cause mortality and death from CAD at 7 years, regardless of initial cholesterol level (NNT ¼ 53) [44]. Patients who have established CAD are categorized as high risk by the ATP III, and the treatment goal established by these guidelines is to obtain a low-density lipoprotein (LDL) cholesterol level less than 100, with use of drug treatment if LDL level is above 130 [45]. The benefits of lipid-lowering therapy on cardiovascular mortality could be extended to high-risk patients who have initial total cholesterol greater than 135, regardless of initial LDL cholesterol (NNT ¼ 66) [46]. In patients who have hypertriglyceridemia (triglyceride O200), non-HDL cholesterol (calculated as total cholesterol – HDL cholesterol) is the closest surrogate marker for highly atherogenic very low-density lipoprotein (VLDL) lipoproteins readily available in clinical practice. Hypertriglyceridemia is a secondary target for lipid-lowering therapy, and should be treated with a combination of lifestyle change and pharmacologic therapy [11]. Low levels of HDL cholesterol have been identified as an independent risk factor for CAD, resulting in an increase in coronary risk of about 2% for every 1% decrease in HDL cholesterol. The use of gemfibrazil to treat low HDL has been shown to lead to lower risk of fatal and nonfatal MI in patients who have normal LDL levels(NNT ¼ 23) [47]. Drugs to raise
872
JUNNILA & RUNKLE
HDL should be considered in patients who have low HDL and normal triglycerides. This can be accomplished with the use of fibrates, 3-Hydroxy-3methylglutaryl coenzyme A (HMG Co-A) reductase inhibitors, or nicotinic acid [11]. Epidemiologic evidence suggests a benefit to a diet high in omega-3 fatty acids, common in fish and marine mammals. Supplementation with fish oil decreases triglycerides and has a variable effect on other lipoprotein levels [48]. Advice to eat at least two weekly portions of fatty fish results in significant decreases in death caused by ischemic heart disease (NNT ¼ 7) and overall mortality (NNT ¼ 8) [49]. A 1992 meta-analysis concluded that there was a benefit in the use of fish oil supplements in the prevention of restenosis following percutaneous intervention (PCI) (NNT ¼ 16) [50]. Hypertension treatment Blood pressure elevation has been consistently linked with CAD mortality in observational studies. Every 20 degrees of systolic blood pressure elevation over 115 and every 10 degrees of diastolic blood pressure elevation over 75 correlates to a doubling of CAD risk [51]. Although controlled trials of antihypertensive treatment have not been conducted on patients who have CAD, treatment of hypertension clearly decreases the risk of cardiovascular events in the general population, and is considered a key element in the management of patients who have CAD [9]. Lowering blood pressure by 12 mmHg over 10 years in patients who have established CAD will result in a significant decline in cardiac death (NNT ¼ 9) [9]. Beta blockers Beta blockers reduce cardiac events following MI, and more limited data support their use in patients who have CAD and who have not had an MI [4]. They appear to be equally to slightly more effective than calcium-channel blockers or long-acting nitrates in controlling angina symptoms, and are as well-tolerated as calcium-channel blockers. The effects of beta blockers combined with nitrates or calcium channel blockers on angina are additive, but these agents may be ineffective or harmful in treating vasospastic angina [9]. Angiotensin-converting enzyme inhibitor therapy Several trials have also shown ACE inhibitors to decrease the risk of cardiovascular death, MI, and stroke [4]. In high-risk patients treated with ramipril, significant reductions in death from cardiovascular causes (NNT ¼ 50), MI (NNT ¼ 42), revascularization (NNT ¼ 43), cardiac arrest (NNT ¼ 200), and heart failure (NNT ¼ 40) were noted, independent of any antihypertensive effect [52]. When patients who had known CAD without heart failure were randomized to receive the ACE inhibitor perindopril or placebo, the ACE inhibitor was shown to have a benefit in prevention of
CORONARY ARTERY DISEASE
873
a nonfatal MI (NNT ¼ 71), as well as a combined cardiovascular end point including total mortality, nonfatal MI, unstable angina, and cardiac arrest (NNT ¼ 19) [53]. Addition of an ACE inhibitor to optimal beta blockade has been shown to reduce exercise-induced myocardial ischemia in patients who have normal left ventricular function [9,54]. Calcium channel blockers Short-acting dihydropyridine calcium channel blockers cause a doserelated increase in the risk of mortality in patients who have a history of MI and unstable angina [55]. Nondihydropyridine calcium channel blockers and long-acting dihydropyridines do not appear to share this risk, and can be used to relieve the symptoms of angina [9]. Nitrates Nitrates relieve the symptoms of angina without increasing risk; however, nitrates do not improve mortality in patients who have CAD. Nitrates have been shown effective in improving exercise tolerance in patients who have chronic stable angina, and result in improved control of anginal symptoms when combined with beta blockers or calcium channel blockers. Sildenafil should not be used within 24 hours of administration of a nitrate product because of the high risk of severe hypotension. Prolonged use of nitrates leads to tolerance of the antianginal effects of the drug, which is generally prevented by maintaining an 8 to 12-hour nitrate-free interval [9]. Revascularization Revascularization can involve CABG using internal mammary artery grafts or vein grafts, or may involve PCI with or without stenting. Early randomized trials comparing CABG with medical management indicated that patients likely to receive a survival advantage from CABG are those who have left main coronary artery disease (NNT ¼ 10), multivessel disease (NNT ¼ 4), or proximal LAD stenosis (NNT ¼ 16) [56]. Patients determined to be at high risk by clinical criteria such as angina severity, history of hypertension, history of MI, and ST depression at rest as measured by an algorithm developed in the Veterans Administration cooperative study also benefit from revascularization (NNT ¼ 9) [56]. Because these trials were conducted, there have been advances in therapy, particularly the use of arterial grafts, and survival after CABG appears to have improved on the basis of observational studies [57]. Based on a meta-analysis, PCI may be superior to medical management in the management of anginal symptoms (NNT ¼ 9), but is more likely to result in the patient proceeding to CABG (NNH ¼ 33) [58]. In a study comparing percutaneous transluminal coronary angioplasty (PTCA) to medical therapy in patients considered suitable for either treatment regimen [59], no difference
874
JUNNILA & RUNKLE
was observed in mortality; however, the use of stents was uncommon in this study, and patients undergoing PTCA were more likely to have a nonfatal MI, (sometimes in conjunction with their procedure) (NNH ¼ 31). Patients randomized to PTCA were also more likely to require CABG (NNT ¼ 48) [59]. Ambulatory ECG monitoring may be able to identify a subgroup of patients likely to obtain a survival benefit from revascularization. Patients undergoing revascularization had improved survival when compared with patients treated medically to relieve anginal symptoms (NNT ¼ 18), as well as when compared with patients treated medically to relieve ischemic changes documented on ambulatory ECG (NNT ¼ 30) [60]. Patients who had asymptomatic ischemia on ambulatory ECG monitoring were more likely to have multivessel disease (58.3% versus 37.1%), or complex plaques (34.2% versus 18.5%) when compared with controls who did not have ambulatory ischemia, possibly explaining the survival benefit observed [61]. On the other hand, medical therapy coupled with aggressive lipid lowering therapy, although less effective in management of angina symptoms (NNT ¼ 8), results in fewer significant ischemic events (NNT ¼ 13) [62]. One study randomized patients who had three-vessel disease to receive either PCI or CABG [63]. The trial failed to show a survival benefit of CABG at 5 years (with the important exception of the subpopulation of diabetic patients who did obtain a survival advantage with CAGB (NNT ¼ 6); however, patients who underwent PCI were much more likely to require repeat revascularization by 5 years after the procedure (NNT ¼ 15), and were more likely to subsequently require CABG (NNT ¼ 56) [63]. By 7 years after randomization, a survival benefit for CABG was seen in the nondiabetic patients as well (NNT ¼ 8) [64]. Another study randomized patients who had severe proximal stenosis of the left anterior descending artery and stable angina to receive medical therapy, balloon angioplasty, or bypass surgery using the left internal mammary artery [65]. The study demonstrated prolonged event-free survival in the patients receiving surgery when compared with patients undergoing PCI (NNT ¼ 3), whereas patients undergoing medical therapy had intermediate results (NNT ¼ 4). There was no significant difference in mortality among the comparison groups [65]. In a trial that used stenting in addition to angioplasty in the PCI group, there was no survival advantage of CABG, though PCI patients were still more likely to require further revascularization (NNT ¼ 8) [66]. Risk factor modification Smoking cessation The link between smoking and the development of cardiovascular disease has been firmly established in numerous observational studies [67]. Smoking also increases the risk of recurrence of MI and increases the risk of sudden death in patients with angina [68]. Smoking has a dose-response relationship in the development of cardiovascular events, and its effects are additive to
CORONARY ARTERY DISEASE
875
other risk factors [69,51]. In an observational study, smoking cessation decreased the risk of cardiovascular events within 5 years in patients who had CAD (NNT ¼ 7) [70]. Risk reduction is evident within a year following MI and increases over time, despite evidence of increased infarct severity in patients who quit [68]. Although because of ethical considerations randomized trials of smoking cessation have not been conducted in patients who have established CAD and who smoke, it remains a key component to reduction of cardiovascular risk in patients who have CAD. Following a cardiac event, patients are particularly receptive to smoking cessation interventions [9]. Weight loss Obesity is associated with increased risk of coronary artery disease mortality, but it appears that most of the risk is caused by other associated risk factors, including impaired glucose tolerance, hypertension, and hyperlipidemia. Obesity, however, remained associated with increased risk of coronary morbidity in the Framingham cohort after adjusting for other known comorbidities, and was predictive of adverse coronary events, even in the small subset of patients who had obesity alone. In addition, weight loss was associated with a significant reduction in morbidity [71]. Exercise training Exercise training has a beneficial effect on lipid profiles, including triglycerides, LDL, and total cholesterol [72], and is useful in the management of other coronary risk factors, including diabetes, hypertension [73], and obesity. Lack of exercise is a risk factor for morbidity from CAD, and most of the benefit can be obtained by moderate levels of physical activity [74]. Regular exercise can prevent CAD, and improves symptoms in patients who have established CAD [75,76]. Following MI, cardiac rehabilitation programs involving exercise decreases cardiovascular mortality (NNT ¼ 56) and total mortality (NNT ¼ 47) [77]. Improvements in exercise tolerance have been clearly documented for ischemic heart disease patients treated with exercise in randomized clinical trials. There is no clear long-term benefit of higher intensity exercise programs, but these programs may be more effective when they involve exercise at least three times per week and last for at least 12 weeks [77]. Even suggestions by a physician on exercise can have benefit [78]. Prescription and supervision of exercise to patients with CAD can reduce mortality, but patient adherence to these regimens remains a problem [75]. There is no significant increase in cardiovascular complications or other serious outcomes demonstrated in CAD patients engaging in exercise programs [77], although patients engaging in regular physical activity are more likely to be injured than those who are sedentary (NNH ¼ 11) [79]. Counseling Type A behavior has been firmly established as an independent risk factor for cardiovascular mortality, and appears to be amenable to counseling,
876
JUNNILA & RUNKLE
resulting in decreased recurrence of cardiac events following MI (NNT ¼ 12) [80]. Moderate alcohol intake One study demonstrated a benefit of current alcohol consumption when compared with never drinking (NNT ¼ 72) and compared with former alcohol consumption (NNT ¼ 39) in men of Japanese ancestry living on Oahu [81]. Moderate alcohol intake (1.5 gm/d to 14.9 gm/d) in women is protective against the development of fatal or severe heart disease, although it is associated with an increased risk of subarachnoid hemorrhage [82]. A case-control study has suggested that increases in HDL cholesterol with alcohol intake may be the source of the observed association with lower cardiovascular morbidity [83]. Cardiac rehabilitation Cardiac rehabilitation services are defined as ‘‘comprehensive, long-term programs involving medical evaluation, prescribed exercise, cardiac risk factor modification, education, and counseling’’ [77]. These programs are designed to avoid morbidity and mortality and to enhance functioning in cardiac patients. Such services are recommended on the basis of improvements in exercise tolerance, symptoms, lipid levels, cigarette smoking, psychosocial well-being and reduction of stress, and mortality [77]. Nurse-led, secondary prevention clinics focusing on use of aspirin, blood pressure and lipid management, lifestyle factors, and behavioral change were able to demonstrate improvements in aspirin therapy, blood pressure management, low fat diet, and exercise at 1 year. All improvements but exercise were sustained at 4 years, with subsequent improvements in survival [84]. Education, counseling, and behavioral modification training as part of cardiac rehabilitation have been demonstrated to improve rates of smoking cessation and prevent relapse, improve lipid levels, reduce anginal symptoms, and improve psychological outcomes, and may promote regression of atherosclerosis and prevent reinfarction. Education alone is unlikely to result in weight loss, but modest weight loss can be attained when educational programs are combined with behavioral interventions. These strategies have not been demonstrated to be effective in improving exercise tolerance, controlling blood pressure, or hastening return to work [77]. Alternative therapies Spinal cord stimulation appears to be effective in reducing anginal symptoms in patients refractory to standard therapy, resulting in increased exercise capacity and decreased frequency of angina. Interestingly, ischemic episodes on ambulatory ECG monitoring were also decreased with therapy [85]. In a randomized trial of CABG compared with spinal cord stimulation
CORONARY ARTERY DISEASE
877
in patients refractory to maximal medical therapy, there was no survival advantage to CABG, and the results in symptom relief were similar [86]. Enhanced external counterpulsation appears effective in angina symptom management as well [87]. For patients with refractory angina not amenable to conventional revascularization, laser transmyocardial revascularization has been found to be significantly more effective in improving anginal symptoms and improving quality of life than continued medical therapy at 3 months (NNT ¼ 2–3) and persisting to 12 months (NNT ¼ 3–5) without observed difference in mortality [88–90]. In one study [91], difference in mortality rate did not reach statistical significance, but favored medical therapy (NNH ¼ 16), leading the authors to recommend against the procedure. Another trial [92] used a percutaneous catheter for transmyocardial revascularization versus continued medical therapy and demonstrated significant improvement in angina (NNT ¼ 5), with no significant difference in survival between groups. Garlic has been shown to decrease cholesterol in the short, but not long term, and there are no data to support its effect on cardiovascular events. Vitamin B6, B12, and folate may lower serum homocysteine levels, which are higher in patients with CAD and may be associated with higher morbidity and mortality; however, the use of such treatment has not been studied in the prevention of coronary events. Acupuncture may increase time to angina with exercise and improve work capacity, though the results of randomized controlled trials are nonhomogeneous. Controlled trials of Chinese herbal therapy have also shown improvements in angina symptoms. Chelation therapy has not been shown to be effective in the treatment of angina [9].
Monitoring patients with known coronary artery disease In the patient who has chronic stable angina, routine testing is of little use without a change in history or physical examination. There is little evidence regarding the follow-up of patients who have known CAD. Based on expert opinion, the ACC/AHA recommends clinical evaluation every 4 to 6 months during the first year of therapy, with annual evaluations to follow. Rather than refer all patients to cardiologists, primary care physicians are encouraged to comanage their patients with alternating visits. The ACC/AHA suggests five questions that should be answered regularly during the follow-up of a patient who is receiving treatment for chronic stable angina (Fig. 2) [93]: Has the patient’s level of physical activity decreased since the last visit? Have the patient’s anginal symptoms increased in frequency or become more severe since the last visit? How well is the patient tolerating therapy? How successful has the patient been in modifying risk factors and improving knowledge about ischemic heart disease?
878
JUNNILA & RUNKLE
Fig. 2. Chronic stable angina. (From Texas Tech University Managed Health Care Network Pharmacy & Therapeutics Committee. Chronic stable angina. University of Texas Medical Branch Correctional Managed Care; 2003. Ó Copyright 2003 University of Texas Medical Branch Correctional Managed Care.)
CORONARY ARTERY DISEASE
879
Has the patient developed any new comorbid illnesses, or has the severity or treatment of known comorbid illnesses worsened the patient’s angina?
Use of cardiac testing during follow-up Cardiac testing should only be considered based on a clinical change in the patient’s status. The ACC/AHA writing committee consensus concluded that the following studies are indicated [93]: Repeated echocardiogram when therapy with medications affecting cardiac conduction are initiated or changed, or when anginal pattern has changed, symptoms or findings suggest a dysrhythmia or conduction abnormality, or near or frank syncope occurs Chest radiography for patients who have evidence of new or worsening congestive heart failure (CHF). The American College of Radiology concurs with this recommendation, noting that typical findings of CHF are often noted on radiograph, and that diseases other than CHF may present with one or more of the signs or symptoms of CHF [94]. Assessment of left ventricular ejection fraction and segmental wall motion by echocardiography or radionuclide imaging in patients who have new or worsening CHF or evidence of intervening MI by history or electrocardiography. One retrospective cohort study [95] confirms that patient history, ECG results, and chest radiograph can all be used to predict the absence of systolic dysfunction. Patients who have normal findings are very unlikely to have left ventricular systolic dysfunction and do not need echocardiography [95]. Echocardiography for patients who have evidence of new or worsening valvular heart disease Treadmill exercise test for patients who had no previous revascularization who have a significant change in clinical status, can exercise, and have none of the following electrocardiogram abnormalities: pre-excitation (Wolff-Parkinson-White) syndrome, electronically paced ventricular rhythm, more than 1 mm of ST-segment depression at rest, or complete left bundle-branch block. The Duke Treadmill Score predicts subsequent cardiac events. Techniques for calculating the Duke Score are described elsewhere [96]. Stress radionuclide imaging or stress echocardiography procedures for patients who had or did not have previous revascularization and who have a significant change in clinical status and cannot exercise, or who have any of the electrocardiogram abnormalities listed above Stress radionuclide imaging or stress echocardiography procedures for patients who have a significant change in clinical status and required a stress imaging procedure on their initial evaluation because of equivocal or intermediate-risk results with exercise electrocardiography testing
880
JUNNILA & RUNKLE
Coronary angiography in patients who have marked limitation of ordinary activity despite maximal medical therapy. Evidence confirms the utility of angiography in patients who have severe stable or unstable anginal symptoms that are resistant to medication [97].
References [1] National Center for Health Statistics, Centers for Disease Control and Prevention, US Department of Health and Human Services. Fast stats A to Z: leading causes of death. Available at: http://www.cdc.gov/nchs/fastats/lcod.htm. Accessed December 12, 2005. [2] American Heart Association. Cardiovascular disease statistics. Available at: http://www. americanheart.org/presenter.jhtml?identifier¼4478. Accessed December 12, 2005. [3] Wackers FJ, Young LH, Inzucchi SE, et al. Detection of silent myocardial ischemia in asymptomatic diabetic subjects: the DIAD study. Diabetes Care 2004;27(8): 1954–61. [4] Gibbons RJ, Balady GJ, Bricker JT, et al. ACC/AHA/2002 guidelines update for exercise testing: summary article: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines. Circulation 2002;106:1883–92. [5] Snow V, Barry P, Fihn SD, et al. Evaluation of primary care patients with chronic stable angina: guidelines form the American College of Physicians. Ann Intern Med 2004;141(7): 57–64. [6] US Preventive Services Task Force. Screening for coronary heart disease: recommendation statement. Ann Intern Med 2004;140:569–72. [7] USPSTF recommendations (USPSTF Task Force February 2004). Available at: http:// www.ahrq.gov/clinic/uspstf/uspsacad.htm. Accessed December 10, 2005. [8] Diamond GA, Forrester JS. Analysis of probability as an aid in the clinical diagnosis of coronary artery disease. N Engl J Med 1979;300:1350–8. [9] Gibbons RJ, Abrams J, Chatterjee K, et al. ACC/AHA 2002 guideline update for the management of patients with chronic stable angina: A report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines (Committee to Update the 1999 Guidelines for the Management of Patients With Chronic Stable Angina). Available at: www.acc.org. Accessed January 23, 2006. [10] Framingham heart study-based risk calculator. Available at: http://hin.nhlbi.nih.gov/atpiii/ calculator.asp?usertype¼prof. Accessed January 19, 2006. [11] Executive Summary of the third report of the National Cholesterol Education Program (NCEP) Expert Panel on Detection. Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III). JAMA 2001;285:2386–97. [12] Coronary Artery Surgery Study (CASS). A randomized trial of coronary artery bypass surgery. Circulation 1983;68:939–50. [13] Fowler-Brown A, Pignone M, Pletcher M, et al. Exercise tolerance testing to screen for coronary heart disease; a systematic review for the technical support for the US Preventive Services Task Force. Ann Intern Med 2004;140(4):W9–24. [14] Gianrossi R, Detrano R, Mulvihill D, et al. Exercise-induced ST depression in the diagnosis of coronary artery disease: a meta-analysis. Circulation 1989;80:87–98. [15] Sox HC, Littenberg B, Garber AM. The role of exercise testing in screening for coronary artery disease. Ann Intern Med 1989;110:456–69. [16] Mahenthiran J, Bangalore S, Yao S, et al. Comparison of prognostic value of stress echocardiography versus stress electrocardiography in patients with suspected coronary artery disease. Am J Cardiol 2005;96(5):628–34. [17] Tsutsui JM, Elhendy A, Anderson JR, et al. Prognostic value of dobutamine stress myocardial contrast perfusion echocardiography. Circulation 2005;112:1444–50.
CORONARY ARTERY DISEASE
881
[18] Geleijnse ML, Elhendy A, Fioretti PM, et al. Dobutamine stress myocardial perfusion imaging. J Am Coll Cardiol 2001;36(7):2017–27. [19] He ZX, Hedrick TD, Pratt CM, et al. Severity of coronary artery calcification by electron beam computed tomography predicts silent myocardial ischemia. Circulation 2000;101: 244–51. [20] Pletcher MJ, Tice JA, Pignone M, et al. Using the coronary artery calcium score to predict coronary heart disease events. Arch Intern Med 2004;164:1285–92. [21] O’Malley PG, Greenberg BA, Taylor AJ. Cost-effectiveness of using electron beam computed tomography to identify patients at risk for clinical coronary artery disease. Am Heart J 2004;148:106–13. [22] Pignone M, Fowler-Brown A, Pletcher M, et al. Screening for asymptomatic coronary artery disease. Systematic evidence review No 22. Rockville (MD): Agency for Healthcare Research and Quality; 2003. Available at: wwwahrq.gov/downloads/pub/prevent/pdfser/chdser.pdf. [23] Bashore TM, Bates ER, Berger PB, et al. American College of Cardiology/Society for Cardiac Angiography and Interventions clinical expert consensus document of cardiac catheterization laboratory standards. J Am Coll Cardiol 2001;37:2170–214. [24] Wang TH, Larson MG, Levy D, et al. Plasma natriuretic peptide levels and the risk of cardiovascular events and death. N Engl J Med 2004;350(7):655–63. [25] Pai JK, Psichon T, Ma J, et al. Inflammatory markers and the risk of coronary heart disease in men and women. N Engl J Med 2004;351(25):2599–610. [26] Wilson PW. CDC/AHA workshop on markers of inflammation and cardiovascular disease: ability of inflammatory markers to predict disease in asymptomatic patients. Circulation 2004;110(25):e568–71. [27] Rakhit DJ, Downey M, Jeffries L, et al. Screening for coronary artery disease in patients with diabetes: a Bayesian strategy of clinical risk evalution and exercise echocardiography. Am Heart J 2005;150:1074–80. [28] Elhendy A, Arruda AM, Mahoney DW, et al. Prognostic stratification of diabetic patients by exercise echocardiography. J Am Coll Cardiol 2001;37(6):155–7. [29] Marwick TH, Case C, Sawada S, et al. Use of stress echocardiography to predict mortality in patients with diabetes and known or suspected coronary artery disease. Diabetes Care 2002; 25(6):1042–8. [30] Lee TH. Use of exercise tolerance testing: AHA/ACC guidelines summary. In: Braunwald E, Zipes DP, Libby P, editors. Heart disease: a textbook of cardiovascular medicine. Philadelphia: WB Saunders; 2001:101. [31] Seto CK. Preparticipation cardiovascular screening. Clin Sports Med 2003;22:23–35. [32] AAFP recommendations for clinical preventive services. Available at: http://www.aafp.org/ x24976.xml. Accessed January 23, 2006. [33] Siscovick DS, Ekelund LG, Johnson JL, et al. Sensitivity of exercise electrocardiography for acute cardiac events during moderate and strenuous physical activity. The Lipid Research Clinics Coronary Primary Prevention Trial. Arch Intern Med 1991;151:325–30. [34] Ashley EA. Preoperative cardiac evaluation: mechanisms, assessment, and reduction of risk. Thorac Surg Clin 2005;15(2):263–75. [35] AAFP summary of ACC/AHA guidelines for ambulatory ECG. Available at: http:// www.aafp.org/afp/20000201/practice.html. Accessed January 23, 2006. [36] Antithrombotic Trialists’ Collaboration. Collaborative metaanalysis of randomised trials of antiplatelet therapy for prevention of death, myocardial infarction, and stroke in high risk patients. BMJ 2002;324:71–86. [37] Berger JSRM, Avanzini F, Pangrazzi I, et al. Aspirin for the primary prevention of cardiovascular events in women and men: a sex-specific meta-analysis of randomized controlled trials. JAMA 2006;295:306–13. [38] Hankey GJSC, Dunbabin DW. Thienopyridines or aspirin to prevent stroke and other serious vascular events in patients at high risk of vascular disease? A systematic review of the evidence from randomized trials. Stroke 2000;31:1779–84.
882
JUNNILA & RUNKLE
[39] Stein PD, Holger J, Schu¨nemann, et al. Antithrombotic therapy in patients with saphenous vein and internal mammary artery bypass grafts: The Seventh ACCP Conference on Antithrombotic and Thrombolytic Therapy. Chest 2004;126(Suppl 3):600s–8s. [40] Anaud SYS. Oral anticoagulant therapy in patients with coronary artery disease: a metaanalysis. JAMA 1999;282:2058–67. [41] Heeschen C, Hamm C. Difficulties with oral platelet glycoprotein IIb/IIIa receptor antagonists. Lancet 2000;355(9201):330–1. [42] Gould ALRJ, Santanello NC, Heyse JF, et al. Cholesterol reduction yields clinical benefit: impact of statin trials. Circulation 1998;97:946–52. [43] Sacks FM, Pfeffer MA, Moye LA, et al. The effect of pravastatin on coronary events after myocardial infarction in patients with average cholesterol levels. Choleterol and Recurrent Events Trial investigators. N Engl J Med 1996;335:1001–9. [44] The Long-Term Intervention with Pravastatin in Ischaemic Disease (LIPID) Study Group. Prevention of cardiovascular events and death with pravastatin in patients with coronary heart disease and a broad range of initial cholesterol levels. N Engl J Med 1998;339:1349–57. [45] Grundy SM, Cleeman J, Merz CN, et al. Implications of recent clinical trials for the National Cholesterol Education Program Adult Treatment Panel III guidelines. Circulation 2004;110: 227–39. [46] Heart Protection Study Collaborative Group. MRC/BHF Heart Protection Study of cholesterol lowering with simvastatin in 20,536 high-risk individuals: a randomised placebocontrolled trial. Lancet 2002;360(9326):7–22. [47] Rubins HB, Robins SJ, Collins D, et al. Gemfibrozil for the secondary prevention of coronary heart disease in men with low levels of high-density lipoprotein cholesterol. N Engl J Med 1999;341(6):410–8. [48] Israel DH, Gorlin R. Fish oils in the prevention of atherosclerosis. J Am Coll Cardiol 1992; 19(1):174–85. [49] Burr M, Fehily A, Gilbert J, et al. Effects of changes in fat, fish, and fibre intakes on death and myocardial reinfarction: Diet and Reinfarction Trial (DART). Lancet 1989;2(8666): 757–61. [50] O’Connor GT, Malenka DJ, Olmstead EM, et al. A meta-analysis of randomized trials of fish oil in prevention of restenosis following coronary angioplasty. Am J Prev Med 1992; 8(3):186–92. [51] National High Blood Pressure Education Program. The seventh report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure: National Institutes of Health, National Heart, Lung, and Blood Institute, National High Blood Pressure Education Program. NIH Publication No. 04–5230. Bethesda (MD): National Institutes of Health; 2004. [52] The Heart Outcomes Prevention Evaluation Study Investigators. Effects of an angiotensinconverting enzyme inhibitor, ramipril, on cardiovascular events in high-risk patients. N Engl J Med 2000;342(3):145–53. [53] Fox K, European Trial On Reduction of Cardiac Events with Perindopril in Stable Coronary Artery Disease investigators. Efficacy of perindopril in reduction of cardiovascular events among patients with stable coronary artery disease: randomised, double-blind, placebocontrolled, multicentre trial (the EUROPA study). Lancet 2003;362(9386):78–8. [54] van der Heuvel AF, Dunselman P, Kingma T, et al. Reduction of exercise-induced myocardial ischemia during add-on treatment with the angiotensin-converting enzyme inhibitor enalapril in patients with normal left ventricular function and optimal beta blockade. J Am Coll Cardiol 2001;37(2):470–4. [55] Furberg CD, Psaty BM, Meyer JV. Nifedipine: dose-related increase in mortality in patients with coronary heart disease. Circulation 1995;92:1326–31. [56] Yusuf S, Zucker D, Peduzzi P, et al. Effect of coronary artery bypass graft surgery on survival: overview of 10-year results from randomised trials by the Coronary Artery Bypass Graft Surgery Trialists Collaboration. Lancet 1994;344(8922):563–70.
CORONARY ARTERY DISEASE
883
[57] Cameron A, Davis K, Green G, et al. Coronary bypass surgery with internal-thoracic-artery graftsdeffects on survival over a 15-year period. N Engl J Med 1996;334(4):216–9. [58] Bucher H, Hengstler P, Schindler C, et al. Percutaneous transluminal coronary angioplasty versus medical treatment for non-acute coronary heart disease: meta-analysis of randomised controlled trials. BMJ 2000;321(7253):73–7. [59] Rita-2 trial participants. Coronary angioplasty verses medical therapy for angina: the second Randomised Intervention Treatment of Angina (RITA-2) trial. Lancet 1997;350:461–8. [60] Davies R, Goldberg A, Forman S, et al. Asymptomatic Cardiac Ischemia Pilot (ACIP) study two-year follow-up: outcomes of patients randomized to initial strategies of medical therapy versus revascularization. Circulation 1997;95(8):2037–43. [61] Sharaf B, Williams D, Miele N, et al. A detailed angiographic analysis of patients with ambulatory electrocardiographic ischemia: results from the Asymptomatic Cardiac Ischemia Pilot (ACIP) study angiographic core laboratory. J Am Coll Cardiol 1997;29(1):78–84. [62] Pitt B WD, Brown WV, van Boven AJ, et al. Aggressive lipid lowering therapy compared with angioplasty in stable coronary artery disease. N Engl J Med 1999; 341(2):70–86. [63] Frye RL, The BARI investigators. Comparison of coronary bypass surgery with angioplasty in patients with multivessel disease. N Engl J Med 1996;335(4):217–25. [64] The BARI investigators. Seven-year outcome in the Bypass Angioplasty Revascularization Investigation (BARI) by treatment and diabetic status. J Am Coll Cardiol 2000;35(5): 1122–9. [65] Hueb WA, Soares PR, Oliveira AD Sr, et al. Five-year follow-up of the Medicine, Angioplasty, or Surgery Study (MASS): a prospective, randomized trial of medical therapy, balloon angioplasty, or bypass surgery for single proximal left anterior descending coronary artery stenosis. Circulation 1999;100(Suppl II):II-107–13. [66] Serruys P, Unger F, Sousa J, et al. Comparison of coronary-artery bypass surgery and stenting for the treatment of multivessel disease. N Engl J Med 2001;344(15):1117–24. [67] McBride P. The health consequences of smoking. Cardiovascular diseases. Med Clin North Am 1992;76(2):333–53. [68] CDC. The health benefits of smoking cessation: a report of the Surgeon General. Rockville (MD): US Department of Health and Human Services, Public Health Service, 1990. DHHS publication number (CDC) 90–8416. [69] Mulcahy R, Hickey N, Graham I, et al. Factors affecting the 5 year survival rate of men following acute coronary heart disease. Am Heart J 1977;93(5):556–9. [70] Vlietstra R, Kronmal R, Oberman A, et al. Effect of cigarette smoking on survival of patients with angiographically documented coronary artery disease. Report from the CASS registry. JAMA 1986;255(8):1023–7. [71] Hubert HB, Feinleib M, McNamara PM, et al. Obesity as an independent risk factor for cardiovascular disease: a 26-year follow-up of participants in the Framingham Heart Study. Circulation 1983;67(5):968–77. [72] Leon AS, Sanchez OA. Response of blood lipids to exercise training alone or combined with dietary intervention. Med Sci Sports Exerc 2001;33:S502–15. [73] Fagard RH. Exercise characteristics and the blood pressure response to dynamic physical training. Med Sci Sports Exerc 2001;33:S484–92. [74] Fletcher GF, Balady G, Blair SN, et al. Statement on exercise: benefits and recommendations for physical activity programs for all americans: a statement for health professionals by the Committee on Exercise and Cardiac Rehabilitation of the Council on Clinical Cardiology, American Heart Association. Circulation 1996;94:857–62. [75] Jolliffe JA, Rees K, Taylor RS, et al. Exercise-based rehabilitation for coronary heart disease. Cochrane Database Syst Rev 2001;1:CD001800. [76] Thompson PD, Buchner D, Pin˜a IL, et al. Exercise and physical activity in the prevention and treatment of atherosclerotic cardiovascular disease: a statement from the Council on Clinical Cardiology (Subcommittee on Exercise, Rehabilitation, and Prevention) and the
884
[77]
[78] [79] [80]
[81] [82]
[83]
[84]
[85] [86]
[87]
[88] [89]
[90]
[91] [92]
[93]
[94] [95]
JUNNILA & RUNKLE
Council on Nutrition, Physical Activity, and Metabolism (Subcommittee on Physical Activity). Circulation 2003;107:3109–16. AHCPR. Cardiac rehabilitation. Agency for Health Care Policy and Research. Available at: http://www.ncbi.nlm.nih.gov/books/bv.fcgi?rid¼hstat2.chapter.6677. Accessed January 23, 2006. Lewis B, Lynch W. The effect of physician advice on exercise behavior. Prev Med 1993;22(1): 110–21. Hootman JM, Macera CA, Ainsworth BE, et al. Epidemiology of musculoskeletal injuries among sedentary and physically active adults. Med Sci Sports Exerc 2002;34(5):838–44. Friedman M, Thoresen C, Gill J, et al. Alteration of type A behavior and its effect on cardiac recurrences in post myocardial infarction patients: summary results of the recurrent coronary prevention project. Am Heart J 1986;112(4):653–65. Yano K, Rhoads G, Kagan A. Coffee, alcohol and risk of coronary heart disease among Japanese men living in Hawaii. N Engl J Med 1977;297(8):405–9. Stampfer M, Colditz G, Willett W, et al. A prospective study of moderate alcohol consumption and the risk of coronary disease and stroke in women. N Engl J Med 1988;319(5): 267–73. Gaziano J, Buring J, Breslow J, et al. Moderate alcohol intake, increased levels of highdensity lipoprotein and its subfractions, and decreased risk of myocardial infarction. N Engl J Med 1993;329(25):1829–34. Murchie PCN, Ritchie LD, Simpson JA, et al. Secondary prevention clinics for coronary heart disease: four year follow up of a randomised controlled trial in primary care. BMJ 2003;326(7380):84. Hautvast R, DeJongste M, Staal M, et al. Spinal cord stimulation in chronic intractable angina pectoris: a randomized, controlled efficacy study. Am Heart J 1998;136(6):1114–20. Ekre O, Eliasson T, Norrsell H, et al. Long-term effects of spinal cord stimulation and coronary artery bypass grafting on quality of life and survival in the ESBY study. Eur Heart J 2002;23:1938–45. Arora R, Chou T, Jain D, et al. The Multicenter Study of Enhanced External Counterpulsation (MUST-EECP): effect of EECP on exercise-induced myocardial ischemia and anginal episodes. J Am Coll Cardiol 1999;33(7):1833–40. Allen K, Dowling R, Fudge T, et al. Comparison of transmyocardial revascularization with medical therapy in patients with refractory angina. N Engl J Med 1999;341(14):1029–36. Burkhoff D, Schmidt S, Schulman S, et al. Transmyocardial laser revascularisation compared with continued medical therapy for treatment of refractory angina pectoris: a prospective randomised trial. ATLANTIC Investigators. Angina treatmentsdlasers and normal therapies in comparison. Lancet 1999;354(9182):885–90. Aaberge L, Nordstrand K, Dragsund M, et al. Transmyocardial revascularization with CO2 laser in patients with refractory angina pectoris. Clinical results from the Norwegian randomized trial. J Am Coll Cardiol 2000;35(5):1170–7. Schofield P, Sharples L, Caine N, et al. Transmyocardial laser revascularisation in patients with refractory angina: a randomised controlled trial. Lancet 1999;353(9152):519–24. Oesterle S, Sanborn T, Ali N, et al. Percutaneous transmyocardial laser revascularisation for severe angina: the PACIFIC randomised trial. Potential class improvement from intramyocardial channels. Lancet 2000;356(9243):1705–10. Snow V, Barry P, Fihn SD, et al. Primary care management of chronic stable angina and asymptomatic suspected or known coronary artery disease: a clinical practice guideline from the American College of Physicians. Ann Intern Med 2004;141:562–7. American College of Radiology (ACR), Expert Panel on Thoracic Imaging. Congestive heart failure. Reston (VA): American College of Radiology (ACR); 2003. p. 4. Talreja D, Gruver C, Sklenar J, et al. Efficient utilization of echocardiography for the assessment of left ventricular systolic function. Am Heart J 2000;139:394–8.
CORONARY ARTERY DISEASE
885
[96] Mark DB, Hlatky MA, Harrell FE Jr, et al. Exercise treadmill score for predicting prognosis in coronary artery disease. Ann Intern Med 1987;106(6):793–800. [97] Finnish Medical Society Duodecim. Coronary angiography and indications for CABG or angioplasty. In: EBM guidelines. Evidence-based medicine [CD-ROM]. Helsinki (Finland): Duodecim Medical Publications Ltd; 2004. Available at: http://wwwguidelines.gov/ summary/summary.aspx?ss¼15&doc_id¼8811&nbr¼4857 (accessed 23 Jan 2006).
Prim Care Clin Office Pract 33 (2006) 887–901
Initiation of Therapy for Patients with Essential Hypertension or Comorbid Conditions Randy Wexler, MD, MPHa,*, David Feldman, MD, PhDb a
Department of Family Medicine, B0902B Cramblett Hall, 456 West 10th Avenue, Columbus, OH 43201, USA b Division of Cardiology at The Ohio State University Medical Center, 244 Davis Heart and Lung Research Institute, 473 West 12th Ave., Columbus, OH 43210, USA
Hypertension is defined as a systolic blood pressure greater than 140 mm Hg, or a diastolic blood pressure in excess of 90 mm Hg. Often a silent disease in its early stages, it is a major contributor to morbidity and mortality worldwide. In the United States, 65 million adults are hypertensive [1]. Hypertension is the number one reason for adult ambulatory office visits in family medicine [2]. One half of all patients who suffer a first-time myocardial infarction (MI), and two thirds of all patients who suffer an initial stroke, have a blood pressure that exceeds 140/90 mm Hg [3]. Additionally, high blood pressure is a major comorbid predictor of heart failure; it precedes this diagnosis in more than 75% of cases [3]. The financial burden of hypertension is staggering. In 2005, the estimated total costs (direct and indirect costs) were $59.7 billion [4]. Ninety-five percent of individuals who have hypertension have essential (primary, benign, idiopathic) hypertension. Typically, its onset occurs after the second decade of life [5]. African Americans suffer disproportionately, with 43% of women and 39% of men afflicted in comparison with their white counterparts (28% and 29%, respectively) [6]. The reason for this disparity is the subject of intense debate. although not understood clearly, an increase in sodium sensitivity in the African American population might account for part of the hypertension disease burden [5,7,8]. Such findings, based on molecular research, are supported by lifestyle studies that have
* Corresponding author. E-mail address:
[email protected] (R. Wexler). 0095-4543/06/$ - see front matter Ó 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.pop.2006.09.006 primarycare.theclinics.com
888
WEXLER & FELDMAN
demonstrated a greater reduction in blood pressure among African Americans while following the Dietary Approaches to Stop Hypertension eating plan in conjunction with a low-sodium diet [9]. Pathophysiology A simplified equation for blood pressure may be characterized as blood pressure ¼ cardiac output peripheral vascular resistance. Although this equation is useful, it underestimates the complexity of the underlying pathophysiology. This includes the renin-angiotensin-aldosterone system, the autonomic nervous system, bradykinin, endothelin, nitric oxide (endothelial-derived relaxing factor), atrial natriuretic peptide, and ouabain [5]. During the initial stages of hypertension, there is an increase in cardiac output, which, in part, is attributable to sympathetic up-regulation, and thereafter begins a cascade of events that leads to a sustained elevation in blood pressure [5,10]. The physiologic response is a sustained increase in peripheral vascular resistance that is followed by the chronic elevation in blood pressure [5]. The genetics of high blood pressure is a rapidly evolving science. Although it is well recognized that blood pressure has a familial predilection, its causes are not well understood at the molecular level. Genetic factors may contribute in up to 30% of all cases of hypertension. Individuals who have one or two parents who have hypertension are twice as likely to develop essential hypertension [5,11]. Although genetics are important, environment and lifestyle factors have a significant impact on patients’ blood pressure. Prognosis Although hypertension is defined as a measurement in excess of 140/90 mm Hg, evidence is accumulating that this degree of blood pressure elevation may be too high. Lewington and colleagues [12] concluded after reviewing 61 prospective studies with more than 1 million participants that with each decade of life, there was a proportional increase in the risk for cardiovascular death when blood pressure increased above 115 mm Hg systolic or 75 mm Hg diastolic. Using the Framingham Heart Study Database, Vasan and colleagues [13] demonstrated that individuals who had a high-normal blood pressure [130–139 mm Hg systolic, and 85–89 mm Hg diastolic] had a twofold increased risk for cardiovascular disease. Tierney and colleagues [14] discovered that a single elevated systolic blood pressure reading could be predictive of an increased risk for stroke and heart disease over the ensuing 5-year period. Although additional studies along this line are needed, the common approach to treating sporadic or isolated increases in systolic blood pressure needs to be examined. It is estimated that only one third of those who have hypertension have a blood pressure that is treated to less than 140/90 mm Hg. In addition,
889
ESSENTIAL HYPERTENSION OR COMORBID CONDITIONS
almost one third of adults who have hypertension are unaware they have hypertension, and only 60% of patients who acknowledge that they have hypertension are being treated [15]. Modest reductions in blood pressure can result in large benefits in patients. A decrease of 5 mm Hg decreases mortality that is due to stroke by 14%, cardiovascular-related mortality by 9%, and all-cause mortality by 7% [16]. These data suggest that aggressive intervention to effect small changes may make a significant difference in morbidity and mortality. Hypertension is treated poorly in the United States [17–21]. Frequently, physicians ignore treatment guidelines or fail to treat mildly elevated systolic pressures [21–24]. This may be due, in part, to a lack of awareness of hypertension treatment guidelines and treatment recommendations in the primary care setting [15]. Hyman and Pavlik [24] surveyed a national sample of primary care physicians to determine their practice patterns for the treatment of hypertension and their familiarity with Joint National Committee [JNC] guidelines. Forty-one percent of those surveyed were not familiar with the guidelines or their recommendations. The lack of adherence to the guidelines seems to have less to do with intellectual disagreements, and more to do with physician awareness. These findings suggest that persistent, untreated hypertension is due, in part, to physician nonadherence to these evidence-based guidelines [22,25–30]. The Seventh Report of the Joint National Committee on Prevention, Detection, Evaluation and Treatment of High Blood Pressure The Seventh Report of The Joint National Committee on Prevention, Detection, Evaluation and Treatment of High Blood Pressure (JNC 7) commenced in the fall of 2002 by the National High Blood Pressure Education Program Coordinating Committee. Its purpose was to evaluate new information in the detection and treatment of hypertension [31]. One of the significant revisions made by JNC 7 was a reclassification of blood pressure stages as well as the creation of a new category, ‘‘prehypertension’’ (Table 1). Prehypertension is not considered to be a disease state. Table 1 The Seventh Report of the Joint National Committee on the Prevention, Detection, Evaluation and Treatment of High Blood Pressure hypertension classification Blood pressure (mm Hg)
Classification
!120/80 120–139/80–89 R140/90 140–159/90–99 160–179/100–109
Normal Prehypertension Hypertension Stage 1 Stage 2
Data from Chobanian A, Bakris G, Black H, et al. Seventh Report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure. Hypertension 2003;42:1206–52.
890
WEXLER & FELDMAN
This designation serves to identify persons who are at risk for developing hypertension so that early interventions may be implemented. Pharmacotherapy is not recommended in persons who have prehypertension unless other compelling indications (eg, diabetes or heart failure) are present. Lifestyle modification is the intervention of choice [15]. JNC 7 also listed several key messages: Increased systolic blood pressure is a greater risk factor for cardiovascular disease than is increased diastolic blood pressure in persons older than 50 years of age [32]. Cardiovascular risk increases as blood pressure exceeds 115/75 mm Hg [12]. In addition, cardiovascular risk doubles for each increase of 20/10 mm Hg greater than 115/75 mm Hg [15]. Patients who are older than 55 years of age have a 90% likelihood of developing hypertension during their remaining years of life [33]. Individuals who have prehypertension should make lifestyle modifications to prevent or delay the onset of hypertension [15]. Thiazide diuretics are the recommended initial drug class for uncomplicated hypertension without significant comorbidities [15]. Many patients who have high-risk conditions (heart failure or diabetes) may benefit from the use of other antihypertensives, including angiotensin-converting enzyme inhibitors (ACE-Is) and angiotensin receptor blockers (ARBs) [15]. Two or more antihypertensives usually are required for patients who have diabetes or kidney disease [15]. If an individual’s initial systolic blood pressure is greater than 20 mm Hg above goal, or a diastolic pressure exceeds 10 mm Hg above goal, the physician should consider initial drug therapy with two agents. Of those considered, a thiazide diuretic should be one of the two initial drugs of choice, if clinically appropriate and tolerated [15,34].
Evaluation Appropriate assessment begins with proper blood pressure monitoring technique. The patient should refrain from the consumption of caffeine, soft drinks, and tobacco products for at least 30 minutes before measurement. Blood pressure should be evaluated after the patient has been seated for at least 5 minutes, preferably in a chair as opposed to an examination table. The bladder of the blood pressure cuff should encompass at least 80% of the forearm, and two measurements should be made with the average of the two representing the accepted blood pressure [15]. The initial laboratory evaluation of a person who has hypertension should include hematocrit, glucose, electrolytes, creatinine, urinalysis, a lipid profile, and an ECG [15]. An elevated C-reactive protein (CRP; a marker of inflammation) may be predictive of future coronary events in healthy
891
ESSENTIAL HYPERTENSION OR COMORBID CONDITIONS
middle-aged men [35] and in the elderly [36,37]. In addition, CRP are associated inversely with cardiorespiratory fitness levels [38]. Although CRP is not recommended by JNC 7 as a routine laboratory measurement, research to determine its role is ongoing, and it may be useful in risk stratification of hypertensive patients. Additional testing may be indicated, depending upon clinical circumstances. Recommended testing for suspected secondary causes of hypertension are listed in Table 2 [15,39]. Treatment: lifestyle modification The treatment of hypertension includes pharmacotherapy and lifestyle modification. Frequently, pharmacotherapy is emphasized, with less attention to lifestyle changes. Lifestyle modification offers a low-cost method to improve overall health. Data clearly demonstrate that lifestyle modifications, including a reduced sodium intake, weight loss, exercise, and moderating alcohol consumption, reduce systolic blood pressure by 21 to 55 mm Hg (Table 3) [9,15,40–45]. Many patients do not adhere to lifestyle modification recommendations. Common reasons include a lack of education from physicians, a lack of access to safe places to exercise, added salt in many prepared foods and restaurant meals, and the higher cost of foods that are lower in sodium and Table 2 Clinical features associated with secondary causes of hypertension Condition
Clinical findings
Diagnostic test
Renovascular hypertension
An increase in serum creatinine following use of an ACE-I; hypertension in a patient who has diffuse atherosclerosis Elevated serum creatinine, abnormal urinalysis Paroxysmal elevations in blood pressure. Triad of: headache, palpitations, and sweating. Unexplained hypokalemia
MRI
Primary renal disease Pheochromocytoma
Primary aldosteronism Cushing’s syndrome
Sleep apnea Coarctation of the aorta Hypothyroidism Primary hyperparathyroidism
Cushingoid features (Moon facies, buffalo hump), history of steroid use Daytime fatigue, snoring, reported nighttime apnea No obtainable blood pressure in the legs, decreased left brachial pulse Weight gain, hair loss, constipation Increased serum calcium
Data from Refs. [15–39].
Estimated glomerular filtration rate 24-h urine for metanephrine and normetanephrine 24-h urinary aldosterone level Dexamethasone suppression test Sleep study CT angiography
Serum thyrotropin Serum parathyroid hormone
892
WEXLER & FELDMAN
Table 3 Reduction in systolic blood pressure for various lifestyle modifications Range of systolic blood pressure reduction (mm Hg)
Lifestyle modification
Recommendation
Weight loss
Maintain a normal body weight based on BMI Diet high in fruits and vegetables, and reduced fat
5–20
Less than 2.4 g/d 30 min of aerobic activity at least 4 d/wk 2 drinks or less per day for men, and 1 drink or less per day for women
2–8 4–9
Dietary Approaches to Stop Hypertension eating plan Low sodium diet Exercise Moderate alcohol consumption
8–14
2–4
Abbreviation: BMI, body mass index. Data from Refs. [9,15,40–45]. Adapted from Chobanian A, Bakris G, Black H, et al. Seventh Report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure [JNC 7]. Hypertension 2003;42:1206–52; with permission.
calories [46]. Physicians can facilitate patient adherence by discussing healthy lifestyle principles at the initiation of treatment and by reinforcing these recommendations during subsequent follow-up visits. This approach was studied by the PREMIER collaborative research group, which used educational and informational protocols to assist physicians in helping their patients to adopt healthy lifestyle habits (in addition to pharmacologic therapy) to lower their blood pressure further [41]. Patient self-management of their chronic hypertension is realistic and feasible with potentially excellent outcomes [47–50]. Self-management techniques are highly effective in chronic disease management programs [50,51]. Risk factor modification in the treatment of hypertension is effective in the primary care setting [51–53]. The significance of these findings is germane to clinical practice for several reasons. Lifestyle modification is less expensive than is pharmacotherapy and is ideally suited to the primary care office. Finally, lifestyle modification gives the patient a sense of control over his/her disease process. Many resources are available from the American Heart Association (AHA) [54], The National Heart Lung and Blood Institute [55], The American Academy of Family Physicians [56], and Improvingchroniccare.org [57]. Resources are listed in Appendix 1. Pharmacotherapy Essential hypertension JNC 7 recommends that a thiazide-type antihypertensive be used as initial treatment for most patients who have essential hypertension [15,31,45].
ESSENTIAL HYPERTENSION OR COMORBID CONDITIONS
893
This recommendation was based primarily on the results of the Antihypertensive and Lipid Lowering Treatment to Prevent Heart Attack Trial, which compared the thiazide-type diuretic chlorthalidone with the ACE-I lisinopril, and the calcium channel blocker (CCB) amlodipine [58]. This randomized, double-blind, active-controlled clinical trial studied 33,357 subjects who had high blood pressure and at least one additional risk factor for heart disease. The study found no difference in the primary end point (combined fatal coronary disease or nonfatal MI among the three groups) among all medications tested. There was a trend toward increased risk for stroke with lisinopril, and an increased risk for heart failure with amlodipine. Based on these data, JNC 7 concluded that thiazide diuretics should be used as first-step in therapy because they are the most cost-effective [15]. The Blood Pressure Lowering Treatment Trialist’s Collaboration Trial (BPLTTC) performed a meta-analysis of 29 randomized controlled trials [59]. It found a linear relationship between better blood pressure control and a reduction in cardiovascular events. This benefit was consistent irrespective of drug class (b-blocker, diuretic, ACE-I, CCB, ARB), although the BPLTTC did note that drug class differences had unique benefits in patients who had particular underlying disease states. In the Swedish Trial in Old Patients with Hypertension-2 Study, the benefits of b-blockers, diuretics, ACE-Is, and CCBs were compared [60]. This trial demonstrated an overall reduction in cardiovascular events with all drug classes related to their ability to lower blood pressure. Patients who have concomitant diagnoses JNC 7 makes several recommendations for the use of specific antihypertensive medications based upon ‘‘compelling indications’’ (underlying disease states) [15]. These recommendations are summarized in Table 4. Heart failure The benefits of ACE-Is in the treatment of patients who have heart failure was demonstrated by the Effect of Enalapril on Survival in Patients with Reduced Left Ventricular Ejection Fractions and Congestive Heart Failure Trial [61] and the Effect of Captopril On Mortality And Morbidity In Patients with Left Ventricular Dysfunction After Myocardial Infarction: Results of the Survival and Ventricular Enlargement Trial [62]. These studies found a relative risk reduction in cardiac events when this class of medication was used in patients who had heart failure. The number of patients needed to treat with ACE-Is for 15 months to save one life was 43 [63]. b-Blockers also are indicated in most patients who have heart failure, unless a contraindication to their use exists [64]. The three that have demonstrated efficacy in this group of patients include carvedilol
894
WEXLER & FELDMAN
Table 4 Drug classes and compelling indications Recommended drug class Compelling indication Diuretic b-Blocker ACE-I ARB CCB Aldosterone antagonist
Heart failure, elevated risk for coronary disease, diabetes, secondary stroke prevention Heart failure, post-MI, elevated risk for coronary disease, diabetes Heart failure, post-MI, elevated risk for coronary disease, diabetes, kidney disease, secondary stroke prevention Heart failure, elevated risk for coronary disease, kidney disease Elevated risk for coronary disease Heart failure, post-MI
Data from Refs. [15,45,61,62,64–79,81,82]. Adapted from Chobanian A, Bakris G, Black H, et al. Seventh Report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure [JNC 7]. Hypertension 2003;42:1206–52; with permission.
[64,65], extended-release metoprolol [66], and bisoprolol [67]. Current JNC 7 and American College of Cardiology (ACC) guidelines for the treatment of heart failure [15,68] recommend that ARBs be used in patients who are intolerant to ACE-Is, although newer studies have demonstrated the efficacy of this drug class in patients who have heart failure [69,70]. Recent data and recommendations from the ACC/AHA guidelines [68] suggests that patients who have hypertension and are at risk for developing heart failure should be started on an ACE-I; patients who have structural heart disease and hypertension should be started on an ACE-I and a b-blocker. Coronary artery disease It is now standard of care to place all patients on a b-blocker after an MI unless a contraindication exists. This recommendation is based on the Beta-Blocker Heart Attack Trial that demonstrated a 16% reduction in nonfatal MI when b-blockers were used in patients who had coronary artery disease [71]. These findings were supported further by the Norwegian Multicenter Study of Timolol after Myocardial Infarction trial that demonstrated a 45% reduction in sudden cardiac death in patients who had coronary artery disease [72]. The conclusions from these studies that were completed in the early 1980s were supported more recently in the CAPRICORN Trial [65]. This trial also was the first to demonstrate a survival difference for b-blockers in patients who had left ventricular dysfunction or heart failure. Although the Prevention of Events with Angiotensin-Converting Enzyme Inhibition Trial [73] failed to demonstrate a benefit of this ACE-I, it is one of the few exceptions to the many other studies that have been published. Studies that support the use of ACE-Is in patients who have hypertension and coronary artery disease include The Trandolapril Cardiac Evaluation [74], Survival of Myocardial Infarction Long-Term Evaluation study [75], Heart Outcomes Prevention Evaluation [76], and the European Trial on
895
ESSENTIAL HYPERTENSION OR COMORBID CONDITIONS
Reduction of Cardiac Events with Perindopril in Patients with Coronary Artery disease [77]. The Comparison of Amlodipine vs. Enalapril to Limit Occurrences of Thrombosis Study found a significant reduction in cardiovascular events (reduction of hospitalization by 42%, reduction of nonfatal MI by 26%, reduction of transient ischemic attack or stroke by 50%) when the CCB amlodipine was used in patients who had hypertension and coronary artery disease and who did not have left ventricular dysfunction or heart failure [78]. Diabetes High blood pressure contributes significantly to the morbidity and mortality of individuals who have diabetes; the goal of treatment is a blood pressure that is less than 130/80 mm Hg [15,64,79]. For every 10-mm Hg reduction in systolic blood pressure, the risk for any complication related to diabetes is reduced by 12% [80]. Because of the renoprotective effect that is conferred by ACE-Is, the American Diabetes Association recommends this class of drugs be used for all diabetics who are older than 55 years and are at risk for coronary artery disease [79]. ACE-Is also reduce proteinuria and slow the decline of the glomerular filtration rate in chronic nephropathies, independently from their blood pressure lowering effect [81]. In patients who have type 2 diabetes, this same effect was seen with the use of the ARB irbesartan [82].
Table 5 Evidence-based recommendations summary Recommendation
Level of evidence
Reference
Lifestyle modifications are recommended for all patients who have hypertension Lifestyle modifications are recommended for all patients who have prehypertension Blood pressure should be treated to less than 140/90 mm Hg for all patients who have essential hypertension Blood pressure should be treated to less than 130/80 mm Hg for all patients who have concomitant chronic disease (eg, heart failure, coronary artery disease, diabetes, renal failure) Thiazide-type diuretics should be used as first-line treatment of uncomplicated essential hypertension Begin treatment with two classes of medications for patients who have a blood pressure in excess of 20 mm Hg systolic of goal or greater than 10 mm Hg diastolic of goal
C
15
C
15
A
15
A
15,68,79
B
15,58
C
15,34
896
WEXLER & FELDMAN
Summary The primary care physician should treat the patient’s underlying disease and his/her blood pressure (achieve an absolute reduction in blood pressure). This two-pronged strategy saves lives and attenuates the progression of chronic diseases. This philosophy of tailoring a patient’s medications is supported by the National Kidney Foundation, AHA, ACC, and the JNC 7 guidelines. Table 5 shows an evidence-based summary of treatment recommendations.
Appendix 1 Epocrates Hypertension Quick Reference for Personal Digital Assistants (http://www.epocrates.com) Includes: Abridged version of JNC 7 How to evaluate the hypertensive patient Lifestyle management recommendations Body mass index calculator Framingham Heart risk calculator Dietary Approaches to Stop Hypertension Diet Plan (http://www.nhlbi.nih.gov/health/public/heart/hbp/dash/) JNC 7 Quick Reference Card (http://www.nhlbi.gov/guidelines/hypertension/jnc7card.htm) Includes: Classification of hypertension Compelling indications Lifestyle recommendations Treatment algorithm Lifestyle modification tips (http://www.nhlbi.gov/hbp/treat/treat/htm) Includes: Healthy eating How to reduce salt intake How to maintain a desired weight Increasing physical activity Limiting alcohol Quitting smoking Shape Up America (http://www.shapeup.org/10000steps.html) A not-for-profit organization promoting 10,000 steps a day for better health.
ESSENTIAL HYPERTENSION OR COMORBID CONDITIONS
897
Includes instructions on how to obtain a pedometer so that patients can track their progress. Books from the American Heart Association (1-800-638-0672; available from book stores and online retailers) AHA Low Salt Cook Book ($22.95) AHA Hypertension Primer (3rd edition)
References [1] Fields L, Burt V, Cutler J, et al. The burden of adult hypertension in the United States 1999– 2000: a rising tide. Hypertension 2004;44:1–7. [2] Franklin SS, Gustin W, Wong ND, et al. Hemodynamic patterns of age-related changes in blood pressure: The Framingham Heart Study. Circulation 1997;96:308–15. [3] American Heart Association. Heart disease and stroke statisticsd2004 update. Dallas (TX): American Heart Association; 2003. [4] American Heart Association. Heart disease and stroke statisticsd2005 update. Dallas (TX): American Heart Association; 2005. [5] Beevers G, Lip G, O’Brien E. ABC of hypertension: the pathophysiology of hypertension. BMJ 2001;322:912–6. [6] Centers for Disease Control and Prevention. National Center for Health Statistics, National Health and Nutrition Examination Survey. Atlanta (GA): Centers for Disease Control and Prevention; 2003. [7] Aviv A, Hollenberg NK, Weder A. Urinary potassium excretion and sodium sensitivity in blacks. Hypertension 2004;43:707–13. [8] Wright JT, Rahman M, Scarpa A, et al. Determinants of salt sensitivity in black and white normotensive and hypertensive women. Hypertension 2003;42:1087–92. [9] Sacks F, Svetkey L, Vollmer W, et al. Effects on blood pressure of reduced dietary sodium and the dietary approaches to stop hypertension [DASH] diet. N Engl J Med 2001;344:3–10. [10] Vikrant A, Tiwari SC. Essential hypertension: pathogenesis and pathophysiology. Indian Acad Clin Med 2002;2:141–61. [11] Kupper N, Willemsen G, Riese H, et al. Heritability of daytime ambulatory blood pressure in an extended twin design. Hypertension 2005;45:80. [12] Lewington S, Clarke R, Qizilbash N, et al. Age specific relevance of usual blood pressure to vascular mortality: a meta analysis of individual data for one million adults in 61 prospective studies. Lancet 2002;360:1903–13. [13] Vasan RS, Larson MG, Leip EP, et al. Impact of high-normal blood pressure on the risk of cardiovascular disease. N Engl J Med 2001;345:1291–7. [14] Tierney W, Brunt M, Kesterson J, et al. Quantifying risk of adverse clinical events with one set of vital signs among primary care patients with hypertension. Ann Fam Med 2004;2:209–17. [15] Chobanian A, Bakris G, Black H, et al. Seventh Report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure. Hypertension 2003;42:1206–52. [16] Whelton P, He J, Appel L, et al. Primary prevention of hypertension: clinical and public health advisory from the national high blood pressure education program. JAMA 2002; 288:1882–8. [17] Lloyd-Jones D, Evans J, Larson M, et al. Treatment and control of hypertension in the community: a prospective analysis. Hypertension 2002;40:640–6. [18] McInnes G. How important is optimal blood pressure control? Clin Ther 2004;26(Suppl A): A3–11.
898
WEXLER & FELDMAN
[19] Franco V, Oparil S, Carretero O. Review: clinical cardiology: new frontiers: hypertensive Therapy Part I. Circulation 2004;109:2953–8. [20] Hajjar I, Kotchen T. Trends in prevalence, awareness, treatment, and control of hypertension in the United States, 1998–200. JAMA 2003;290:199–206. [21] Berlowitz R, Ash A, Hickey E, et al. Inadequate management of blood pressure in a hypertensive population. N Engl J Med 1998;339:1957–63. [22] Holmes JS, Shevrin M, Goldman B, et al. Translating research into practice: are physicians following evidence-based treatment guidelines in the treatment of hypertension? Med Res Rev 2004;61:453–73. [23] Hyman DJ, Pavlik VN, Vallbona C. Physician role in the lack of awareness and control of hypertension. J Clin Hypertens 2000;2:324–30. [24] Hyman DJ, Pavlik VN. Self-reported hypertension treatment practices among primary care physicians: Blood pressure thresholds, drug choices, and the role of guidelines and evidence based medicine. Arch Intern Med 2000;160:2281–6. [25] Psaty BM, Manolio TA, Smith NL, et al. Time trends in high blood pressure control and the use of antihypertensive medications in older adults: the cardiovascular health study. Arch Intern Med 2002;162:2325–32. [26] Milchak JL, Carter BL, James PA, et al. Measuring adherence to practice guidelines for the management of hypertension: an evaluation of the literature. Hypertension 2004;44: 602–8. [27] Knight EL, Glynn RJ, Levin R, et al. Failure of evidence based medicine in the treatment of hypertension in older patients. J Gen Intern Med 2000;15:702–9. [28] Troein M, Gardell B, Selander S, et al. Guidelines and reported practice for the treatment of hypertension and hypercholesterolemia. J Gen Intern Med 1997;242:173–8. [29] Mehta SS, Wilcox CS, Schulman KA. Treatment of hypertension in patients with comorbidities: results from the study of hypertension prescribing practices. Am J Hypertens 1999;12: 333–40. [30] Cuspidi C, Michev I, Lonati L, et al. Compliance to hypertension guidelines in clinical practice: a multicentre pilot study in Italy. J Hum Hypertens 2002;16:699–703. [31] Chobanian A, Bakris G, Black H, et al. The Seventh Report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure: the JNC 7 report. JAMA 2003;289(19):2560–72 [erratum appears in JAMA 2003;290(2):197]. [32] Franklin SS, Laeson MG, Khan SA, et al. Does the relation of blood pressure to coronary heart disease risk change with aging? The Framingham Heart Study. Circulation 2001;103: 1245–9. [33] Vasan RS, Beiser A, Seshadri S, et al. Residual lifetime risk for developing hypertension in middle-aged women and men: The Framingham Heart Study. JAMA 2002;287: 1003–10. [34] Sica DA. Rationale for fixed-dose combinations in the treatment of hypertension: the cycle repeats. Drugs 2002;62:443–62. [35] Koenig W, Sund M, Frohlich M, et al. C-reactive protein, a sensitive marker of inflammation, predicts future risk of coronary heart disease in initially healthy middle aged men: Results from the MONICA [Monitoring Trends and Determinants in Cardiovascular Disease] Augsburg cohort study, 1984–1992. Circulation 1999;99:237–42. [36] Strandberg T, Tilvis RS. C-reactive protein, cardiovascular risk factors, and mortality in a prospective study in the elderly. Arterioscler Thromb Vasc Biol 2000;20:1057–60. [37] Tracy RP, Lemaitre RN, Psaty BM, et al. Relationship of C-reactive protein to risk of cardiovascular disease in the elderly: results from the cardiovascular health study and the rural health promotion project. Arterioscler Thromb Vasc Biol 1997;17:1121–7. [38] Church TS, Barlow CE, Earnest CP, et al. Associations between cardiorespiratory fitness and C-reactive protein in men. Arterioscler Thromb Vasc Biol 2002;22:1869–76. [39] Kaplan NM. Initial evaluation of the hypertensive patient. UpToDate Online 13.1, April 20, 2004, Available at: http://www.utdol.com. Accessed May 6, 2005.
ESSENTIAL HYPERTENSION OR COMORBID CONDITIONS
899
[40] Campbell N. A brief overview of 2004 recommendations of the Canadian Hypertension Education Program. Can Fam Phys 2004;50:1411–5. [41] PREMIER Collaborative Research Writing Group. Effects of comprehensive lifestyle modification on blood pressure control. JAMA 2003;289:2083–93. [42] Appel L, Moore T, Obarzanek E, et al. A clinical trial of the effects of dietary patterns on blood pressure. N Engl J Med 1997;336:1117–24. [43] Vollmer W, Sacks F, Ard J, et al. Effects of diet and sodium intake on blood pressure: subgroup analysis of the DASH-Sodium Trial. Ann Intern Med 2001;135:1019–28. [44] Paul J, Whelton P, Appel L, et al. Long-term effects of weight loss and dietary sodium reduction on incidence of hypertension. Hypertension 2000;35:544–9. [45] Black HR. JNC 7 & treatment guidelines: goals and recommendations. Cardiol Rev 2004;21: 37–44. [46] US Department of Health and Human services. Public Health Service, Centers for Disease Control and Prevention, National Center for Health Statistics, 2002. [47] Bodenheimer T, Wagner E, Grumbach K. Improving primary care for patients with chronic disease. JAMA 2002;288:1775–9. [48] Bodenheimer T, Lorig K, Holman H, et al. Patient self management of chronic disease in primary care. JAMA 2002;288:2469–75. [49] Bodenheimer T, Wagner E, Grumbach K. Improving primary care for patients with chronic illness: the chronic care model, Part 2. JAMA 2002;288:1909–14. [50] Warsi A, Wang P, LaValley M, et al. Self-management education programs in chronic disease: a systemic review and methodological critique of the literature. Arch Intern Med 2004; 164:1641–9. [51] Mattila R, Malmivaara A, Kastarinen M, et al. Effectiveness of multidisciplinary lifestyle intervention for hypertension: a randomized controlled trial. J Hum Hypertens 2003;17: 199–205. [52] Korhonen M, Kastarinen M, Uusitupa M, et al. The effect of intensified diet counseling on the diet of hypertensive subjects in primary health care: A 2-year open randomized controlled trial of lifestyle intervention against hypertension in eastern Finland. Prev Med 2003;36: 8–16. [53] Kastarinen M, Puska P, Korhonen M, et al. Non-pharmacological treatment of hypertension in primary health care: a 2-year open randomized controlled trial of lifestyle intervention against hypertension in eastern Finland. J Hypertens 2002;20:2505–12. [54] The American Heart Association. Available at: http://www.americanheart.org. Accessed June 2006. [55] The National Heart Lung and Blood Institute. Available at: http://www.nhlbi.nih.gov/ index.htm. Accessed June 2006. [56] The American Academy of Family Physicians. Available at: http://www.familydoctor.org. Accessed June 2006. [57] Available at: http://Improvingchroniccare.org. Accessed June 2006. [58] The ALLHAT Officers and Coordinators for the ALLHAT Collaborative Research Group. Major outcomes in high-risk hypertensive patients randomized to angiotensinconverting enzyme inhibitor or calcium channel blocker vs. diuretic: the antihypertensive and lipid-lowering treatment to prevent heart attack trial [ALLHAT]. JAMA 2002;288: 2981–97. [59] Blood Pressure Lowering Treatment Trialist’s Collaboration. Effects of different blood-pressure-lowering regimens on major cardiovascular events: results of prospectively-designed overviews of randomized trials. Lancet 2003;362:1527–35. [60] Hansson L, Lindholm L, Ekbom T, et al. Randomised trial of old and new antihypertensive drugs in elderly patients: cardiovascular mortality and morbidity the Swedish Trial in Old Patients with hypertension-2 study. Lancet 1999;354:1751–6. [61] The SOLVD Investigators. Effect of enalapril on survival in patients with reduced left ventricular ejection fractions and congestive heart failure. N Engl J Med 1991;325:293–302.
900
WEXLER & FELDMAN
[62] The SAVE Investigators. Effect of captopril on mortality and morbidity in patients with left ventricular dysfunction after myocardial infarction. Results of the survival and ventricular enlargement trial. N Engl J Med 1992;327:669–77. [63] Flather MD, Yusuf S, Kobler L, et al. Long-term ACE inhibitor therapy in patients with heart failure or left ventricular dysfunction: a systematic overview of data from individual patients. Lancet 2000;355:1575–81. [64] The CAPRICORN Investigators. Effect of carvedilol on outcome after myocardial infarction in patients with left-ventricular dysfunction: the CAPRICORN randomized trial. Lancet 2001;357:1385–90. [65] Packer M, Coats A, Fowler MB, et al. Effect of carvedilol on survival in severe chronic heart failure. N Engl J Med 2001;344:1651–8. [66] Fagerberg B, et al, for the MERIT-HF Study Group. Effect of metoprolol CR/XL in chronic heart failure: Metoprolol CR/XL randomized intervention trial in congestive heart failure [MERIT-HF]. Lancet 1999;353:2001–7. [67] CIBIS II Investigators. The Cardiac Insufficiency Bisoprolol Study II [CIBIS-II]: a randomised trial. Lancet 1999;353:9–13. [68] Hunt SA, Abraham WT, Chin M, et al. ACC/AHA 2005 Guideline Update for the Diagnosis and Management of Chronic Heart failure in the Adult: A report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines (Writing Committee to Update the 2001 Guidelines for the Evaluation and Management of Heart Failure). J Am Coll Cardiol 2005;46:1–82. [69] Cohn JN, Tognoni G, et al. A randomized trial of the angiotensin receptor blocker valsartan in chronic heart failure. N Engl J Med 2001;345:167–75. [70] Young JB, Dunlap ME, Pfeffer MA, et al. Mortality and morbidity reduction with candesartan in patients with chronic heart failure and left ventricular systolic dysfunction: results of the CHARM low-left ventricular ejection fraction trials. Circulation 2004;110: 2618–26. [71] Beta-Blocker Heart Attack Research Group. A randomized trial of propranalol in patients with acute myocardial infarction, I: mortality results. JAMA 1982;247:1707–14. [72] Timolol-induced reduction in mortality and reinfarction in patients surviving acute myocardial infarction. N Engl J Med 1981;304:801–7. [73] The PEACE Trial Investigators. Angiotensin-converting-enzyme inhibition in stable coronary artery disease. New Engl J Med 2004;351:2058–68. [74] Kober L, Torp-Pedersen C, Carlsen JE, et al, for the TRACE study group. A clinical trial of the angiotensin-converting-enzyme inhibitor trandolapril in patients with left ventricular dysfunction after myocardial dysfunction. New Engl J Med 1995;333:1670–6. [75] Ambrosioni E, Borghi C, Magnani B, for the SMILE Study Group. The effect of the angiotensin-converting-enzyme inhibitor zofenopril on mortality and morbidity after myocardial infarction. N Engl J Med 1995;332:80–5. [76] The Heart Outcomes Prevention Evaluation Study Investigators. Effects of an angiotensinconverting-enzyme inhibitor, ramipril, on cardiovascular events in high risk patients. N Engl J Med 2000;342:145–53. [77] The EURopean trial On reduction of cardiac events with Perindopril in stable coronary Artery disease Investigators. Efficacy of perindopril in reduction of cardiovascular events among patients with stable coronary artery disease; randomized, double-blind, placebo-controlled, multicentre trial [the EUROPA study]. Lancet 2003;362:782–8. [78] Nissen SE, Tuzcu EM, Libby P, et al. Effect of antihypertensive agents on cardiovascular events in patients with coronary disease and normal blood pressure: The CAMELOT Study: a randomized controlled trial. JAMA 2004;292:2217–26. [79] American Diabetes Association. Treatment of hypertension in adults with diabetes. Diabetes Care 2003;26:S80–2. [80] National Diabetes Information Clearing House. Available at: http://diabetes.niddk.nih.gov/ dm/pubs/statistics/#7. Accessed February 18, 2005.
ESSENTIAL HYPERTENSION OR COMORBID CONDITIONS
901
[81] The Gisen Group. Randomised placebo controlled trial of effect of ramipril on decline in glomerular filtration rate and risk of terminal renal failure in proteinuric, non-diabetic nephropathy. Lancet 1997;349:1857–63. [82] Lewis EJ, Hunsiker LG, Clarke WR, et al. Renoprotective effect of the angiotensin receptor antagonist irbesartan in patients with nephropathy due to type 2 diabetes. N Engl J Med 2001;335:851–60.
Prim Care Clin Office Pract 33 (2006) 903–921
Common Questions in Managing Hyperlipidemia Irene M. Rosen, MD*, Richard W. Sams II, MD Department of Family Medicine, Madigan Army Medical Center, 9040 Fitzsimmons Drive, Tacoma, WA 98431, USA
Hyperlipidemia is the seventh most common diagnosis that primary care physicians encounter. In the 45- to 64-year-old age group, it climbs to the fourth most common [1]. The National Cholesterol Education Program’s (NCEP) Coordinating Committee recently released an update to the 2001 Adult Treatment Panel (ATP) III guideline on the detection, evaluation, and treatment of hyperlipidemia [2]. Most significantly, recommendations were made for lowering the treatment goal of low-density lipoprotein (LDL) cholesterol for patients who are at highest risk for coronary heart disease (CHD); a therapeutic option of lowering the LDL to less than 70 mg/dL in persons who are at highest risk for cardiovascular events was suggested. Subsequent investigators have been asking if we should ‘‘treat to new targets.’’ This and other recommendations by the report, as well as the results of more recent trials, have generated several questions and a good bit of controversy [3]. Clinicians are confronted with several practical questions when approaching the diagnosis and management of hyperlipidemia: Who should be treated? How significant is the impact of treatment? What are the treatment goals for LDL? What are the possible adverse effects of treatment? Should elevated triglycerides or low high-density lipoprotein (HDL) cholesterol be treated? Is it safe and effective to combine dyslipidemics? What other therapies besides statins have proven to be effective in improving patient outcomes? What laboratory values should be checked and how often? The answers to these common questions are difficult to find given the multitude of studies and their conflicting results. Additionally, results from trials on hyperlipidemia often are exaggerated by emphasizing relative * Corresponding author. E-mail address:
[email protected] (I.M. Rosen). 0095-4543/06/$ - see front matter Ó 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.pop.2006.09.007 primarycare.theclinics.com
904
ROSEN & SAMS
risk reductions. Succinct evidence-based noncommercial answers to the above questions are needed for the busy clinician. This article attempts to answer the questions posed. An emphasis is placed on patient-oriented evidence that matters, to address outcomes that are relevant to patients and clinicians. A brief evidence-based bottom line provides the clinician with the essential information to answer the question. This is followed by a brief review of the evidence that justifies the answer. The strength of recommendation (SOR) is used in grading the evidence [4]. Each question is concluded with the authors’ interpretation of the evidence, practice patterns, and style. For a summary of recommendations, see Table 1. Which patients who have high low-density lipoproteins should be treated with medications? Bottom line Elevated LDL is associated clearly with an increased risk for cardiovascular disease. Increases in LDL pose an incremental increase in the risk for CHD (SOR: A, based on well-designed consistent cohort studies). Recommendations on when to initiate medical therapy given the LDL level are dependent on other risk factors for CHD, and vary by organization. The NCEP ATP III guideline is the most widely used set of recommendations (SOR: C, based on expert opinion) [5]. Table 1 Strength of recommendations for managing dyslipidemia Key clinical recommendation
Strength of recommendation
Patients in the highest risk category for CHD should be on a statin if tolerated, regardless of their baseline LDL Patients with a !10% 10-year risk for CHD do not need statin therapy for primary prevention Liver function testing should be followed regularly in patients who are on statin monotherapy or combination therapy A diet containing omega-3 polyunsaturated fatty acids reduces the risk for cardiovascular events A Mediterranean diet decreases the incidence of recurrent cardiovascular events in patients who have CHD In patients who do not have CHD, but who have low HDL or elevated triglycerides and a family history of premature CHD, consider adding a fibrate or niacin if goals are not met with lifestyle changes In patients who have a family history of CHD or continue to smoke, consider a statin, even if their LDL is !130 mg/dL
A
A A
A B
C
C
MANAGING HYPERLIPIDEMIA
905
Initiating pharmacologic treatment to decrease the LDL for any given patient is a decision that is based on the level of risk for CHD and expected risk reduction from treatment. Patients who are at highest risk tend to benefit most from LDL reduction. Evidence summary Various organizations have put forth treatment guidelines, and some of their recommendations, risk estimates, and treatment approaches differ. The ATP III guideline is the most widely accepted set of recommendations. The American Heart Association, the National Heart, Lung and Blood Institute, and the American College of Cardiology Foundation are three major organizations that have adopted them. The ATP III follows a step-based approach to individual risk assessment. A patient’s 10-year risk for CHD is determined, and the patient’s risk category determines treatment goals for LDL. Patients are categorized as high risk (O20% risk for CHD in 10 years), moderate risk (10%–20%), or low risk (!10%). Web-based and hand-held–based ATP III calculators exist, that are quick, easy to use and effective at calculating a patient’s risk category [6]. Quick reference guides that delineate the steps also are readily available [5]. Box 1 contains a brief summary of the steps. There is no cholesterol level at which a patient has zero risk for CHD. Recommendations for cut points for treatment are continuing to evolve based on results from randomized controlled trials (RCTs). Most treatment trials in the last 15 years have been with HMG CoA reductase inhibitors (statins). Cardiovascular benefit of cholesterol lowering with statins has been demonstrated in patients who have CHD and who do or do not have hyperlipidemia. Predictably, the benefit is highest for those who have CHD plus hyperlipidemia, although recent trials demonstrated some benefit of statin therapy in patients who had CHD and what was considered average or normal cholesterol levels [5]. A large RCT showed that patients who had CHD or CHD equivalents and who were older than 65 years of age had the greatest benefit. Other trials, however, failed to show a greater benefit in the over-65 population [8]. Clearly, in the elderly population, it is appropriate to consider the chronologic and the physiologic age. Patients with a limited life expectancy because of comorbid illnesses may not be candidates for therapy, but highly functioning patients should not be denied treatment based on age alone. Limited data exist on primary prevention in the elderly, with no clear morbidity or mortality benefit. Because more than 50% of elderly patients die of cardiovascular disease, some investigators recommend primary prevention in those with two or more cardiac risk factors, and an LDL of greater than 160 mg/dL [9]. Diabetic patients also show a clear benefit from cholesterol reduction. More than half of all diabetics die of cardiovascular causes. Decreasing their
906
ROSEN & SAMS
Box 1. Step-based approach to individual risk assessment 1. Obtain a fasting lipid profile, which stratifies total cholesterol, LDL, HDL, and triglyceride levels. 2. Identify CHD ‘‘risk equivalents’’ (ie, those risk factors that place patients at the same risk for CHD as that of established CHD) [5]. Atherosclerotic disease (peripheral disease, abdominal aortic aneurysm, symptomatic coronary artery disease) Diabetes mellitus Multiple risk factors that confer a combined 10-year risk for CHD of more than 20% 3. Identify major CHD risk factors, except for LDL [5]. Age (>45 years for men, >55 years for women) Hypertension (>140/90 mm Hg, or on medication) Low HDL (<40 mg/dL). HDL greater than 60 mg/dL removes one risk factor from the total count. Family history of premature CHD (first-degree male relative <55 years of age, first-degree female relative <65 years of age) Cigarette smoking 4. For patients with two or more risk factors (other than LDL), calculate the 10-year CHD risk using the Framingham risk tables. This risk category determines LDL goal of therapy, and the need for dietary, physical activity, and pharmacologic interventions. If patients exceed established cut points given their level of risk, drug therapy is recommended to achieve the LDL goal. As a caveat, be aware that the Framingham risk tables may overestimate the CHD risk in some populations, particularly Japanese and Hispanic men and Native American women [7]
cholesterol decreases the risk for death, as well as the incidence of individual cardiac events. Guidelines on cholesterol lowering in primary prevention are less clear. Although certain subgroups that have hyperlipidemia and additional risk factors can benefit, treatment for patients who have average cholesterol may not be cost-effective. Clinical commentary Patients who are at highest risk for CHD or recurrent cardiovascular events will benefit most from pharmacologic therapy. The higher the
MANAGING HYPERLIPIDEMIA
907
baseline risk, the greater the absolute risk reduction and potential benefit for the patient (see later discussion). The patient’s baseline risk is the key to clinical decision making, and it is worth the clinician’s time to do a comprehensive risk determination.
What are the treatment goals for low-density lipoprotein? Bottom line Patients who are in the highest risk category (Box 2)dthose who have established CHD, CHD risk equivalents, and those with a greater than 20% 10-year risk of CHD based on the ATP III guidelinedshould be on a statin (if tolerated), regardless of their baseline LDL (SOR: A, RCTs with consistent findings). The benefit of pursuing a target LDL of less than 70 mg/dL (instead of the previous ATP III goal of less than 100 mg/dL) in such patients is marginal and may increase the risk for side effects and discontinuation rates (SOR: B, RCTs with inconsistent findings). The goal should be tailored to specific patient preferences and risk profiles. The LDL goal for most patients who are at moderate risk for CHDdthose with a 10% to 20% 10-year riskdshould remain at less than 130 mg/dL pending results from ongoing trials. The absolute risk reduction for such patients with a family history of CHD or who continue to smoke may justify initiating a statin if the LDL is already less than 130 mg/dL (SOR: C, expert opinion). The ATP III goals for patients with a less than 10% 10-year risk for CHD should remain in place pending further studies (SOR: A, consistent RCTs). Evidence summary The benefit of more intensive lipid therapy, which was recommended by the recent NCEP report for patients who have known CHD, is considered modest by some investigators, and significant by others. The Heart Protection Study (HPS) was one of the primary trials that prompted the change in the recommendations [10]. The HPS followed more than 20,000 patients who were at high risk for CHD for 5 years. The patients were randomized to simvastatin, 40 mg, or placebo. During the study, 12.9% of the patients
Box 2. Risk categories and corresponding LDL treatment goals CHD and CHD equivalent (goal: <100 mg/dL [5]; revised goal: <70 mg/dL [2]) Multiple (2+) risk factors (goal: <130 mg/dL [5]; revised goal: <100 mg/dL [2]) 0 to 1 risk factors (goal: <160 mg/dL [5])
908
ROSEN & SAMS
who were on simvastatin died compared with 14.7% of the patients who were on placebo (number needed to treat [NNT] ¼ 56); simvastatin conferred an overall mortality benefit. This benefit was derived from a decrease in cardiovascular mortality (7.6% versus 9.1%; NNT ¼ 67), which included reductions in fatal myocardial infarctions (MIs) and strokes. There also was a reduction in coronary and noncoronary revascularizations (9.1% versus 11.7%; NNT ¼ 39). Any major vascular event decreased from 25.2% to 19.8% (NNT ¼ 19). Notably, these reductions occurred over a range of cholesterol values, including levels of less than 77 mg/dL. There was a consistent relative risk reduction of 21% to 25% for most cardiovascular events, regardless of the patients’ starting cholesterol or baseline risk. Adverse events were rare and there were no differences between the two arms of the trial, although many patients who had adverse effects may have been weeded out in the ‘‘run-in’’ phase of the study, where more than 11,000 patients were excluded from the subsequent randomization phase. The recently released TNT (‘‘Treating to New Targets’’) study enrolled 10,003 patients who had a known LDL of less than 130 mg/dL. They were selected from a larger cohort that initially was treated with a run-in dose of atorvastatin to lower their LDL. Patients were randomized to atorvastatin, 10 mg or 80 mg, and followed for an average of 5 years [11]. Higher-dose atorvastatin lowered LDL to an average of 77 mg/dL (versus 101 mg/dL with 10 mg). The composite cardiovascular events were fewer in the arm that took 80 mg (8.7% versus 10.9%; NNT ¼ 45), primarily because of fewer nonfatal MIs. There was no difference in all-cause mortality (5.6% versus 5.7%). Adverse events were more common in the group that took 80 mg (8.1% versus 5.8%; number needed to harm [NNH] ¼ 43), as were discontinuation rates (7.2% versus 5.3%; NNH ¼ 52). The Anglo-Scandinavian Cardiac Outcomes Trial–Lipid Lowering Arm (ASCOTT-LLA) was a primary prevention study that was used to advocate more intensive treatment for patients who were at moderate risk for CHD [12]. More than 19,000 patients with at least three cardiovascular risk factors and an average cholesterol of 135 mg/dL were randomized to receive atorvastatin, 10 mg, or placebo. The study was halted early because of its positive results. The primary outcome, nonfatal MI and fatal CHD, was less common in the treatment arm (1.9% versus 3.0%; NNT ¼ 94). Notably, subgroup analyses showed that women did not benefit from treatment, and there was no improvement in all-cause mortality. The PROVE IT study was a 2-year trial that compared standarddose pravastatin, 40 mg (n ¼ 2063) with high-dose atorvastatin, 80 mg (n ¼ 2099) started after a patient was diagnosed with acute coronary syndrome [13]. Patients who took atorvastatin had a lower mean cholesterol and lower cardiac combined end point (NNT ¼ 25 to prevent one cardiovascular event). Notably, however, 72% of the patients had starting LDL levels of less than 125 mg/dL; the benefit of higher-dose treatment was not statistically significant in this large subgroup.
MANAGING HYPERLIPIDEMIA
909
Clinical commentary These and other less significant trials reveal some consistent trends. For a given mg/dL change in LDL, the change in relative risk for CHD is the same whether starting at higher or lower LDL levels. This is true for patients who previously were considered within goal, based on ATP III. Additionally, when a patient has a higher baseline risk for CHD secondary to other risk factors, the likelihood for benefit from medical therapy is greater than for a person who has a lower baseline risk, because of the greater change in absolute risk [2]. Several caveats should cause the clinician to pause before deciding to lower a patient’s LDL goal to the new suggested levels. First, patients in the cited trials may not be representative of the typical primary care patient. In the HPS, the study with the most robust data, a run-in phase of the study selected out patients who experienced adverse effects or were noncompliant. Only patients whose LDL was ‘‘responsive’’ to statins were eligible. Direct LDL levels were measured, which are 15% higher than calculated LDL levelsdthe method that most laboratories continue to use. Additionally, a large percentage of the population pays for its own medicines out of pocket. A month’s supply of the higher-dose statins can cost hundreds of dollars, which can make them prohibitively expensive. From a public health perspective, it is unclear if a 2% to 4% absolute risk reduction in cardiovascular events justifies higher doses and costs to the individual and to society. What does all of this mean for our patients who have established CHD or who are at high risk for CHD? Again, the higher the baseline risk, the more likely the patient is to benefit from more intensive therapy. Treatment needs to be individualized based on the patient’s baseline risk, ability to tolerate medical therapy, economic situation, and motivation for treatment. Motivated patients who are at high risk and who can acquire high-dose statins, tolerate them, and are ‘‘LDL responsive’’ are likely to achieve a modest decrease in risk. For patients who are at low to moderate risk, a shared decision-making process should guide treatment decisions.
Should elevated triglycerides or low high-density lipoproteins be treated medically? Bottom line In patients who have CHD or diabetes and a low HDL or elevated triglycerides, consider adding gemfibrozil or niacin to decrease the risk for cardiovascular events further (SOR: B, based on a single RCT for each medicine). In patients who do not have CHD, but who have elevated triglycerides or low HDL and a family history of premature CHD, consider treating with a fibrate or niacin if goals are not met with therapeutic lifestyle changes (SOR: C, expert opinion).
910
ROSEN & SAMS
Evidence summary There is strong evidence that low HDL is a powerful risk factor for CHD and that high HDL protects against the disease [5]. It is less certain if elevated triglycerides definitively increase the risk for CHD independent of other factors [14–16]. Elevated triglycerides and low HDL often occur together in the setting of several potential contributing factors [5]: Overweight and obesity Physical inactivity Cigarette smoking Type II diabetes b-Blockers, anabolic steroids, progestins Hereditary factors Very high carbohydrate intake In the final risk assessment steps of the ATP III, the investigators provide recommendations on goals and treatment of elevated triglycerides and low HDL. Unless triglycerides are elevated markedly (O500 mg/dL), which puts the patient at risk for pancreatitis or chylomicronemia syndrome, they are treated indirectly by targeting the non-HDL cholesterol, based on the rationale that triglyceride-rich lipoproteins, particularly very lowdensity lipoproteins, are atherogenic. Identify and treat metabolic syndrome. This includes treating obesity, physical inactivity, hypertension, elevated triglycerides, and low HDL. Treat elevated triglycerides. Treat initially with therapeutic lifestyle changes [5]. If triglycerides are greater than 200 mg/dL, despite treating LDL and initiating therapeutic lifestyle changes, LDL therapy with a statin can be intensified. If this does not decrease the non-HDL cholesterol sufficiently, a fibrate or nicotinic acid should be added. If the above goals are met and HDL cholesterol is still less than 40 mg/dL and the patient has CHD or a CHD equivalent, a fibrate or nicotinic acid should be considered. Despite the mounting epidemiologic evidence that associates elevated triglycerides and low HDL with CHD, only a few trials definitively demonstrated that lowering triglycerides and raising HDL improves patientoriented outcomes. The VA-HIT (Veterans Affairs High-density Lipoprotein Cholesterol Intervention Trial) study showed a lower incidence of cardiovascular events when elevated triglycerides were lowered with gemfibrozil (NNT ¼ 23) in men who had CHD [17]. The Coronary Drug Project, one of the first major cardiovascular placebo-controlled trials, is the only study to investigate niacin’s ability as a monotherapy to affect cardiovascular end points [18]. It showed decreased incidences of nonfatal MIs (from 13.9% to 10.1%; NNT ¼ 26; P!.001) and cerebrovascular events (from 11.3% to 8.4%;
MANAGING HYPERLIPIDEMIA
911
NNT ¼ 34; P!.001). All-cause mortality and cardiovascular mortality were not affected. Clinical commentary Some trials offers evidence that treating elevated triglycerides and a low HDL improves patient-oriented outcomes. Given the relative paucity of data to support this intervention, significant collaboration with the patient is necessary. In appropriate patients who have elevated triglycerides or a low HDL, using gemfibrozil is a good first choice given the evidence, side effect profile (it usually is tolerated well), and lower cost. Niacin is more effective at elevating HDL, but is more complicated to use. Immediate-release niacin causes flushing in as many as 92% of patients, and also can cause gastrointestinal symptoms, pruritis, and an increased risk for gout and liver injury at higher doses. Some of the flushing can be alleviated by premedicating with aspirin (325 mg 30 minutes before taking niacin) or by using the new extended-release preparation Niaspan, which causes less flushing, but can cost in excess of $200 per month at higher dosages [19]. How significant is the impact of medications on patient outcomes? Bottom line Statins modestly decrease all-cause mortality and cardiovascular mortality in patients who have established CHD and in those who are at higher risk for CHD (SOR: A, based on several meta-analyses). Evidence summary All-cause and cardiovascular mortality Several meta-analyses of RCTs have found a significant decrease in allcause mortality with statin treatment. One meta-analysis of 97 RCTs with more than 275,000 enrolled patients showed a risk ratio of 0.87 for the group that took statins versus the control group (95% CI, 0.81–0.94), with an NNT of 50 (95% CI, 38–87) over 4.4 years to prevent one death in patients who had CHD, and an NNT of 228 (95% CI, 123–2958) to prevent an additional death in patients who did not have CHD [8]. The same study also showed a decrease in overall mortality with omega-3 fatty acids (fish oil and linoleic acid) in patients who had CHD (NNT ¼ 44; 95% CI, 31–84 over 4.4 years), but not for patients who did not have CHD. The study did not show any decrease in all-cause mortality with niacin or fibrates. Another meta-analysis of 14 RCTs with more than 90,000 enrolled patients supported a decrease in allcause mortality with statin treatment (NNT ¼ 86; 95% CI, 36–49) and a proportional 12% risk reduction in all-cause mortality for each mmol/L
912
ROSEN & SAMS
reduction in LDL [20]. The same study showed a decrease in cardiovascular deaths (NNT ¼ 40; 95% CI, 36–49); the benefit was seen regardless of age or sex, but predictably, the highest-risk patients derived the greatest benefit. A third meta-analysis of 17 RCTs with more than 21,000 enrolled patients also showed a significant reduction in all-cause mortality with statin treatment (odds ratio [OR], 0.76; 95% CI, 0.67–0.86), as well as a decrease in nonfatal heart attacks [21]. As in the previous studies, the benefit for individual patients varied by their baseline risk, with greater risk conferring greater potential benefits. A fourth study, an RCT that specifically looked at potential statin benefits in patients who had normal cholesterol levels and no known CHD (primary prevention), showed a decreased risk for a first major coronary event, but no change in all-cause mortality [22]. This was corroborated by a meta-analysis of 4 RCTs with 10,000 enrolled patients that also looked at cholesterol lowering in primary prevention. There was a decrease in cardiovascular mortality (OR, 0.71; 95% CI, 0.56–0.91; NNT ¼ 238 for 5 years) but no change in all-cause mortality [23]. Stroke A meta-analysis of 65 RCTs with a total of 200,000 enrolled patients showed a decrease in fatal and nonfatal stroke risk in patients who did and did not have CHD, with high-risk patients benefiting more (NNT ¼ 617; 95% CI; 463–1111) than did low-risk ones (NNT ¼ 2778; 95% CI; 2083–5000) [24]. A second meta-analysis of 5 RCTs with a total of 1700 enrolled patients showed no evidence of benefit or harm in lowering cholesterol in patients who had a history of cerebrovascular disease only (OR, 0.96; 95% CI, 0.71–1.30) [25]. Finally, an RCT with 9000 enrolled patients showed a small improvement in stroke prevention with statin treatment in high-risk patients, with an absolute risk reduction (ARR) of 0.8% in the risk for ischemic stroke (NNT ¼ 128 over 6 years), but no significant difference in the risk for hemorrhagic stroke [26]. This was corroborated by a second RCT with 20,000 enrolled patients that looked at stroke prevention in a high risk population. It showed an ARR with statin treatment of 1.4% (NNT ¼ 72 over 4.3 years) for risk for ischemic stroke, and once again no change in the risk for hemorrhagic stroke [27]. Diabetes mellitus A decision analysis using the results of the Scandinavian Simvastatin Survival Study (average LDL reduction of 35 points, and an average HDL increase of 8 points), and applying these to the Cardiovascular Disease Life Expectancy Model estimated that 25.4 million person-years of life would be saved with lipid control in diabetics as opposed to 16 million person-years of life saved in persons who have CHD. This translates to 3 to 3.4 years of life in the average patient who has diabetes mellitus versus 2.4 to 2.7 years of life in the average patient who has CHD [28]. A meta-analysis of 19 RCTs supported this conclusion; it showed a larger effect on cardiovascular
MANAGING HYPERLIPIDEMIA
913
outcomes in diabetics with cholesterol lowering than with blood glucose control. The study results showed an NNT of 106 (95% CI, 62–366) to prevent one cardiac event with cholesterol lowering, whereas glucose lowering alone in patients who did not have CHD did not show a statistically significant decrease in cardiac events [29]. Osteoporotic fractures One case-control study of more than 80,000 enrolled patients showed no association between statin use and a reduced risk for osteoporotic fractures (OR, 1.01; 95% CI, 0.88–1.16). This contrasts with the results of previous studies that did show some minor benefits of statin use in osteoporosis. At this point, more prospective randomized trials are needed to answer this question definitively [30]. Clinical commentary The data in the previous sections paint a consistent picture for approaching patients in the management hyperlipidemia: the higher the risk, and the higher the cholesterol level (and the lower the HDL), the greater the benefit. Determine the patient’s baseline risk, and collaborate with him or her in making treatment decisions.
Which drug combination therapies positively affect patient outcomes in hyperlipidemia? Bottom line No combination therapies have been compared with monotherapy in head to head trials. Adding niacin, a fibrate, or ezetimibe to a statin further reduces the LDLdand in some casesdincreases the HDL (SOR: A, based on multiple RCTs). Evidence summary In patients who cannot achieve their treatment goals on statin monotherapy, it is possible to combine a statin with niacin, which causes an additional 10% to 15% LDL reduction over statins alone, along with a 15% to 35% elevation in HDL. An RCT of 160 patients who had known coronary artery disease, low HDL, and low LDL showed that the combination of niacin and simvastatin resulted in a 0.4% regression in coronary plaque and one fewer clinical coronary event, whereas patients who were on antioxidants, placebo, or a simvastatin/niacin/antioxidant combination had an increase in mean percent stenosis as well as more clinical coronary events [31]. Finally, a meta-analysis of 9 RCTs that evaluated combination treatment with niacin and a statin showed a 25% to 57% reduction in LDL and
914
ROSEN & SAMS
a concomitant 13% to 36% increase in HDL [32]. Other studies have shown that the combination of LDL-lowering statins and HDL-raising niacin is a good option for patients who have complicated dyslipidemias, with a favorable impact on the lipid profile and a reduction in cardiovascular events [33]. The combination of statins and bile-acid sequestrants was shown to achieve an additional 10% to 20% reduction in LDL over statins alone, with gastrointestinal side effects being a possible deterrent to patient compliance. Bile-acid sequestrants can be a good stand-alone choice for women in whom pregnancy is a possibility. Statins and fibrates combine to achieve a modest additional reduction in LDL, but can increase HDL 10% to 35%, and reduce triglycerides 20% to 50%. Although there are some case reports of myopathy and rhabdomyolysis with this drug combination, the extent of the problem has not been quantified well [34]. In a meta-analysis of 36 clinical trials, the incidence of myopathy with a statin-fibrate combination was not significantly higher than with statins alone [35]. Ezetimibe is a medication that blocks cholesterol absorption in the intestine; it is used most often in combination with statins to avoid high-dose statins that can result in adverse drug reactions. For example, a combination of atorvastatin, 10 mg, and ezetimibe, 10 mg, achieves similar results to atorvastatin, 80 mg. The combination can result in an additional LDL decrease of 9% to 15%, and an additional HDL increase of 1% to 5%. At this point, there is little information on patient-oriented outcomes with a statin/ezetimibe combination [36]. Clinical commentary There are several second-line agents that can be combined with statins safely and effectively to help a patient achieve treatment goals. A patient’s treatment goal, and, thus, his or her appropriate choice of treatment, must be determined on an individual basis, with close clinician/patient collaboration. What is the potential harm of dyslipidemia treatment? Bottom line In low to medium doses of statins, side effects are low and risks are minimal. High doses may increase the rate of side effects and discontinuation of therapy. Generally, combination therapy of statins with niacin or a fibrate is tolerated well, but necessitate closer monitoring of creatine kinase (CK) levels. Liver injury is a rare event, but liver function tests (LFTs) should be followed (see later discussion) (SOR: A, based on meta-analyses and RCTs).
MANAGING HYPERLIPIDEMIA
915
Evidence summary One of the most important physician mandates is ‘‘first, do no harm.’’ Because no drug therapy is completely without potential side effects, we must be cognizant of potential adverse side effects when we start patients on dyslipidemia medications. Some treatment combinations that were seen previously as unsafe are now recognized as acceptable options in the setting of appropriate patient monitoring. Historically, there has been concern about an increased rate of myopathy and liver damage with statin monotherapy and with combination therapy. Studies showed that approximately 5% of patients on a statin drug get nonspecific muscle aches and pains, a rate that is similar to placebo. Severe myopathy and a CK level that is more than ten times normal occur in approximately 0.08% of patients; predisposing factors include age older than 80 years, small frame, multisystem disease, perioperative period, and polypharmacy [37]. A meta-analysis of controlled trials showed that after 5 years, only 0.01% more patients developed rhabdomyolysis than in the control group [20]. The combination of statin and fibrate or statin and niacin used to be viewed with suspicion, but it is now an option under ATP III guidelines. With careful monitoring, only about 1% of patients get significant CK elevations of more than three times the upper limit of normal. Patients should be encouraged to report any muscle aches, weakness, or brown urine to their physicians immediately. A baseline CK level should be considered at the onset of treatment; if during the treatment course the CK level increases to more than 10 times normal, therapy should be discontinued. Do not forget to consider thyrotropin assessment in a patient who has elevated CK, because thyroid disease also can increase levels of this enzyme. In a patient who has muscle aches and no or moderate CK elevation (3–10 times normal), it is reasonable to monitor CK levels weekly until symptoms improve or worsen to the point where treatment needs to be discontinued. For asymptomatic patients who have significant CK elevations, one can consider a drug holiday followed by a rechallenge with close CK monitoring [37]. LFT elevations and possible liver toxicity also are concerns with statin treatment. Studies have shown that approximately 0.5% of patients discontinue statin use because of elevated LFTs, and that with careful monitoring it is not necessary to discontinue therapy unless values increase to more than three times the normal level [32]. At this point, there is some concern about the disproportionately high rate of reported adverse drug reactions with rosuvastatin, particularly because more than 60% of the observed adverse effects occurred at the range of doses that are used commonly in clinical practice [38]. Another potential concern about dyslipidemia medication safety has focused on whether there is an increase in ‘‘nonillness mortality’’ (NIM)d suicide, violence, and accidentsdwith statin use. A meta-analysis of clinical
916
ROSEN & SAMS
trials showed no evidence of increased NIM, and no relationship between the degree of cholesterol reduction and NIM. The study did note a trend toward increased deaths and violence with diet and nonstatin drugs [39], which was corroborated by another meta-analysis that showed an increase in mortality with fibrates (NNH ¼ 132 over 4.4 years; 95% CI, 69–662) [20]. Finally, a population-based prospective cohort study of more than 2200 elderly patients with an average age of 72 showed that patients in the lowest quartile of total, LDL, and non-HDL cholesterol were almost twice as likely to die over the 3-year study period as were those in the highest quartile, after adjusting for possible confounding variables. It is prudent to remember that, especially in the elderly, lower cholesterol is not always better, and it may be an indicator of frailty or subclinical disease [40]. Clinical commentary As with any other pharmacologic treatment, risks must be weighed against benefits with any dyslipidemia monotherapy or combination therapy. Overall, currently used dyslipidemia medications are safe in most patients, as long as a good follow-up plan is in place. The patient should be made aware of potential side effects, and encouraged to report any suspicious symptoms to his or her provider.
What laboratory tests should be ordered in the management of hyperlipidemia? Bottom line In a patient who does not have unusual risks or family history of dyslipidemia, cholesterol screening should be initiated at 20 years of age and repeated every 5 years until a concerning elevation is detected, at which point more frequent testing should be started and intervention considered. Check LFTs at the onset of therapy, at 6 to 12 weeks, and then annually or as needed. Some sources also recommend baseline CK at initiation of statin monotherapy or combination therapy (SOR: C, expert opinion). Evidence summary ATP III guidelines state that adults aged 20 years and older should have a fasting lipid profile done every 5 years. If it is impossible to obtain fasting levels, then only the total cholesterol and HDL yield valid results; however, they still can provide useful information, because studies have shown good agreement between fasting and nonfasting total cholesterol and HDL [41]. If an initial screening test shows a cholesterol level of more than 200 mg/dL, but a fasting LDL of less than 130 mg/dL and an HDL of more than 40 mg/dL, a patient is at average risk for a lipid-related event
MANAGING HYPERLIPIDEMIA
917
over the next 1 to 2 years. In this case, laboratory tests should be repeated every 1 to 2 years, because total cholesterol and LDL tend to increase with age. Also, a patient may develop a comorbid condition, such as nephrotic syndrome, diabetes mellitus, or thyroid disease, which can manifest itself initially as a dyslipidemia [2]. Some sources recommend risk-stratifying patients further by using C-reactive protein (CRP), because preliminary evidence shows that patients with low (!149 mg/dL) LDL but elevated CRP may benefit from treatment [42]. Initial blood work for someone who is beginning treatment with antidyslipidemic agents should include a fasting lipid panel, LFTs, baseline CK, and thyrotropin. Special tests, such as CRP, can be considered to define a subset of patients that is at excess risk for CHD. Follow-up tests should include LFTs 6 to 12 weeks after initiation of therapy, every 6 months for 1 year, and on a yearly basis thereafter. CK should be monitored if indicated by symptoms, but routine periodic retesting is not recommended. Clinical commentary Awareness of appropriate laboratory tests in the setting of cholesterol screening as well as during treatment initiation and continuation is key to ensuring patient safety, monitoring progress, and adjusting therapeutic regimens.
What alternative therapies have proven effective in treating hyperlipidemia? Bottom line Following an enriched diet that contains omega 3 polyunsaturated fatty acids by adding supplements or by increasing dietary sources (fatty fish [eg, salmon], flaxseed, canola and soybean oil and nuts) reduces the risk for cardiovascular events (SOR: A, based on a well-performed meta-analysis of controlled trials). A Mediterranean diet (high in fruits, vegetables, whole grains, beans, nuts, seeds, and olive oil with low to moderate amounts of fish, poultry, and dairy products and little red meat) decreases the incidence of recurrent cardiovascular events in patients who have CHD (SOR: B, based on one blinded controlled trial). Garlic, soy, red yeast rice, fiber, and green tea may decrease cholesterol or LDL levels, but no studies on cardiovascular outcomes have been performed on any of these dietary items (SOR: B, based on varying qualities of meta-analyses and RCTs). Evidence summary A large Finnish cohort of 1871 men who did not have CHD at entry into the study was followed for an average of 10 years. Men with the highest
918
ROSEN & SAMS
intake of fish oils had a decreased incidence of cardiac events compared with those with a low intake of fish oils (9.1% versus 14.9%; NNT ¼ 17) [43]. A well-performed meta-analysis of 11 RCTs showed that fish oil supplements or increased dietary intake of foods that contain polyunsaturated fats significantly decreased the incidence of fatal CHD and overall mortality. Only 24 high-risk patients need to be treated for 1.5 years to prevent one death [44]. In a secondary prevention study, 605 patients who had CHD were randomized to their usual French diet or to a Mediterranean diet. After 4 years, there was a decreased incidence of cardiovascular events in the intervention group (9.63% versus 18.74%; P!.0002; NNT ¼ 11)da larger effect than was seen in any medical intervention trial for secondary prevention of cardiovascular events [45]. Trials on garlic have had variable effects on cholesterol. A meta-analysis of 13 RCTs found an average 5.8% decrease in cholesterol with garlic compared with placebo. When only the higher-quality studies were included, no effect was found [46]. A meta-analysis of 38 controlled trials found that compared with a control diet, a diet that is high in soy protein (average intake of 47 g/d) lowered total cholesterol by 9.3%, LDL by 12.9%, and triglycerides by 10%, and increased HDL by 2.4% [47]. Red yeast rice, a fermented rice product that is used in Chinese foods, contains a family of naturally occurring statinlike substances. In a double-blind RCT, 83 patients who had hyperlipidemia were assigned randomly to receive red yeast rice or placebo for 8 weeks. At the end of the study, the total cholesterol in the arm that received red yeast rice had decreased by 17% and the LDL had decreased by 23%. In a meta-analysis of trials that studied soluble fiber, each gram of additional soluble fiber in the diet reduced LDL by an average of 2.2 mg/dL [48]. Finally, a single RCT in 240 people found that once daily green tea extract decreased LDL by 26.1% compared with placebo [49]. Clinical commentary In compliant patients, dietary changes may be more effective than are medical interventions for treatment of hyperlipidemia. The NNT for the Mediterranean diet and for diets that are rich in polyunsaturated fatty acids are much more robust than any statin study to date, although only smaller trials have been conducted. We need to do a better job of educating ourselves and our patients on the specifics of a healthy diet, and encouraging compliance with a healthy lifestyle for maximum health benefits.
Summary Hyperlipidemia is a serious disease that affects the health and well-being of many, and further complicates other chronic illnesses. When treating a patient who has a lipid disorder, it is wise to take a global approach to the
MANAGING HYPERLIPIDEMIA
919
problem by assessing the patient’s history and risk factors, collaborating on developing a healthy lifestyle plan to which the patient can commit, and initiating appropriate therapy when indicated.
References [1] Number of office visits (in thousands) by patients age 45 to 64 to selected specialties ranked by the 20 most frequent principal diagnoses by the physician (ICD-9 three digit code in parentheses), 2002. Available at: http://www.aafp.org/x25058.xml. Accessed on December 28, 2005. [2] Grundy SM, Cleeman JI, Merz CN, et al. Implications of recent clinical trials for the National Cholesterol Education Program Adult Treatment Panel III guidelines. Circulation 2004;110:227–39. [3] Pitt B. Low-density lipoprotein cholesterol in patients with stable coronary artery diseasedis it time to shift our goals? [editorial]. N Engl J Med 2005;352:1483–4. [4] Ebell MH, Siwek J, Weiss BD, et al. Strength of Recommendation Taxonomy (SORT): A patient-centered approach to grading evidence in the medical literature. Am Fam Physician 2004;69:549–57. [5] Third report of the National Cholesterol Education Program (NCEP) Expert Panel on detection, evaluation, and treatment of high blood cholesterol in adults (Adult Treatment Panel III). Circulation 2002;106:3143–421. [6] National Heart, Lung and Blood Institute. http://www.nhlbi.nih.gov/guidelines/ cholesterol. Accessed December 2005. [7] D’Agostino RB, Grundy S, Sullivan LM, et al. Validation of the Framingham Coronary Heart Disease Prediction Scores: results of a multiple ethnic groups investigation. JAMA 2001;286:180–7. [8] Studer M, Briel M, Leimenstoll B, et al. Effects of different antilipidemic agents and diets on mortality. Arch Intern Med 2005;165:725–30. [9] Kaiser FE. Cholesterol and the older adult. South Med J 1993;86:2511–4. [10] Vaughan CJ, Buckley BM. MRC/BHF Heart Protection Study of cholesterol lowering with simvastatin in 20,536 high-risk individuals: a randomized placebo-controlled trial. Lancet 2002;360:7–22. [11] LaRosa JC, Grundy SM, Waters DD, et al, for the Treating to New Targets (TNT) Investigators. Intensive lipid lowering with atorvastatin in patients with stable coronary artery disease. N Engl J Med 2005;352:1425–35. [12] Sever PS, Dahlof B, Poulter NR, et al. Prevention of coronary and stroke events with atorvastatin in hypertensive patients who have average or lower-than-average cholesterol concentrations, in the Anglo-Scandinavian Cardiac Outcomes Trial–Lipid Lowering Arm (ASCOTT-LLA): a multicentre randomized controlled trial. Lancet 2003;361:1149–58. [13] Cannon CP, Braunwald E, McCabe CH, et al. Pravastatin or Atorvastatin Evaluation and Infection Therapy–Thrombolysis in Myocardial Infarction 22 Investigators. Intensive versus moderate lipid lowering with statins after acute coronary syndromes. N Engl J Med 2004;350:1495–504. [14] Austin MA, Hokanson JE, Edwards KL. Hypertriglyceridemia as a cardiovascular risk factor. Am J Cardiol 1998;81:7B–12B. [15] Assmann G, Schulte H, Funke H, et al. The emergence of triglycerides as a significant independent risk factor in coronary artery disease. Eur Heart J 1998;19(Suppl M):M8–14. [16] Haim M, Benderly M, Brunner D, et al. Elevated serum triglyceride levels and long-term mortality in patients with coronary heart disease. Circulation 1999;100:475–82. [17] Rubins H, Robins S, Collins D, et al. Gemfibrozil for the secondary prevention of coronary heart disease in men with low levels of high density lipoprotein cholesterol. N Engl J Med 1999;341:410–8.
920
ROSEN & SAMS
[18] The Coronary Drug Project Research Group. Clofibrate and niacin in coronary heart disease. JAMA 1975;231:360–81. [19] Boehringer SK. Niacin use: an update. Pharmacist’s Letter/Prescriber’s Letter 2005;21(12): 211207. [20] Baigent C, Keech A, Kearney PM, et al. Efficacy and safety of cholesterol-lowering treatment: prospective meta-analysis of data from 90,056 participants in 14 randomised trials of statins. Lancet 2005;366:1267–78. [21] Ross DS, Allen IE, Connelly JE, et al. Clinical outcomes in statin treatment trials. A metaanalysis. Arch Intern Med 1999;159:1793–802. [22] Downs JR, Clearfield M, Weis S, et al. Primary prevention of acute coronary events with lovastatin in men and women with average cholesterol levels. Results of AFCAPS/TexCAPS. JAMA 1998;279:1615–22. [23] Pignone M, Phillips C, Mulrow C. Use of lipid lowering drugs for primary prevention of coronary heart disease: a meta-analysis of randomised trials. BMJ 2000;321:1–5. [24] Briel M, Studer M, Glass TR, et al. Effects of statins on stroke prevention in patients with and without coronary heart disease: a meta-analysis of randomized controlled trials. Am J Med 2004;117:596–606. [25] Manktelow B, Gillies C, Potter JF. Interventions in the management of serum lipids for preventing stroke recurrence. Cochrane Databases Syst Rev 2002;(3):CD002091. [26] White HD, Simes RJ, Anderson NE, et al. Pravastatin therapy and the risk of stroke. N Engl J Med 2000;343:317–26. [27] Collins R, Armitage J, Parish S, et al. Effects of cholesterol-lowering with simvastatin on stroke and other major vascular events in 20,536 people with cerebrovascular disease or other high-risk conditions. Lancet 2004;363:757–67. [28] Grover SA, Coupal L, Zowall H, et al. Evaluating the benefits of treating dyslipidemia: the importance of diabetes as a risk factor. Am J Med 2003;115:122–8. [29] Huang ES, Meigs JB, Singer DE. The effects of interventions to prevent cardiovascular disease in patients with type 2 diabetes mellitus. Am J Med 2001;111:633–42. [30] Van Staa TP, Wegman S, de Vries F, et al. Use of statins and risk of fractures. JAMA 2001; 285:1850–5. [31] Brown BG, Zhao X-Q, Chait A, et al. Simvastatin and niacin, antioxidant vitamins, or the combination for the prevention of coronary disease. N Engl J Med 2001;345:1583–92. [32] Xydakis AM, Ballantyne CM. Combination therapy for combined dyslipidemia. Am J Cardiol 2002;90(10B):21K–9K. [33] Larsen ML, Illingworth DR. Drug treatment of dyslipoproteinemia. Med Clin North Am 1994;78:225–45. [34] Kehoe WA. Combination cholesterol-lowering therapy. Pharmacist’s Letter/Prescriber’s Letter 2003;19:190304. [35] Shek A, Ferrill MJ. Statin-fibrate combination therapy. Ann Pharmacother 2001;35: 908–17. [36] Ballantyne CM, Houri J, Notarbartolo A, et al. Effect of ezetimibe coadministered with atorvastatin in 628 patients with primary hypercholesterolemia. Circulation 2003;107: 2409–15. [37] Pasternak RC, Smith SC, Bairey-Merz CN, et al. ACC/AHA/NHLBI clinical advisory on the use and safety of statins. J Am Coll Cardiol 2002;40:567–72. [38] Alsheikh-Ali AA, Ambrose MS, Kuvin JT, et al. The safety of rosuvastatin as used in common clinical practice: a postmarketing analysis. Circulation 2005;111:3051–7. [39] Muldoon MF, Manuck SB, Mendelsohn AB, et al. Cholesterol reduction and non-illness mortality: meta-analysis of randomised clinical trials. BMJ 2001;322:11–5. [40] Schupf N, Costa R, Luchsinger J, et al. Relationship between plasma lipids and all-cause mortality in nondemented elderly. J Am Geriatr Soc 2005;53:219–26. [41] Craig SR, Amin RV, Russell DW, et al. Blood cholesterol screening. J Gen Intern Med 2000; 15:395–9.
MANAGING HYPERLIPIDEMIA
921
[42] Ridker PM, Rifai N, Clearfield M, et al. Measurement of c-reactive protein for the targeting of statin therapy in the primary prevention of acute coronary events. N Engl J Med 2001;344: 1959–65. [43] Rissanen T, Voutilainen S, Nyyssonen K, et al. Fish oil-derived fatty acids, docosahexaenoic acid and docosapentaenoic acid, and the risk of acute coronary events: the Kuopio ischaemic heart disease risk factor study. Circulation 2000;102:2677–9. [44] Bucher HC, Hengstler P, Schindler C, et al. N-3 polyunsaturated fatty acids in coronary heart disease: a meta-analysis of randomized controlled trials. Am J Med 2002;112:298–304. [45] De Lorgeril M, Salen P, Martin JL, et al. Mediterranean diet, traditional risk factors, and the rate of cardiovascular complications after myocardial infarction. Circulation 1999;99: 779–85. [46] Stevinson C, Pittler MH, Ernst E. Garlic for treating hypercholesterolemia. A meta-analysis of randomized clinical trial. Ann Intern Med 2000;133:420–9. [47] Anderson JW, Johnstone BM, Cook-Newell ME. Meta-analysis of the effects of soy protein intake on serum lipids. N Engl J Med 1995;333:276–82. [48] Brown L, Rosner B, Willett WW, et al. Cholesterol-lowering effects of dietary fiber: a metaanalysis. Am J Clin Nutr 1999;69:30–42. [49] Maron DJ, Lu GP, Cai NS, et al. Cholesterol-lowering effect of a theaflavin-enriched green tree extract: a randomized controlled trial. Arch Intern Med 2003;163:1448–53.
Prim Care Clin Office Pract 33 (2006) 923–941
An Evidence-Based Approach to the Management of Depression Douglas Maurer, DO*, Ross Colt, MD, MBA Madigan Army Medical Center, Building 9040, Fitzsimmons Drive, Tacoma, WA 98431, USA
Despite a multitude of medical and complementary treatments, depression continues to exact a huge toll on the population worldwide. Depression accounts for more than 4% of the worldwide disease burden, on par with coronary heart disease and diarrheal diseases [1]. In the United States, the prevalence of depression is 5.4% to 8.9% [2]. Depression affects 1 out of 10 patients and affects 5% to 13% of outpatients [3]. The cost of depression is staggering, accounting for more than $43.7 billion in medical expenses and lost productivity [4]. Depression often is not treated adequately [5]. Furthermore, more than 75% of patients who have depression have recurrent episodes, and 10% to 30% have residual symptoms [6,7]. Depression has been associated with poorer outcomes in patients who have coronary artery disease, diabetes, and stroke [8–10]. Treatment of depression may reduce mortality from coronary artery disease and stroke, as well as help to prevent suicide [11–13]. It is obvious that depression is of the utmost importance to primary care physicians; however, providers frequently have inadequate guidance to choose between the myriad of options that is available [14]. Increasing numbers of patients are using complementary medicine for the treatment of depression, which complicates management [15]. What is the evidence in support of one medication over another? What medications are safe to use in children and pregnant women? Is there any evidence supporting over-the-counter supplements? These are just a few of the questions that primary care physicians face on a daily basis. This article attempts to answer these questions and many others in an evidence-based approach to the management of depression, which focuses on diagnosis, medical management, and complementary treatments. The views expressed in the article (book, speech, etc.) are those of the author(s) and do not reflect the official policy of the Department of the Army, the Department of Defense or the U.S. Government. * Corresponding author. E-mail address:
[email protected] (D. Maurer). 0095-4543/06/$ - see front matter Ó 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.pop.2006.09.003 primarycare.theclinics.com
924
MAURER & COLT
What is the underlying cause of depression? Studies of twins and families clearly demonstrate that depression is heritable. No specific gene abnormalities have been identified, although variants in the gene that encodes the serotonin-transporter protein may play a role [16,17]. Furthermore, the environmental effects on genetic predisposition are unclear. The neurotransmitter hypothesis for depression is supported well in the literature. Deficiencies in serotonin, norepinephrine, dopamine, and g-aminobutyric acid, and overactivity of acetylcholine, corticotrophin-releasing factor, and substance P all play a role in the pathogenesis of depression [17]. Finally, brain imaging using positron-emission tomography has demonstrated structural and neurotransmitter abnormalities in depressed patients, including the cingulate cortex, prefrontal cortex, insula, hippocampus, thalamus, amygdale, and brain stem [18]. What are the diagnostic criteria for depression? The diagnostic criteria of depression are as follows: 1. At least five of the following symptoms have been present during the same 2-week period, nearly every day, and represent a change from previous functioning. At least one of the symptoms must be depressed mood or loss of interest or pleasure: Depressed mood (or alternatively can be irritable mood in children and adolescents) Markedly diminished interest or pleasure in all, or almost all, activities Significant weight loss or weight gain when not dieting Insomnia or hypersomnia Psychomotor agitation or retardation Fatigue or loss of energy Feelings of worthlessness or excessive or inappropriate guilt Diminished ability to think or concentrate Recurrent thoughts of death, recurrent suicidal ideation without a specific plan, or a suicide attempt or a specific plan for committing suicide 2. Symptoms are not accounted for better by a mood disorder that is due to a general medical condition, a substance-induced mood disorder, or bereavement 3. Symptoms are not accounted for better by a psychotic disorder [19]. Should physicians screen for depression and which screening tool should be used? According to the US Preventative Services Task Force (USPSTF), screening adults for depression should be performed in clinical practices that have systems in place to assure accurate diagnosis, effective treatment,
DEPRESSION: AN EVIDENCE-BASED APPROACH
925
and follow-up. The USPSTF gives adult depression screening a grade B recommendation. However, the USPSTF finds insufficient evidence for or against routine screening for depression in children and adolescents [20]. In regards to specific screening instruments, physicians have many from which to choose. USPSTF finds little evidence to recommend one over another; however, asking two simple questions about mood and anhedonia (‘‘Over the past 2 weeks, have you felt down, depressed, or hopeless’’? and ‘‘Over the past 2 weeks, have you felt little interest or pleasure in doing things’’?) is as effective as longer instruments [20–23]. The Cochrane Collaboration found that routine screening had minimal impact on the detection, management, or outcome of depression, however [24]. Are there any meaningful differences between newer and older antidepressants? Despite a multitude of pharmaceutical company advertisements and provider educational programs, there is little evidence that newer antidepressants are more effective than are older ones. A meta-analysis of 315 trials found that newer antidepressants (selective serotonin reuptake inhibitors [SSRIs], venlafaxine, mirtazapine, nefazodone, bupropion) were as effective as tricyclic antidepressants (TCAs). Patients reported a greater number of side effects with TCAs, but patients were equally likely to discontinue a newer agent as they were an older one [25]. Anderson and Tomenson [26], specifically examined discontinuation rates among SSRIs and TCAs; SSRIs were associated with only 10% fewer dropouts than were TCAs. Likewise, other studies of SSRIs versus older TCAs found no difference in efficacy, but fewer side effects. A 2003 study of more than 2900 patients commented on the small sample sizes of the studies, pharmaceutical industry sponsorship, and poor methodologic quality [27]. The Cochrane Collaboration examined more than 98 trials that compared SSRIs with TCAs, and found no significant differences between the drugs [28]. A review by the Agency for Health Care Policy and Research came to similar conclusions [29]. A more recent meta-analysis examined which of the newer antidepressants were safer and more effective. Forty-six studies were examined (85% were sponsored by drug companies); 20 found no difference between the antidepressants. Sertraline (Zoloft) and venlafaxine (Effexor) were slightly more effective than was fluoxetine (Prozac). Sexual side effects were the highest with the SSRIs and lowest with bupropion (Wellbutrin). Weight gain was highest with mirtazapine (Remeron) and lowest with fluoxetine (Prozac) [30]. A study by Kroenke and colleagues [31] compared the SSRIs paroxetine, fluoxetine, and sertraline and found no difference in efficacy and safety. The study also suggested that patients who do not respond to one SSRI might respond to another. This was an open-label study, and by the end of the study, less than half of the patients were taking their original medication [32].
926
MAURER & COLT
What are the primary side effects of antidepressants? Physicians and patients are concerned about potential side effects of medications prescribed. There has been considerable media coverage on the side effects of both old and new antidepressants, particularly in the areas of sexual dysfunction and suicide risk. Safety in children, pregnancy, and in special patient groups is covered elsewhere in this article. The sexual dysfunction that is associated with antidepressants includes sexual arousal disorder, orgasmic dysfunction, and sexual desire disorder [33]. Gregorian and colleagues [34] reviewed more than 200 articles on sexual dysfunction associated with antidepressants. The researchers found that SSRIs were most likely to cause sexual dysfunction (30%–60%), whereas bupropion and nefazodone were least likely (%10%). Two meta-analyses that compared bupropion with SSRIs for efficacy and safety concluded that both treatments were equally effective, but that the SSRIs were associated with much greater sexual dysfunction [33,35]. A smaller study of 107 outpatients who had depression found that 73% of patients on SSRIs reported worsening sexual function, compared with only 14% of those on bupropion [36]. Some investigators even speculate why bupropion is not the most frequently prescribed antidepressant, because it is as efficacious as the more popular SSRIs, yet is associated with less weight gain and sexual dysfunction [37]. Nurnberg and colleagues [38] studied the use of sildenafil (Viagra) in men who had sexual dysfunction associated with antidepressant use. They found that 54.5% of those who took sildenafil improved compared with 4.4% of those who took placebo (number needed to treat [NNT] ¼ 2). It has been well established that depression carries risks for suicidal ideation and completed suicide; however, great controversy has existed over whether the medications that are used to treat depression actually increase the risk for suicide. Fergusson and colleagues [39] completed a systematic review as part of the Cochrane Collaboration and examined more than 702 trials. They concluded that SSRIs were associated with an increased suicide risk (odds ratio [OR], 2.28; 95% CI, 1.14–4.55). Another meta-analysis by Gunnell and colleagues [40] found no such association. Likewise, other recent studies found no conclusive evidence that SSRIs or TCAs increase the risk for suicide [41,42]. Note that all of these studies only included adult patients.
What drugs should primary care physicians prescribe first-line for depression? There is little evidence to support prescribing one antidepressant over another. Studies that compare SSRIs with each other or with TCAs or newer antidepressants, such as bupropion or seratonin-norepinephrine reuptake inhibitors (SNRIs), show little difference in efficacy [25–32,43–45].
DEPRESSION: AN EVIDENCE-BASED APPROACH
927
A meta-analysis of 32 trials found that the SNRI venlafaxine, was more effective than were SSRIs, but not TCAs [46]. Some investigators suggest that initial therapy should be based solely on cost, side effect profiles, comorbidities, depressive symptoms, history of response to therapy, and patient preference [25,26,29–32,47,48]. Clinical practice guidelines for the treatment of depression exist, but many are of poor quality [14]. The Institute for Clinical Systems Improvement (ICSI) recently published a thorough evidence-based guideline for the management of depression in adults [49]. The ICSI guideline recommends SSRIs, venlafaxine, mirtazapine, and bupropion equally for firstline treatment of depression. The ICSI cites these medications’ simplicity of use, favorable side effect profiles, and community standards for their choices. They recommend secondary amine tricyclics as second-line therapy, and recommend restricting the use of monoamine oxidase inhibitors for nonresponders. If patients are not responding to treatment, what should be done next? Patients who are on antidepressant therapy should be evaluated for adequate response in 4 to 6 weeks. A reasonable criterion for an adequate response to therapy is a 25% reduction in baseline symptom severity [48,49]. Often, it is unclear about what should be done if patients are not improving. Patients should be evaluated for medication compliance, drug interactions, comorbid medical conditions, and proper diagnosis. Various investigators recommend increasing the dose of the current medication, switching to another drug within the same class as the current medication versus switching classes entirely, combining the current medication with one from another class, or augmenting treatment with nonantidepressants (eg, a mood stabilizer or anticonvulsant) [47,48,49]. Patients who improve by at least 25%, but who are not yet in remission should have the dose of their current medication increased. Patients who have not improved by at least 25% should be switched to a new medication, either from the same class or a different one [48,49]. Patients who do not respond to one SSRI or TCA have up to a 70% chance of responding to a different one [31,50]. Another strategy is to combine the current medication with another medication from a different class. This strategy uses the concept of multiple mechanisms of action (eg, combining an SSRI with bupropion, combining an SSRI with a TCA, an SSRI, and an SNRI). Obviously, the risk for drug interactions and adverse effects is much greater with this strategy. A review of combination therapy examined 27 studies of 667 patients; the overall response rate was 62.2%; however, 22 of the studies were uncontrolled, open-label trials [51]. Finally, augmentation with other medications can be used for treatmentresistant depression. Lithium is considered a first-line augmentation
928
MAURER & COLT
medication and has the most evidence to advocate its use [48,49,52,53]. Most of the evidence with lithium comes from studies of severe depression and electroconvulsive therapy trials, however, and not from the primary care literature. Studies of anticonvulsants, such as phenytoin, carbamazepine, lamotrigine, and valproic acid are promising, but the quality of the evidence is poor [52,54]. Many of the anticonvulsant data come from studies of bipolar disorder. Thyroid hormone has been studied to augment depression, even in the absence of hypothyroidism. A meta-analysis by Altshuler and colleagues [55] found that augmentation with triiodothyronine in patients who were on TCAs improved symptoms by about 10 days. Studies of thyroid hormone augmentation in patients who are on SSRIs are less convincing. The literature contains only small, mostly uncontrolled trials [56]. Many patients who have depression suffer from concomitant anxiety. Benzodiazepines are used commonly along with antidepressants in patients who have newly diagnosed depression and significant anxiety symptoms. One small, double-blind trial of 80 patients demonstrated improvements in depression rating scales in those who were treated with fluoxetine and clonazepam [57]. A Cochrane review of nine studies with 679 patients found that those who were on combination therapy were less likely to drop out of the study and reported improved symptoms at 4 weeks than were those who were on antidepressants alone. The results were not significant at 6 weeks, however, and Cochrane cautioned use, citing the risks for dependence and accident proneness. Of note, this Cochrane review was updated last in 1999 [58]. How long should patients who have depression be treated with antidepressants? Many physicians and patients struggle with the decision of how long to treat depression. Antidepressant therapy can be divided into three phases: the acute phase, the continuation phase, and the maintenance phase. The acute phase is the first 3 months of therapy during which remission is the goal. The continuation phase is the next 6-month period during which medication is used for continued alleviation of symptoms. The maintenance phase is an additional 6- to 24-month period during which remission is maintained with or without medication [47,49]. Evidence suggests that treatment for less than 6 months results in an unacceptably high relapse rate [59]. Most guidelines advocate treating for at least 6 to 12 months after induction of remission [47,49,60]. The ICSI guideline recommends 6 to 12 months of treatment for a first episode, 3 years of treatment for a second episode, and lifetime therapy for treatment of a third episode or second episodes with complicating factors (eg, dysthymia) [49].
DEPRESSION: AN EVIDENCE-BASED APPROACH
929
What is the antidepressant of choice in children and adolescents? Recent publicity surrounding unpublished ‘‘negative’’ studies has brought the medical management of depression in children and adolescents to the forefront. A meta-analysis by Whittington and colleagues [61] of published and unpublished studies found two published studies that showed beneficial effects of fluoxetine, and several unpublished studies that also showed benefit. One published study of paroxetine and two published studies of sertraline suggested only minimal benefit, whereas unpublished studies of these drugs, as well as citalopram and venlafaxine, showed more harm than benefit (increased suicidal ideation and attempt, hostility, self-harm). Wagner and colleagues [62] conducted two randomized placebo-controlled clinical trials using sertraline that included 327 patients. The researchers found 40% improvement in depression scores (P ¼ .05), but only followed patients for 10 weeks. Sixty-nine percent of patients improved on sertraline, versus 59% on placebo; only 1 in 10 patients on sertraline improved [62,63]. This study supports the opinion of some investigators, that the few positive studies on the medical treatment of depression in children is a result of a high placebo response rate [63–65]. A Cochrane review of tricyclics in children evaluated 13 trials that included 506 patients. It found no evidence of benefit (OR, 0.84; 95% CI, 0.56–1.25) and significant adverse effects including vertigo, orthostatic hypotension, tremor, and dry mouth [66]. Clearly, more research on antidepressants in children is needed. For now, experts recommend fluoxetine, the only SSRI that is approved for pediatric use, if medication is to be used. Psychotherapy has been shown to be beneficial and safe in children [47,49,63].
Which antidepressants are safe in pregnancy? Untreated depression in pregnant women can have devastating results on mother and child, but recent studies again are questioning the safety of some antidepressants in pregnancy. There are no antidepressants with a pregnancy risk category A. Bupropion and maprotiline are category B. Most others are category C. Imipramine, nortriptyline, and, as of December 2005, paroxetine are category D. None are rated Category X [49]. Studies have examined the effects of using antidepressants specifically during the first and third trimesters, and on the fetus in the immediate postpartum period. Most of the studies involve fluoxetine. A case-control study of more than 500 women who were exposed to fluoxetine during the first trimester reported that it was associated with minor anomalies, whereas exposure in the third trimester was associated with premature delivery, poor neonatal adaptation, lower birth rate, and shorter birth length. Major structural anomalies and spontaneous pregnancy loss were not associated with fluoxetine use [67].
930
MAURER & COLT
Two separate meta-analyses of more than 900 patients who used fluoxetine, newer SSRIs, and TCAs during the first trimester found no significant difference in the rate of major fetal malformations [68,69]. Studies on the use of venlafaxine and bupropion during the first trimester found no increase in the risk for major malformations; however, both studies were small, and the use of bupropion was associated with an increase in spontaneous abortions (15% with bupropion versus 12% with other antidepressants versus 7% with nonteratogens) [70,71]. Recently, numerous cases of a neonatal toxicity versus withdrawal-like syndrome, termed the ‘‘poor neonatal adaptation syndrome,’’ were reported in infants who were born to mothers on TCAs, SSRIs, and SNRIs. The US Food and Drug Administration (FDA) has modified the labeling of these drugs to reflect these developments [69,72–74]. The syndrome consists of transient jitteriness, poor muscle tone, weak cry, and respiratory distress; no fatalities have been reported. It has been seen in up to 30% of infants exposed [72,73]. The FDA advocated tapering the dosage over the last trimester to avoid exposure at birth; many investigators argue against this, and see greater risk in the untreated maternal depression. No specific treatments have been studied or universally recommended; however, these children should be observed carefully [49,73]. Evidence regarding the long-term outcomes of children who are exposed to antidepressants in utero is lacking [49,72].
Should antidepressants be used in patients who have experienced a myocardial infarction or stroke? The evidence is mounting that depression is linked to coronary artery disease, and patients who have had a stroke are at high risk for depression. It is still unclear whether treating depression improves outcomes in these patients [8,47,75]. Many investigators debate the safety of antidepressants in these conditions. Glassman and colleagues [76] studied the safety of sertraline in 369 depressed patients who had been hospitalized recently for acute myocardial infarction. Sertraline did not affect ejection fraction, ventricular arrhythmias, or QT interval. The group that took sertraline had improved scores on a depression scale and a nonsignificant lower risk for adverse cardiac events. A smaller study that compared paroxetine with nortriptyline in 81 depressed patients who had coronary artery disease found that both drugs were effective in treating depression, but there was a higher risk for cardiac events with nortriptyline [47,77]. The literature on the use of antidepressants in patients who have suffered a stroke is even more controversial. Antidepressants that affect blood pressure, such as TCAs and SNRIs, should be avoided [47,78–80]. A small study of 121 patients who were treated with fluoxetine for depression after a stroke showed improvement in symptoms, but the patients only were followed for
DEPRESSION: AN EVIDENCE-BASED APPROACH
931
45 days [81]. A study of 104 patients who had had a stroke and were treated with fluoxetine or nortriptyline showed reduction in mortality (59.2% versus 36.4%; P ¼ .03), regardless of whether they were depressed. Fluoxetine and nortriptyline were equally effective [12]. A systematic review by Hackett and colleagues [82] that examined seven treatment trials of 615 patients and nine prevention trials of 479 patients found insufficient evidence to support the routine use of antidepressants for the prevention of depression or to improve recovery from stroke. The investigators cited difficulty with the analysis because of a lack of standardized diagnostic and outcome criteria and differing analytic methods.
Does counseling provide a measurable therapeutic effect in depressed patients? There have been several Cochrane reviews to assess the effectiveness and cost-effectiveness of counseling in primary care. In general, counseling seems to provide significantly greater clinical effectiveness in the short term, with high patient satisfaction; however, this effect does not persist long-term. One difficulty in evaluating the effectiveness of counseling is that different trials use a multitude of different measurement tools. The Cochrane group attempted to make the measurements homogeneous by transforming continuous data from different measuring instruments into a standard effect size by dividing mean values by standard deviations. They also performed tests of heterogeneity to assess the feasibility of aggregating measures of outcome from trials. Examining seven trials that were completed before June of 2001, the group found that counseling was associated with significantly better clinical outcomes in the short term [83]; however, this effect did not persist in the long term. Of note, the total costs of counseling versus usual care over the long-term were similar, although the economic analyses were likely to be underpowered. One particular type of psychotherapy is cognitive behavioral treatment (CBT). A small randomized controlled trial (nonblinded) of 40 patients evaluated the use of CBT to prevent relapse in patients who were tapering off from antidepressant drugs [84]. The researchers defined CBT as ‘‘strategies and techniques designed to help patients correct distorted life views and maladaptive beliefs, and to recognize that depression is the consequence of a maladaptive lifestyle which does not take life stress, interpersonal friction, excessive work, and inadequate rest into proper account.’’ Patients in the group that used CBT reported significantly lower residual symptoms at the 20-week mark. At 2 years, the group that used CBT had a significantly lower relapse rate (25%) than did the group that was managed with traditional methods (80%), yielding a NNT of 1.8. It is worth noting that in the above mentioned Cochrane review, other studies found no difference between CBT and standard counseling for the general treatment of depression.
932
MAURER & COLT
Therefore, CBT’s niche may be in decreasing relapses in depression management. A more specific area of interest for many providers is the use of psychosocial interventions by general practitioners, rather than other health care professionals, such as counselors. Another Cochrane review evaluated eight studies in the literature that were available in this area and published before January of 2002 [85]. They found that there is good evidence that problemsolving treatment by general practitioners is effective for major depression.
Should health care providers recommend an exercise program for their patients who suffer from depression? Exercise has proven to be a beneficial addition to the treatment program of a variety of maladies. It is not surprising to learn that exercise can be an effective treatment for depression. Blumenthal and colleagues [86] examined the effects of a supervised walking program on a group of 156 depressed older patients (ageO49 years). In this randomized, single-blinded, controlled trial, patients were assigned to treatment with the antidepressant sertraline (Zoloft), a supervised exercise program, or both. The exercise program consisted of 30 minutes of walking or jogging at 70% to 80% of maximal aerobic intensity three times a week. After 4 months, 60% to 69% of the patients were no longer depressed, and there were no statistically significant differences in depression scores between any of the three groups. Of note, patients who were on drug therapy responded more quickly, but the exercisers caught up by the end of the trial. The results of this study may be tempered by the high drop-out rate (15% of patients on medication and 26% of patients who were exercising); however, an exercise program still can be considered as an alternative to medications for treating depression in select groups of older patients. A second study that was performed by Dunn and colleagues [87] looked at what could be described as the dose-related effect of exercise on depression. The researchers looked at a group of 80 adults, ages 20 to 45 years, who had been diagnosed with mild to moderate major depressive disorder. The patients were assigned randomly to one of five exercise treatment groups: high- or low-dose exercise based on the number of kcal/kg the patients burned divided into three or five sessions per week, or flexibility exercises three sessions per week as a control. The high-dose exercise is consistent with public health recommendations for exercise, equivalent to a 70-kg man exercising on a treadmill to a heart rate of 145 beats per minute for 30 minutes. Patients who were assigned to the high-dose exercise showed a significant clinical response, defined as a 50% or more reduction in the mean Hamilton Rating Scale for Depression, compared with the control group (42% versus 23%; NNT ¼ 5). The results from the groups that were assigned low-dose exercise were equivocal. Of note, 12.5% of the
DEPRESSION: AN EVIDENCE-BASED APPROACH
933
patients (10 patients) were lost to follow-up at 12 weeks. Thus, recommending an exercise program at a level that is consistent with current public health recommendations likely can provide some treatment efficacy in adults who are suffering from mild to moderate depression.
Are there any herbal remedies that have proven benefit in the treatment of depression? St. John’s wort is an herbal remedy from the Hypericum perforatum plant that has long been used for treating depression. The name comes from the old English term for a plant (wort) that generally bloomed in June during the historical feast of John the Baptist. A recently updated Cochrane review of 37 randomized trials that involved 4925 patients evaluated the efficacy of St. John’s wort in the treatment of depression [88]. Most of the trials were short in duration (4–6 weeks). Compared with previous reviews of hypericum, the most recent placebo-controlled studies seem to show only minor benefits in patients who have mild to moderate depression. In the six largest trials, the combined response rate (RR) for hypericum versus placebo was 1.15 (95% CI, 1.02–1.29). In six smaller trials, the RR was 2.06 (95% CI, 1.65–2.59). Comparing hypericum with standard antidepressants showed a RR of 0.98 (95% CI, 0.85–1.12; six trials) for SSRIs and 1.03 (95% CI, 0.93–1.14; seven trials) for tri- or tetracyclic antidepressants. Of note, patients who took hypericum reported significantly less adverse effects compared with older antidepressants, but the differences were nonsignificant compared with SSRIs. Older trials were performed almost exclusively in German-language countries and tended to be smaller, shorter, and less likely to use a placebo run-in design. The investigators noted that St John’s wort can have significant interactions with several frequently used drugs, and recommended that physicians regularly ask their patients about hypericum use. They also commented on the high variability of bioactive components in hypericum preparations, and recommended avoiding products that do not provide explicit content information, such as the amount of total extract (eg, 900 mg) that is found in the preparation.
What is evidence behind hormonal supplements? Dehydroepiandrosterone (DHEA) is an over-the-counter hormonal agent that has been reported to have some antidepressant effects. The data supporting its use in the treatment of depression is sparse. In a small cross-over trial of 46 adults, Schmidt and colleagues randomized patients to receive DHEA or matched placebo for 6 weeks with a 1- to 2-week washout before crossover. The dosage was 30 mg three times a day for the first 3 weeks and 150 mg 3 times a day for the last 3 weeks. The primary outcome measured was a 50% or greater reduction in baseline Hamilton
934
MAURER & COLT
Depression Rating Scale. A total of 23 subjects who took DHEA (50%) reported a 50% reduction in the score versus 13 of the patients (28%) who took placebo (P!.01). Of note, women received a 10-day course of medroxyprogesterone acetate at the end of the trial to cause shedding of any endometrial growth that could have been stimulated by the DHEA. Although the results of this single small trial are promising, more studies are needed. The Agency for Health care Research and Quality (AHRQ) conducted a search of the literature to evaluate the effects of S-adenosyl-L-methionine (SAMe) on depression as well as on several other disease entities. They found 39 unique studies focused on depression that presented randomized clinical trials on human subjects, but most were small and the quality varied considerably. From this group they included 28 studies in a meta-analysis to determine the efficacy of SAMe in decreasing the symptoms of depression. Treatment with SAMe was associated with an approximately 6-point increase in the Hamilton Rating Scale for Depression measured at 3 weeks (95% CI, 2.2–9.0). The researchers considered this result to be statistically and clinically significant. Compared with conventional antidepressant pharmacology, there was not a statistically significant difference in outcomes when using SAMe for treatment. The AHRQ concluded that a need exists for additional review studies, clinical studies, and studies to help understand the pharmacology of SAMe better. They noted that good dose-escalation studies have not been performed to help determine the most effective oral dose of SAMe.
Is there any proven benefit of light therapy in the treatment of depression? Bright light therapy has been used with some success for seasonal affective disorder, but there is less evidence to support its use in nonseasonal depression. The Cochrane group recently searched for randomized controlled trials that compared bright light with inactive placebo treatments (eg, dim light) for nonseasonal depression [89]. Their review included 20 studies, most of which applied bright light as adjunctive treatment to drug therapy, the rather counterintuitive treatment of sleep deprivation, or both. The results as a whole did not reach clinical significance, but showed a nonsignificant trend toward higher efficacy with bright light. In a subgroup that included only the high-quality studies, the response to bright light was significantly better than in the control group (standardized mean difference 0.90; 95% CI, 1.50–0.31). The benefit of bright light treatment also achieved statistical significance in studies that applied morning light treatment (standardized mean difference, 0.38; 95% CI, 0.62–0.14) and in sleep deprivation responders (standardized mean difference, 1.02; 95% CI, 1.60–0.45). One important adverse effect was that hypomania was more common with bright light therapy than with control treatment
DEPRESSION: AN EVIDENCE-BASED APPROACH
935
(RR, 4.91; 95% CI, 1.66–14.46; number needed to harm 8; 95% CI, 5–20). The researchers concluded that light therapy offers ‘‘modest though promising’’ antidepressive efficacy, especially in the above-noted subgroups. They noted that because of the heterogeneity of studies, with many studies being of poor quality, these results should be interpreted with caution.
What about collaborative therapy and disease management programs? Given the constraints that generally are faced in a primary care office, health care providers often may feel that they do not have sufficient time or resources to treat depression on their own. Being able to use a team approach offers one possible solution. Katon and colleagues [90] sought to determine if a collaborative approach to treating persistent or recurrent depression was effective in a randomized controlled trial. The researchers looked at a group of 2699 patients who had received a new antidepressant medication through their large health maintenance organization. These patients were contacted 8 weeks later and asked whether they were having residual or recurrent depressive symptoms. This subgroup of 228 patients that were still having symptoms was randomized to receive standard care from their primary care physician or a ‘‘stepped collaborative care intervention.’’ The latter intervention entailed a book and accompanying videotape, two sessions with a psychiatrist, and phone calls in between visits to review their progress. The psychiatrist worked with the primary care physician to adjust medications as necessary, and a minority of patients (12%) also was sent for psychotherapy. Although the collaboration did not show a significant effect in patients who had severe depression, patients who had moderate depression showed a continuing improvement over the 28 months of the study. Mean symptom check list scores were 0.9/4.0 versus 1.2/4.0 for traditional care, and disability scores were 3.09/10 versus 3.58/10 for traditional care (P ¼ .004). It may be surprising to note that overall average costs for health care were about 8% lower with intervention, although depression treatment costs were higher. The investigators concluded that a collaborative care intervention improved outcomes for depression without additional health care costs in patients who had persistent depressive symptoms. A recent meta-analysis on the subject of disease management programs demonstrated similarly positive clinical outcomes but higher costs in general. Neumeyer-Gromen and colleagues [91] performed a literature search for randomized controlled trials that compared a disease management program with traditional care for depression. This yielded 10 studies that encompassed 4196 patients, generally in the environment of primary care clinics of managed care organizations. Most of the disease management programs included patient education, provider education, pharmacy monitoring (by a psychiatrist, psychologist, or nurse), and collaborative care with
936
MAURER & COLT
consultants. They found that patients in disease management programs had a greater likelihood of improvement in disease severity (RR, 0.75; 95% CI, 0.70–0.81) and better patient satisfaction (RR, 0.57; 95% CI, 0.37–0.87). Patient adherence to the treatment regimen also improved significantly. The drawback was a significant increase in costs, which varied from $9051 to $49,500 per quality-adjusted life-year, depending on the cost-effectiveness analysis. The investigators concluded that disease management programs significantly enhanced the quality of care for depression, and that the increased costs were within the range of other widely accepted public health improvements.
Is there any proven benefit for the use of alternative or complementary medicine? Acupuncture has been used for centuries in China and Japan to correct a variety of health maladies. Traditional Chinese medicine described the use of acupuncture to maintain a state of health by correcting the imbalance of energy in the body. A recent Cochrane review evaluated seven trials with a total of 517 patients to determine if the evidence supports the efficacy and adverse effects of acupuncture for the treatment of depression [92]. Five of the trials included a comparison of acupuncture with medication, whereas two trials compared acupuncture with sham acupuncture or a wait list control. The scientific study design was noted to be poor. There was no significant difference between acupuncture and medication in reducing the severity of depression (weighted mean difference, 0.53; 95% CI, 1.42–2.47) or decreasing the incidence of remission (RR, 1.2; 95% CI, 0.94–1.51). Adverse effects reported in the groups that received acupuncture generally were mild and included sleep disturbances, headaches, fatigue, palpitations, and dry mouth. The investigators concluded that there is insufficient evidence to determine the efficacy of acupuncture in the management of depression because of a lack of welldesigned, randomized controlled trials and the overall small number of people studied.
Are there any other treatment modalities that are supported by evidence? This article by no means provides an exhaustive list of all treatment modalities that could be used to treat depression. Some therapies, such as electroconvulsive treatment (likely effective for refractory depression or in elderly patients), transcranial magnetic stimulation (no strong evidence), and vagus nerve stimulation (no strong evidence), have not been discussed here because of their presumed limited use in a primary care setting [93–96].
DEPRESSION: AN EVIDENCE-BASED APPROACH
937
References [1] The world health report 2002dreducing risks, promoting healthy life. Geneva: World Health Organization; 2002. [2] Narrow WE, Rae DS, Robins LN, et al. Revised prevalence estimates of mental disorders in the United States: using a clinical significance criterion to reconcile 2 survey’s estimates. Arch Gen Psychiatry 2002;59:115–23. [3] Coyne JC, Fechner-Bates S, Schwenk TL. Prevalence, nature, and comorbidity of depressive disorders in primary care. Gen Hosp Psychiatry 1994;51:8–19. [4] Chisholm D. The economic costs of depression. In: Dawson A, Tylee A, editors. Depression: social and economic timebomb. London: BMJ Books; 2001. p. 10–2. [5] Goldman LS, Nielsen NH, Champion HC. Awareness, diagnosis, and treatment of depression. J Gen Intern Med 1999;14:569–80. [6] Mueller TL, Leon AC, Keller MB, et al. Recurrence after recovery from major depressive disorder during 15 years of observational follow-up. Am J Psychiatry 1999;156: 1000–6. [7] Keller MB, Lavori PW, Rice J, et al. The persistent risk of chronicity in recurrent episodes of nonbipolar major depressive disorder: a prospective follow-up. Am J Psychiatry 1986;143: 24–8. [8] Ford DE. Depression is a risk factor for coronary artery disease in men. The precursors study. Arch Intern Med 1998;158:1422–6. [9] Ciechanowski PS, Kanton WJ, Russo JE. Depression and diabetes: impact of depressive symptoms on adherence, function, and costs. Arch Intern Med 2000;160:3278–85. [10] Robinson RG, Bolduc P, Price TR. Two-year longitudinal study of poststroke mood disorders: diagnosis and outcome at one and two year follow-up. Stroke 1987;18:837–43. [11] Taylor CB, Youngblood ME, Catellier D, et al. Effects of antidepressant medication on morbidity and mortality in depressed patients after acute myocardial infarction. Arch Gen Psychiatry 2005;62:792–8. [12] Jorge RE, Robinson RG, Arndt S, et al. Mortality and poststroke depression: a placebocontrolled trial of antidepressants. Am J Psychiatry 2003;160:1823–9. [13] Gibbons RD, Hur K, Bhaumik DK, et al. The relationship between antidepressant use and rate of suicide. Arch Gen Psychiatry 2005;62:165–72. [14] Cornwall PL, Scott J. Which clinical practice guidelines for depression? An overview for busy practitioners. Br J Gen Prac 2000;50:908–11. [15] Knaudt PR, Connor KM, Weisler RH, et al. Alternative therapy use by psychiatric outpatients. J Nerv Ment Dis 1999;187:692–5. [16] Kendler KS, Davis CG, Kessler RC. The familial aggregation of common psychiatric and substance use disorders in the National Comorbidity Survey: a family history study. Br J Psychiatry 1997;170:541–8. [17] Nemeroff CB, Vale WW. The neurobiology of depression: inroads to treatment and new drug discovery. J Clin Psychiatry 2005;66:5–13. [18] Milak MS, Parsey RV, Keilp J, et al. Neuroanatomic correlates of psychopathologic components of major depressive disorder. Arch Gen Psychiatry 2005;62:397–408. [19] American Psychiatric Association. Diagnostic and statistical manual of mental disorders. 4th edition: primary care version. Washington, DC: American Psychiatric Association; 1995. [20] US Preventative Services Task Force. Screening for depression: recommendations and rationale. Ann Intern Med 2002;136(10):760–4. [21] Williams JW, Noel PH, Cordes JA, et al. Is this patient clinically depressed? JAMA 2002;287: 1160–70. [22] Arroll B, Khin N, Kerse N. Screening for depression in primary care with two verbally asked questions: cross sectional study. BMJ 2003;327:1144–6. [23] Whooley MA, Avins AL, Miranda J, et al. Case-finding instruments for depression. J Gen Intern Med 1997;12:439–45.
938
MAURER & COLT
[24] Gilbody S, House AO, Sheldon TA. Screening and case finding instruments for depression. Cochrane Database Syst Rev 2005;(4):CD002792. [25] Williams JW, Mulrow CD, Chiquette E, et al. A systematic review of newer pharmacotherapies for depression in adults: evidence report summary. Ann Intern Med 2000;132:743–56. [26] Anderson IM, Tomenson BM. Treatment discontinuation with selective serotonin reuptake inhibitors compared with tricyclic antidepressants: a meta-analysis. BMJ 1995;310:1433–8. [27] MacGillivray S, Arroll B, Hatcher S, et al. Efficacy and tolerability of selective serotonin reuptake inhibitors compared with tricyclic antidepressants in depression treated in primary care: systematic review and meta-analysis. BMJ 2003;326:1014–26. [28] Geddes JR, Freemantle N, Mason J, et al. SSRIs versus other antidepressants for depressive disorder. Cochrane Database Syst Rev 2000;(2):CD001851. [29] Treatment of depressiondnewer pharmacotherapies. Summary, evidence report/technology assessment. Agency for Health Care Policy and Research, Rockville, MD. Available at: http://www.ahrq.gov/clinic/deprsumm.htm. Accessed November 9, 2006. [30] Hansen RA, Gartlehner G, Lohr KN, et al. Efficacy and safety of second-generation antidepressants in the treatment of major depressive disorder. Ann Intern Med 2005;143:415–26. [31] Kroenke K, West SL, Swindle R, et al. Similar effectiveness of paroxetine, fluoxetine, and sertraline in primary care. A randomized trial. JAMA 2001;286:2947–55. [32] Straton J, Cronholm P. Are paroxetine, fluoxetine, and sertraline equally effective for depression? J Fam Pract 2002;51:285–6. [33] Nieuwstraten CE, Dolovich LR. Bupropion versus selective serotonin-reuptake inhibitors for treatment of depression. Ann Pharmacother 2001;35:1608–13. [34] Gregorian RS, Golden KA, Bahce A, et al. Antidepressant-induced sexual dysfunction. Ann Pharmacother 2002;36:1577–89. [35] Thase ME, Haight BR, Richard N, et al. Remission rates following antidepressant therapy with bupropion or selective serotonin reuptake inhibitors: a meta-analysis of original data from 7 randomized controlled trials. J Clin Psychiatry 2005;66:974–81. [36] Modell JG, Katholl CR, Modell JD, et al. Comparative sexual side effects of bupropion, fluoxetine, paroxetine, and sertraline. Clin Pharmacol Ther 1997;61:476–87. [37] Zimmerman M, Posternak MA, Attiullah N, et al. Why isn’t bupropion the most frequently prescribed antidepressant? J Clin Psychiatry 2005;66:603–10. [38] Nurnberg HG, Hensley PL, Gelenberg AJ, et al. Treatment of antidepressant-associated sexual dysfunction with sildenafil. A randomized controlled trial. JAMA 2003;289:56–64. [39] Fergusson D, Doucette S, Glass KC, et al. Association between suicide attempts and selective serotonin reuptake inhibitors: systematic review of randomized controlled trials. BMJ 2005;330:396–9. [40] Gunnell D, Saperia J, Ashby D. Selective serotonin reuptake inhibitors (SSRIs) and suicide in adults: meta-analysis of drug company data from placebo controlled, randomized controlled trials submitted to the MHRA’s safety review. BMJ 2005;330:385–91. [41] Khan A, Khan S, Kolts R, et al. Suicide rates in clinical trials of SSRIs, other antidepressants and placebo: analysis of FDA reports. Am J Psychiatry 2003;160:790–2. [42] Martinez C, Rietbrock S, Wise L, et al. Antidepressant treatment and the risk of fatal and non-fatal self harm in first episode depression: nested case-control study. BMJ 2005;330: 389–93. [43] Simon GE, VonKorff M, Heiligenstein JH, et al. Initial antidepressant of choice in primary care. JAMA 1996;275:1897–902. [44] Simon GE, Heiligenstein J, Revicki D, et al. Long-term outcomes of initial antidepressant drug choice in a ‘‘real world’’ randomized trial. Arch Fam Med 1999;8:319–25. [45] Burke WJ. Selective versus multi-transmitter antidepressants: are two mechanisms better than one? J Clin Psychiatry 2004;4:37–45. [46] Smith D, Dempster C, Glanville J, et al. Efficacy and tolerability of venlafaxine compared with selective serotonin reuptake inhibitors and other antidepressants: a meta-analysis. Br J Psychiatry 2002;180:396–404.
DEPRESSION: AN EVIDENCE-BASED APPROACH
939
[47] Mann JJ. The medical management of depression. N Engl J Med 2005;353:1819–33. [48] Ellis PM, Smith DA. Treating depression: the beyondblue guidelines for treating depression in primary care. ‘‘Not so much what you do but that you keep doing it.’’ Med J Aust 2002;20: 77–83. [49] Institute for Clinical Systems Improvement. Major depression in adults for mental health care. Bloomington (MN): Institute for Clinical Systems Improvement; 2004. [50] Thase ME, Rush AJ, Howland RH, et al. Double-blind switch study of imipramine or sertraline treatment of antidepressant-resistant chronic depression. Arch Gen Psychiatry 2002;59: 233–9. [51] Lam RW, Wan DD, Cohen NL, et al. Combining antidepressants for treatment-resistant depression. J Clin Psychiatry 2002;63:685–93. [52] DeBattista C, Lembke A. Update on augmentation of antidepressant response in resistant depression. Curr Psychiatry Rep 2005;7:435–40. [53] Bschor T, Lewitska U, Sasse J, et al. Lithium augmentation in treatment-resistant depression: clinical evidence, serotonergic and endocrine mechanisms. Pharmacopsychiatry 2003;36:230–4. [54] Barbosa L, Berk M, Vorster M. A double-blind, randomized, placebo-controlled trial of augmentation with lamotrigine or placebo in patients concomitantly treated with fluoxetine for resistant major depressive episodes. J Clin Psychiatry 2003;64:403–7. [55] Altshuler LL, Bauer M, Frye MA, et al. Does thyroid hormone supplementation accelerate tricyclic antidepressant response? A review and meta-analysis of the literature. Am J Psychiatry 2001;158:1617–22. [56] Iosifescu DV, Nierenberg AA, Mischoulon D, et al. An open study of triiodothyronine augmentation of selective serotonin reuptake inhibitors in treatment-resistant major depressive disorder. J Clin Psychiatry 2005;66:1038–42. [57] Smith WT, Londburg PD, Glaudin V, et al. Short-term augmentation of fluoxetine with clonazepam in the treatment of depression: A double-blind study. Am J Psychiatry 1998;155: 1339–45. [58] Furukawa TA, Streiner DL, Young LT. Antidepressants and benzodiazepines for major depression. Cochrane Database Syst Rev 2002;(1):CD001026. [59] Loonen AJ, Peer PG, Zwanikken GJ. Continuation and maintenance therapy with antidepressive agents. Meta-analysis of research. Pharm Weekbl Sci 1991;13:167–75. [60] Geddes JR, Carney SM, Davies C, et al. Relapse prevention with antidepressant drug treatment in depressive disorders: a systematic review. Lancet 2003;361:653–61. [61] Whittington CJ, Kendall T, Fonagy P, et al. Selective serotonin reuptake inhibitors in childhood depression: systematic review of published versus unpublished data. Lancet 2004;363: 1341–5. [62] Wagner KD, Ambrosini P, Rynn M, et al. Efficacy of sertraline in the treatment of children and adolescents with major depressive disorder. Two randomized controlled trials. JAMA 2003;290:1033–41. [63] Garland EJ. Facing the evidence: antidepressant treatment in children and adolescents. CMAJ 2004;170:1–4. [64] Khan A, Khan SR, Walens G, et al. Frequency of positive studies among fixed and flexible dose antidepressant clinical trials: an analysis of the food and drug administration summary basis of approval reports. Neuropsychopharmacology 2003;28:552–7. [65] Emslie GJ, Ryan ND, Wagner KD. Major depression disorder in children and adolescents: clinical trial design and antidepressant efficacy. J Clin Psychiatry 2005;66:14–20. [66] Hazell P, O’Connell D, Heathcote D, et al. Tricyclic drugs for depression in children and adolescents. Cochrane Database Syst Rev 2002;(2):CD002317. [67] Chambers CD, Johnson KA, Dick LM, et al. Birth outcomes in pregnant women taking fluoxetine. N Engl J Med 1996;335:1010–5. [68] Addis A, Koren G. Safety of fluoxetine during the first trimester of pregnancy: a metaanalytical review of epidemiological studies. Psychol Med 2000;30:89–94.
940
MAURER & COLT
[69] Wisner KL, Gelenberg AJ, Leonard H, et al. Pharmacologic treatment of depression during pregnancy. JAMA 1999;282:1264–9. [70] Einarson A, Fatoye B, Sarkar M, et al. Pregnancy outcome following gestational exposure to venlafaxine: a multicenter prospective controlled study. Am J Psychiatry 2001; 158:1728–30. [71] Chun-Fai-Chan B, Koren G, Fayez I, et al. Pregnancy outcome of women exposed to bupropion during pregnancy: a prospective comparative study. Am J Obstet Gynecol 2005;192: 932–6. [72] Huntington J, Zantop V. Antidepressant medications in pregnancy. Am Fam Physician 2004;70:2195–6. [73] Koren G, Matsui D, Einarson A, et al. Is maternal use of selective serotonin reuptake inhibitors in the third trimester of pregnancy harmful to neonates? CMAJ 2005;172: 1457–9. [74] Kallen B. Noenate characteristics after maternal use of antidepressants in late pregnancy. Arch Pediatr Adolesc Med 2004;158:312–6. [75] Wulsin LR, Singal BM. Do depressive symptoms increase the risk of for the onset of coronary disease? A systematic quantitative review. Psychosom Med 2003;65:201–10. [76] Glassman AH, O’Connor CM, Califf RM, et al. Sertraline treatment of major depression in patients with acute MI or unstable angina. JAMA 2002;288:701–9. [77] Roose SP, Laghrissi-Thode F, Kennedy JS, et al. Comparison of paroxetine and nortriptyline in depressed patients with ischemic heart disease. JAMA 1998;279:287–91. [78] Thase ME, Tran PV, Wiltse C, et al. Cardiovascular profile of duloxetine, a dual reuptake inhibitor of serotonin and norepinephrine. J Clin Psychopharmacol 2005;25: 132–40. [79] Dugan SE, Fuller MA. Duloxetine: a dual reuptake inhibitor. Ann Phamacother 2004;38: 2078–85. [80] Ramasubbu R. Cerebrovascular effects of selective serotonin reuptake inhibitors: a systematic review. J Clin Psychiatry 2004;65:1642–53. [81] Wiart L, Petit H, Joseph PA, et al. Fluoxetine in early poststroke depression: a double-blind placebo-controlled study. Stroke 2000;31:1829–32. [82] Hackett ML, Anderson CS, House AO. Management of depression after stroke: a systematic review of pharmacological therapies. Stroke 2005;36:1098–103. [83] Bower P, Rowland N, Mellor Clark J, et al. Effectiveness and cost effectiveness of counseling in primary care. Cochrane Database Syst Rev 2006;(3):CD001025. [84] Fava GA, Rafanelli C, Grandi S, et al. Prevention of recurrent depression with cognitive behavioral therapy. Preliminary findings. Arch Gen Psychiatry 1998;55:816–20. [85] Huibers MJ, Beurskens AJ, Bleijenberg G, et al. The effectiveness of psychosocial interventions delivered by general practitioners. Cochrane Database Syst Rev 2003;(2):CD003494. [86] Blumenthal JA, Babyak MA, Moore KA, et al. Effects of exercise training on older patients with major depression. Arch Intern Med 1999;159:2349–56. [87] Dunn AL, Trivedi MH, Kampert JB, et al. Exercise treatment for depression. Efficacy and dose response. Am J Prev Med 2005;28:1–8. [88] Linde K, Mulrow CD. St John’s Wort for depression. Cochrane Database Syst Rev 2005;(2): CD000448. [89] Schmidt PJ, Daly RC, Bloch M, et al. Dehydroepiandrosterone monotherapy in midlifeonset major and minor depression. Arch Gen Psychiatry 2005;62:154–62. [90] S-Adenosyl-L-methionine for treatment of depression, osteoarthritis, and liver disease. Summary, Evidence Report/Technology Assessment: Number 64. 2002. Rockville, Agency for Healthcare Research and Quality, Publication No. 02–E033. [91] Tuunainen A, Kripke DF, Endo T. Light therapy for non-seasonal depression. Cochrane Database Syst Rev 2004;(2):CD004050. [92] Katon W, Russo J, von Korff M, et al. Long-term effects of a collaborative care intervention in persistently depressed primary care patients. J Gen Intern Med 2002;17:741–8.
DEPRESSION: AN EVIDENCE-BASED APPROACH
941
[93] Neumeyer-Gromen A, Lampert T, Stark K, et al. Disease management programs for depression: a systematic review and meta-analysis of randomized controlled trials. Med Care 2004; 42:1211–21. [94] Van der Wurff FB, Stek ML, Hoogendijk WL, et al. Electroconvulsive therapy for the depressed elderly. Cochrane Database Syst Rev 2003;(2):CD003593. [95] Martin JLR, Barbanoj MJ, Schlaepfer TE, et al. Transcranial magnetic stimulation for treating depression. Cochrane Database Syst Rev 2002;(2):CD003493. [96] Institute for Clinical Systems Improvement (ICSI). Major depression in adults in primary care. Bloomington (MN): Institute for Clinical Systems Improvement (ICSI); 2004. p. 78.
Prim Care Clin Office Pract 33 (2006) 943–951
Osteoporosis Milisa K Rizer, MD, MPH Department of Family Medicine, The Ohio State University, 2231 North High Street, Columbus, OH 43201, USA
Osteoporosis, as defined by the National Osteoporosis Foundation, is a disease that is characterized by low bone mass and structural deterioration of bone tissue, which leads to bone fragility and an increased susceptibility to fractures [1]. Aging is only one factor that contributes to the development of osteoporosis. Genetics, suboptimal nutrition, deficiency of calcium and vitamin D, lifestyle, smoking, decrease in sex hormone production, and medications also contribute to skeletal fragility. In the United States, more than 10 million individuals have osteoporosis and more than 34 million have low bone mass or osteopenia of the hip [2]. It has been estimated that 1.5 million individuals suffer a bone disease– related fracture each year [3]. Osteoporosis is responsible for more than 700,000 vertebral fractures and more than 300,000 hip fractures every year [3,4]. Women who are older than the age of 50 years have a 50% chance of suffering a fracture; men of the same age have a 25% risk. Those who have a hip fracture have a 10% to 20% mortality in the first year, and less than 50% regain their prefracture level of mobility and independence [2]. Mortality within 90 days of osteoporotic fractures in individuals who are older than 65 years is substantially higher than predicted; for some fractures, the risk for early death increases by nearly sevenfold [5]. Osteoporotic fractures are a frequent and important cause of disability and medical costs worldwide. Fortunately, osteoporotic fractures are preventable. Guidelines for the prevention, screening, diagnosis, and management of osteoporosis have been established by the US Preventive Services Task Force (USPSTF), the National Osteoporosis Foundation, Surgeon General, American College of Obstetrician and Gynecologists, American Association of Clinical Endocrinologists, and the Osteoporosis Society of Canada. Although some are consistent and similar, others are not.
E-mail address:
[email protected] 0095-4543/06/$ - see front matter Ó 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.pop.2006.09.004 primarycare.theclinics.com
944
RIZER
Prevention Exercise is an important component of any osteoporosis prevention program. All types of physical activity can contribute to bone health. Activities that are weight bearing or involve impact are most useful for increasing or maintaining bone mass [2]. Recent small studies found that the use of vibrating platforms increased bone mineral density (BMD) and slowed bone loss [2]. A systematic review of randomized trials shows that impact and nonimpact exercises have a positive effect at the lumbar spine in pre- and postmenopausal women. Impact exercise probably has a positive effect at the femoral neck. More studies are required to determine the optimal intensity and type of exercise [6]. Throughout life, men and women should be encouraged to participate in exercise, particularly in weight-bearing exercises, which include impact as a component. A review of eight trials found that exercise programs showed a trend toward prevention of bone loss in the lumbar spine with impact and nonimpact exercise [6].
Calcium Calcium supplementation has a small, positive effect on bone density. The data show a trend toward reduction in vertebral fractures, but do not address, in a meaningful way, the possible effect of calcium on reducing the incidence of nonvertebral fractures [7]. Hip fractures are related strongly to low BMD, cost more to repair, and cause more disability than does any other type of osteoporotic fracture [8]. A study by Dawson-Hughes and colleagues [9] concluded that healthy older postmenopausal women with a daily calcium intake of less than 400 mg can reduce bone loss significantly by increasing their calcium intake to 800 mg per day. They also showed that supplementation with calcium citrate maleate was more effective than was supplementation with calcium carbonate at the doses used in the study.
Vitamin D Vitamin D decreases vertebral fractures and may decrease nonvertebral fractures. A meta-analyses of the efficacy of vitamin D treatment in preventing osteoporosis in postmenopausal women showed that it reduced the incidence of vertebral fractures [relative risk (RR), 0.63; 95% CI, 0.45–0.88; P ! .01], and showed a trend toward reduced incidence of nonvertebral fractures [RR, 0.77; 95% CI, 0.57–1.04; P ¼ .09] [10]. The use of standard or hydroxylated vitamin D has not been studied well. In a study of 3270 healthy ambulatory women with a mean age of 84 6 years, 1634 received 800 IU/d of vitamin D3 along with 1200 mg of elemental calcium, and 1632 received double placebo. The bone density of the proximal femur increased 2.7% in the group that received vitamin D and decreased 4.6% in the group
945
OSTEOPOROSIS
that received placebo (P ! .001). The number of hip fractures was 43% lower (P ! .043) and the total number of nonvertebral fractures was 32% lower (P ! .015) in the group that took vitamin D plus calcium [11]. Randomized controlled trials indicate that supplemental calcium and vitamin D reduce the risk for hip fractures and other nonvertebral fractures in elderly women [4]. Table 1 outlines the current recommended amounts of calcium and vitamin D in the diet.
Screening Risk factors The USPSTF recommends that women aged 65 years and older be screened routinely for osteoporosis, although the USPSTF does not define ‘‘routinely.’’ This screening should begin at age 60 years for women who are at increased risk for osteoporotic fractures [2]. To estimate the benefits of routine screening for women in different age groups, the USPSTF used estimates from recent studies to project the number of fractures that would be prevented over 5 years from a hypothetic cohort of 10,000 postmenopausal women [12]. For women 65 to 69 years of age, the numbers needed to screen were 731 to prevent one hip fracture and 248 to prevent one vertebral fracture in 5 years [13]. For women with low bone density, the number needed to screen was 88 to prevent one hip fracture and 30 to prevent one vertebral fracture. The number needed to screen became more favorable as age advanced [14]. In addition, it found three clinical risk factors that consistently predicted increased risk for fracture: advanced age, low weight or body mass index, and nonuse of hormone replacement therapy (HRT). The presence of any of the three risk factors increased the risk for fracture by 70% (RR, 1.7) [13,14]. The USPSTF makes no recommendation for or against routine osteoporosis screening in postmenopausal women who are younger than 60 years or Table 1 Recommended daily dosages of calcium and vitamin D Age group
Calcium (mg/d)
[SOR]
Vitamin D (IU/d)
[SOR]
Prepubertal children (ages 4–8 y) Adolescents (ages 9–18 y) Women (ages 19–50 y) Women over 50 y Pregnant or lactating women (R18 y) Men (ages 19–50 y) Men over 50 y
800 1300 1000 1500 1000 1000 1500
[B] [B] [A] [A] [A] [C] [C]
No data No data 400 800 400 400 800
[D] [A] [D] [D] [A]
Abbreviation: SOR, strength of recommendation. Data from Whiting SJ, Calvo MS. Dietary recommendations for vitamin D: a critical need for functional end points to establish an estimated average requirement. J Nutr 2005;135:304–9.
946
RIZER
in women who are aged 60 to 64 years who are not at increased risk for osteoporotic fractures [2]. No study has evaluated the effect of screening in reducing fractures in this younger population [4,13]. Although several studies have tested screening tools, the Osteoporosis Society of Canada recommends targeted case finding strategies for those at increased risk, using at least one major or two minor risk factors (Box 1), along with BMD measurement with central dual-energy x-ray absorptiometry (DEXA) at age 65 years [13]. Low bone BMD along with the major risk factors of previous fragility fracture, age, and family history of osteoporosis stand out as predictors of fracture related to osteoporosis. Clinically, a fragility fracture may be defined as one that occurs as a result of minimal trauma, such as a fall from a standing height or less, or no identifiable trauma [15]. These risk factors
Box 1. Risk factors for osteoporosis Major risk factors Age older than 65 years Vertebral compression fracture Fragility fracture after age 40 Family history of osteoporotic fracture Systemic glucocorticoid therapy of longer than 3 months Malabsorption syndrome Primary hyperparathyroidism Propensity to fall Osteopenia apparent on radiograph Hypogonadism Early menopause before age 45 Minor risk factors Rheumatoid arthritis Low dietary calcium intake Chronic anticonvulsant therapy History of clinical hyperthyroidism Smoking Excessive alcohol intake Excessive caffeine intake Weight less than 57 kg Chronic heparin therapy Weight loss greater than 10% of weight at age 25 Data from Brown JP, Josse RG. 2002 Clinical practice guidelines for the diagnosis and management of osteoporosis in Canada. CMAJ 2002;167:S1–34.
OSTEOPOROSIS
947
have a cumulative effect such that, for example, if a person has a low BMD in addition to a fragility fracture or is older than 65 years of age and has a BMD in the range associated with osteoporosis, he or she should be considered to be at high risk for fracture and a candidate for therapy. A meta-analysis of 29 published cross-sectional studies of smoking and bone density indicated that postmenopausal bone loss was greater in current smokers than in nonsmokers [4]. Other studies found that the risk for hip fracture was higher for thinner smokers than for normal or overweight smokers [4]. Alcohol use is an inconsistent predictor of bone mass and fractures [4]. Caffeine intake is associated inconsistently with low bone density and fractures [4]. Another reasonable recommendation is to screen postmenopausal women who are younger than 65 years who have low weight (or body mass index) or who have never used HRT [16]. People who receive 7.5 mg of prednisone daily for more than 3 months should be assessed for initiation of a bone-sparing therapy. People who receive more than 2.5 mg of prednisone daily should be regarded as being at increased risk for fragility fracture and require further assessment, with at least BMD measurement [15]. Dual-energy x-ray absorptiometry scanning DEXA is the technical standard for measuring BMD because it measures at important sites of osteopathic fractures, has high precision and accuracy, is inexpensive, and has modest radiation exposure [17]. Typically, measurement of BMD is taken at two locations, spine and hip, and reported as a Tscore and a Z-score. The T-score is the difference between an individual’s BMD and the mean BMD for a reference population. It is expressed in standard deviation units. A score of 0 indicates a BMD equal to the mean, a score of þ1 indicates one standard deviation above the mean, and a score of 1 is one standard deviation below the mean. Osteoporosis is defined as a T score that is less than 2.5 [16]. The Z-score is used to compare an individual with others in the same age group [16]. Measuring BMD remains the single best predictor of fracture risk available [2]. Central (hip and spine) DEXA is the most accurate tool for evaluating BMD in clinical settings [15]. Access to BMD measurement should not be limited by decision tools based on clinical risk factors [15]. Screening frequency Optimal screening frequency has not been studied, but the USPSTF suggests a frequency of not more than every 2 years for older women or every 5 years for younger postmenopausal women [14]. For younger individuals who initially were screened based upon risk factors but who have normal BMD values, repeat testing every 5 to 10 years may be helpful [14]. Those who have a borderline low BMD or may lose bone rapidly (eg, exposure
948
RIZER
to high doses of glucocorticoids) require repeat screening in 2 to 3 years [15]. DEXA is useful in monitoring patients who are on medical therapy for osteoporosis [15]. Patients should be rescanned using central DEXA scanning 2 years after initiating medication therapy. Ultrasound Quantitative ultrasound may assist in the diagnosis of osteoporosis, but is not useful for monitoring. Sound waves that are used to assess bone mass do not emit radiation. It can measure bone density in a variety of peripheral sites. Most devices use a formula to calculate a bone density equivalent and are not sufficiently precise for follow-up or monitoring [15].
Treatment The National Osteoporosis Foundation recommends starting pharmacologic intervention for those patients who are determined to be at increased risk for an osteoporotic-related fracture: patients with a T-score of 2.0 or lower by hip DEXA with no risk factors, those with a T-score of 1.5 or lower and one or more risk factors, or those with a previous hip or vertebral fracture [1]. Antiresorptives or anabolic agents can be used. Antiresorptives include HRT, bisphosphonates, calcitonin, and selective estrogen receptor modulator. The only anabolic agent available is teriparatide. There is convincing, patient-oriented evidence for nonvertebral fracture reduction for only two agents: risedronate (Actonel) and alendronate (Fosamax) [18]. Bisphosphonates The bisphosphonates are a first-line preventive therapy in postmenopausal women with low bone density, and first-line treatment for postmenopausal women who have osteoporosis, especially those with pre-existing vertebral fractures [15]. For weekly dosing regimens, once-weekly alendronate is slightly better than is once-weekly risedronate in increasing the bone density in the trochanter and spine [19]. Bisphosphonates also are the first-line therapy for prevention of glucocorticoid-induced osteoporosis and for the treatment of glucocorticoid-induced osteoporosis in patients who require prolonged glucocorticoid therapy [15]. Bisphosphonates also are indicated as first-line treatment for men who have low bone mass or osteoporosis [15]. In premenopausal women who have osteopenia or osteoporosis, the use of bisphosphonates has not been examined, and is not recommended in the absence of an identified secondary cause of osteoporosis [15]. Table 2 outlines the important uses of the bisphosphonates in the treatment of osteoporosis.
949
OSTEOPOROSIS
Table 2 Bisphosphonates in the treatment of osteoporosis Strength of recommendation Condition
Alendronate
Risedronate
Etidronate
Preventive therapy in postmenopausal women with low bone density Treatment of postmenopausal women who have osteoporosis Preventive therapy for glucocorticoid-induced osteoporosis Treatment of glucocorticoid-induced osteoporosis Treatment for men with low bone mass
A
A
No data
A
A
B
A
A
B
A
A
B
A
No data
B
Data from Brown JP, Josse RG. 2002 Clinical practice guidelines for the diagnosis and management of osteoporosis in Canada. CMAJ 2002;167:S1–34.
Calcitonin Nasal or parenteral calcitonin is a first-line treatment for pain associated with acute vertebral fractures [15]. It is considered as a second-line treatment for postmenopausal women who have osteoporosis [15]. Nasal calcitonin was shown to increase BMD and reduce lumbar spine fractures by 36%, but has not been shown to reduce hip fractures [20]. Hormone replacement therapy HRT is a first-line preventive therapy in postmenopausal women with low bone density [15]; however, when used only for the prevention of postmenopausal osteoporosis, the risks of HRT may outweigh the benefits. HRT is a first-line preventive therapy for women who experience menopause before age 45 [15]. It is a second-line treatment for postmenopausal women who have osteoporosis [15]. With prolonged use of HRT taken only for the treatment of postmenopausal osteoporosis, the substantial risks for cardiovascular disease, stroke, and invasive breast cancer may lead to an unfavorable risk-benefit ratio [17]. It should be administered for the shortest period at the lowest possible dose. Raloxifene Raloxifene (Evista) is a selective estrogen receptor modulator that is designed to provide the benefits of estrogen on BMD with a lower risk for breast cancer, endometrial cancer, and cardiovascular disease [20]. Raloxifene is a first-line therapy in the prevention of further bone loss in postmenopausal women with low bone density, and a first-line treatment for osteoporosis in postmenopausal women [15]. Raloxifene was shown to increase BMD and reduce the risk for vertebral fractures, but there is no evidence that it reduces nonvertebral fractures [20]. The adverse effects that are
950
RIZER
associated with raloxifene include venous thromboembolic disease, pulmonary embolism, and hot flashes [20]. Teriparatide Teriparatide (Forteo) is a once-daily, subcutaneously administered anabolic agent that stimulates osteoblastic bone formation at trabecular and cortical sites [16]. It is approved for treatment of osteoporosis in postmenopausal women who are at high risk for fracture or for whom other therapies have failed. In conjunction with adequate calcium and vitamin D, it produces increases in bone mass of 10% to 15% per year [5,16]. It also reduces all vertebral fractures by approximately 67% and nonvertebral fracture rates by approximately 50%. Parathyroid hormone Parathyroid hormone (PTH) is tolerated well, although some patients experience leg cramps and dizziness. Because PTH caused an increase in the incidence of osteosarcoma in rats, patients with an increased risk for osteosarcoma (eg, patients who have Paget’s disease of bone, previous radiation therapy of the skeleton, bone metastases, hypercalcemia, or a history of skeletal malignancy) should not receive PTH therapy. The safety and efficacy of PTH have not been demonstrated beyond 2 years of treatment [20,21]. Table 3 summarizes the options for the screening, prevention, and treatment of osteoporosis. Table 3 Summary of recommendations for the screening, prevention and treatment of osteoporosis Recommendation Treatment should be initiated to reduce fracture risk in postmenopausal women who have experienced a fragility or low-impact fracture. [1,15] Treatment should be instituted in those postmenopausal women with BMD T scores less than 2 by central DEXA in the absence of risk factors and in women with T scores less than 1.5 in the presence of one or more risk factors. [1,15] First-line pharmacologic options determined by the FDA to be safe and effective for osteoporosis prevention should be used. [1,15] First-line pharmacologic options determined by the FDA to be safe and effective for osteoporosis treatment should be used. [1,14] Women should be counseled about the following preventive measures: adequate calcium, vitamin D, exercise, smoking cessation, moderate alcohol intake, and fall prevention strategies. [1,15] BMD testing should be recommended to all postmenopausal women 65 y Lears of age or older. [1,2,12] BMD testing should be recommended for postmenopausal women younger than 65 years of age who have one or more risk factors for osteoporosis. [1,2,12]
Strength of recommendation A A
A A B
B B
OSTEOPOROSIS
951
References [1] National Osteoporosis Foundation. Fast facts about osteoporosis. Available at: www.nof. org/osteoporosis/diseasefacts.htm. Accessed January 23, 2006. [2] US Department of Health and Human Services. Bone health and osteoporosis: a report of the Surgeon General. Available at: www.surgeongeneral.gov/library/bonehealth. Accessed January 22, 2006. [3] Riggs BL, Melton LJ III. The worldwide problem of osteoporosis: insights afforded by epidemiology. Bone 1995;17(5 Suppl):505S–11S. [4] HSTAT. Guide to clinical preventive services. 3rd edition: recommendations and systematic evidence reviews, guide to community preventive services. Available at: www.ncbi.nlm.nih. gov. Accessed January 29, 2006. [5] Heaney RP. Advances in therapy for osteoporosis. Clin Med Res 2003;1(2):93–9. [6] Wallace BA, Cumming RG. Systematic review of randomized trials of the effect of exercise on bone mass in pre- and postmenopausal women. Calcif Tissue Int 2000;67(1):10–8. [7] Shea B, Wells G, Cranney A, et al. Meta-analyses of therapies for postmenopausal osteoporosis. VII. Meta-analysis of calcium supplementation for the prevention of postmenopausal osteoporosis. Endocr Rev 2002;23(4):552–9. [8] Cummings S, Melton LJ III. Epidemiology and outcomes of osteoporotic fractures. Lancet 2002;359:1761–67. [9] Dawson-Hughes B, Dallal GE, Krall EA, et al. A controlled trial of the effect of calcium supplementation on bone density in postmenopausal women. N Engl J Med 1990;323(13): 878–83. [10] Papadimitropoulos E, Wells G, Shea B, et al. Meta-analyses of therapies for postmenopausal osteoporosis. VII. Meta-analysis of the efficacy of vitamin D treatment in preventing osteoporosis in postmenopausal women. Endocr Rev 2002;23(4):560–9. [11] Chapuy MC, Arlot ME, Duboeuf F, et al. Vitamin D3 and calcium to prevent hip fractures in the elderly women. N Engl J Med 1992;327(23):1637–42. [12] The US Preventive Services Task Force. Screening for osteoporosis in postmenopausal women: recommendations and rationale. Available at: www.ahrq.gov/clinic/3rdduspstf/ osteoporosis/osteorr.htm. Accessed December 27, 2005. [13] Nelson H, Helfand M, Woolf S, et al. Screening for postmenopausal osteoporosis: a review of the evidence for the US Preventive Services Task Force. Ann Intern Med 2002;137:529–41. [14] Cronholm P. Densitometry identifies women in whom treatment will reduce fracture risk. J Fam Pract 2003;52(2). [15] Brown J, Josse R. 2002 Clinical practice guidelines for the diagnosis and management of osteoporosis in Canada. CMAJ 2002;167(Suppl 10):S1–34. [16] South-Paul JE. Osteoporosis: Part 1. Evaluation and assessment. Am Fam Physician 2001; 63:897–904, 908. [17] Neff M. Practice guidelines: ACOG releases guidelines for clinical management of osteoporosis. Available at: www.aafp.org/afp/20040315/practice.html. Accesed February 5, 2006. [18] Cranney A, Guyatt G, Griffith L, et al. Meta-analysis of osteoporosis therapies for postmenopausal osteoporosis. Endocr Rev 2002;23(4):570–8. [19] Allen ML, Watt L. Guidelines for the diagnosis, screening and treatment of osteoporosis in women. Adv Stud Med 2005;5(10):518–23. [20] National Osteoporosis Foundation. Pharmacologic options for drug treatment of osteoporosis. Accessed at www.nof.org/physguide/pharmacologic.htm on 1/29/2006. [21] Rosen CJ, Hochberg MC, Bonnick SL, et al. Treatment with once-weekly alendronate 70 mg compared with once-weekly risedronate 35 mg in women with postmenopausal osteoporosis: a randomized double-blind study. J Bone Miner Res 2005;20(1):141–51.
Prim Care Clin Office Pract 33 (2006) 953–963
Hormone Replacement Therapy Pamela Dull, MD OSU Department of Family Medicine, 1492 East Broad Street, Suite 1302, Columbus, OH 43205, USA
With many women in or near menopause and multiple news stories of hormone risks, there has been renewed interest in the treatment of menopause. Vasomotor symptoms or hot flashes occur in 14% to 51% of women before menopause, in 35% to 50% in the perimenopausal period, and in 30% to 80% after menopause [1]. Menopause and its transition time also may cause vaginal dryness or painful intercourse, sleep disturbances, mood changes, and other somatic complaints. Recent large studies have shown that hormone replacement therapy (HRT) should not be used for disease prevention and have shown risks that may offset their benefits. With the United States’ baby-boom generation approaching menopause, many women are asking their doctors about the usefulness of HRT and whether they should use alternatives. This article reviews the studies that have shaped our current perspective and outlines the recommendations of major medical organizations.
Evidence summary Women’s Health Initiative HRT has been available for many years. The Women’s Health Initiative (WHI) studies introduced concerns that challenge the routine use of HRT. The WHI study involved more than 16,000 postmenopausal women, aged 50 to 79 years. It was randomized and controlled at 40 different clinical centers across the United States. All patients who were on any postmenopausal treatment had a required 3-month ‘‘wash out’’ period of no medication before being randomized for trial. Those with a uterus received placebo or a combination pill of 0.625 mg of conjugated equine estrogen (CEE)/2.5 mg of medroxyprogesterone (MPA) daily. The smaller E-mail address:
[email protected] 0095-4543/06/$ - see front matter Ó 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.pop.2006.09.008 primarycare.theclinics.com
954
DULL
subpopulation of women without a uterus was randomized to placebo or CEE, 0.625 mg. The original plan was an 8-year study to examine effect on secondary diseases; however, combination HRT treatment was terminated early at 5.2 years because of an increased rate of breast cancer [2]. For every 10,000 women per year who were taking combination HRT, there were eight more invasive breast cancers, seven more coronary heart disease (CHD) events, eight more strokes, and eight more pulmonary embolisms than in their counterparts who were taking placebo. Despite the benefits of six fewer colon cancers and five fewer hip fractures per 10,000 women [2–5], labeling of combination HRT preparations was changed in 2002 directing that these products be used only for symptom relief and not for the prevention of disease. Because the biggest increase in breast cancer occurred after 5 years of use, physicians and patients were encouraged to use HRT for the shortest amount of time possible. The additional risks were per 10,000 women per year, which made the likelihood for an event to occur in the average woman sitting in your office to be extremely unlikely. In 2004, estrogen-only (CEE) treatment in women who previously had undergone hysterectomy was stopped because of an increase in adverse events. This arm demonstrated an additional 12 strokes per 10,000 women per year. There was a 23% reduction in breast cancers in this group and six fewer hip fractures per 10,000 women. There was no effect on colorectal cancer rates or CHD events, and a minimal increase in pulmonary embolism [6,7]. The limitations of the WHI trials are that they included older women who were postmenopausal for many years. The results may not extrapolate to younger or more newly menopausal women. One small study that followed the WHI demonstrated a shift of patient attitudes away from using HRT [8]. In this study, 100% of the random subjects had heard of the WHI studies and 52% reported that the study changed their use of HRT. Fifty-two percent also reported that they were less likely to trust information that their doctors gave them about HRT [8]. Of those who stopped therapy because of the WHI report, another study reported that 25% resumed therapy despite the risks [1]. A later subset of the WHI trial, the Women’s Health Initiative Memory Study, investigated the rates of dementia and cognitive impairment in women who were taking combination HRT or CEE alone. Although there was no apparent worsening in cognitive impairment, the risk for dementia was twofold higher in the group that was taking combination HRT. This translates into 23 more cases of dementia per 10,000 women per year [9]. The hazard ratio was 1.76 in women after hysterectomy who were taking CEE alone. Cognitive impairment was slightly higher in women who were taking CEE compared with placebo, with a hazard ratio of 1.34 [10]. Therefore, based upon this study, neither CEE nor combination HRT should be used for the prevention or treatment of dementia or cognitive impairment [9,10].
HORMONE REPLACEMENT THERAPY
955
Million Women Study The Million Women Study in England between 1996 and 2001 studied the prevalence of breast and uterine malignancies among women who were and were not taking HRT [11,12]. There was an increased risk for breast cancer (relative risk [RR], 1.66; P!.0001) and an increased risk for death from that cancer (RR, 1.22; P ¼ .05) [11]. Past use of HRT did not increase breast cancer or its mortality. There was no difference between preparations or dose. The duration of therapy is important. After 10 years of use, there were 19 additional breast cancers per 1000 women using combination HRT and five additional breast cancers per 1000 women using estrogen-only therapy [11]. Women using combination HRT had a lower risk for endometrial cancer (RR, 0.71) as compared with those who had never used HRT. Estrogen-only therapy, however, increases this risk (RR, 1.43) [12]. There is a greater risk for breast and endometrial cancer in women who use combined HRT, with no significant difference between continuous or cyclic use [11,12]. Netherlands Trial After the Million Women Study and WHI trials results were published, a small, unnamed Netherlands study found a smaller decrease in HRT prescribing rates after the WHI trial, and a significantly larger decrease after the Million Women Study. In 2000, the HRT prescribing rate was 107 per 1000 women aged 45 to 69 years; it decreased to 87 per 1000 in 2003. Although previous studies discouraged the long-term use of combination HRT, the proportion of long-term HRT users (O3 years) remained unchanged [13]. This indicates that although women are aware of the risks of hormone treatment, they feel comfortable continuing treatment. Other studies and the type of breast cancer Although the WHI and the Million Women Study showed increased breast cancer rates in users of combined HRT, the breast tumors that were found in the WHI were less localized and were associated with more positive lymph nodes [14]. A subsequent analysis of the data showed improved survival rates in postmenopausal women who were taking HRT and subsequently diagnosed with breast cancer [15]. Some investigators postulate that HRT makes patients’ breasts more dense, which makes mammography more difficult to read, and, thus, delays the diagnosis [15]. A subsequent study confirmed a reduction in mammographic sensitivity, although it also reported smaller, more differentiated (grade I) breast tumors among HRT users versus nonusers [16]. Another study showed an increased rate of breast cancer in HRT users, but less invasive disease with better survival (0% mortality in the group that took HRT versus 13% mortality in the group that did not take HRT at 8 years) [15]. Some investigators postulate
956
DULL
that HRT may increase the growth of pre-existing tumors to the point of earlier detection [15]. If there are concerns of breast tissue density interfering with examination and mammography, HRT should be stopped 2 weeks before the study [15].
Guidelines and recommendations from national organizations United States Preventative Services Task Force The Unites States Preventive Services Task Force (USPSTF) found good evidence that combined HRT use reduces rates of fracture and fair evidence that it reduces rates of colon cancer [17]. It concluded that there is good evidence for increased risk for coronary heart disease, venous thromboembolism, and breast cancer in HRT users. There is fair evidence of increased risk for stroke, cholecystitis, dementia, and diminished cognitive function. There is good evidence that estrogen alone reduces the number of fractures. There is fair evidence that it increases the risk for venous thromboembolism, stroke, and dementia and decreases cognitive function. The USPSTF found insufficient evidence to assess its effects on the incidence of ovarian cancer, mortality from breast cancer or coronary heart disease, or all-cause mortality. There also is insufficient evidence to assess colon or breast cancer risk in users of unopposed estrogen. The USPSTF recommends against the routine use of HRT for the prevention of chronic conditions in postmenopausal women (D-recommendation: good to fair evidence of many risks or possible harms may outweigh the benefits). The USPSTF recommends that physicians discuss the benefits and risks of therapy with their patients who have symptoms. Unopposed estrogen only should be considered in women without a uterus [17]. American College of Obstetricians and Gynecologists The American College of Obstetricians and Gynecologists (ACOG) does not recommend estrogen-only or combination HRT be used for disease prevention. Hormonal therapy is still the most effective treatment of menopausal symptoms, but a woman should only use HRT if the benefits outweigh the risks. The smallest effective dose should be used for the shortest time needed with yearly assessment for continued use [18]. North American Menopause Society The North American Menopause Society (NAMS) created an evidencebased position statement stating that prescription systemic estrogen products are still the standard for moderate to severe menopause-related hot flashes. Clinicians are advised to review the risks and benefits of this therapy with patients. Mild symptoms should be treated with lifestyle changes, either
HORMONE REPLACEMENT THERAPY
957
alone or combined with nonprescription therapy. Alternatives include progestins, venlafaxine, paroxetine, fluoxetine, and gabapentin. Treatment with HRT should be reassessed periodically with the understanding that vasomotor symptoms usually abate spontaneously in time [19]. Estrogen-only therapy is only recommended for women without a uterus. The shortest duration and smallest effective dose should be used [20].
Low-dose therapy After the WHI and national guidelines releases, studies were designed to examine lower doses of estrogen and estrogen combined with progestin. The Women’s Health, Osteoporosis, Progestin, Estrogen Study confirmed that lower-dose estrogen combined HRT was as effective as traditional doses. In estrogen-only therapy, the traditional 0.625-mg dosage of CEE was slightly more effective than were lower dosages for the treatment of hot flashes. Vaginal atrophy improved similarly in all treatment groups [21]. Another study found that the lower-dose estrogen combination HRT had fewer episodes of any type of vaginal bleeding than did the common 0.625-mg CEE/2.5-mg MPA combination HRT [22]. There was no difference in endometrial cancer rates between lower-dose estrogen regimens and standard-dose regimens over a 2-year period [23]. Low-dose estrogen HRT alternatives include 0.45-mg CEE/1.5-mg MPA, 0.3-mg CEE/1.5-mg MPA, and 2.5-mg ethinyl estradiol/0.5-mg norethindrone acetate. Low-estrogen alone formulations include 0.3-mg CEE, 0.45-mg CEE, and 0.50-mg estradiol. A new low-dose combination of 0.5-mg estradiol/1-mg drospirenone was approved recently for use by the US Food and Drug Administration (FDA) [24]. An alternative to prepackaged combination hormone therapy is to add a cyclic progestin (2.5–5 mg MPA daily on days 16 to 25 of each month) to an estrogen (eg, 0.5 mg of daily estradiol) [25]. Cyclic therapy is not used frequently, mainly because of its relative dosing complexity, adherence issues, and increased incidence of vaginal bleeding. Cyclic therapy of less than 12 days of progestin showed increased uterine hyperplasia changes, which suggested possible uterine precancerous changes [25].
Nonoral estrogens Few data exist that compare topical or transdermal HRT with systemic therapy [26]. A recent review found no significant differences between CEE, oral 17-b-estradiol, and transdermal 17-b-estradiol in the treatment of hot flashes [27]. The most commonly reported side effects of transdermal preparations are breast tenderness and vaginal bleeding. More women discontinued using the transdermal preparations because of skin irritation
958
DULL
[28]. Currently available preparations include various biweekly and weekly doses of estradiol (eg, Vivelle, Esclim, Alora, Climara). Estrogen-only patches only should be used in women without a uterus. Two combination HRT transdermal patches are available: estradiol, 0.05 mg/d, combined with norethindrone acetate, 0.14 mg/d or 0.25 mg/d (CombiPatch) and estradiol, 0.45 mg/d, combined with levonorgestrel, 0.15 mg/d (Climara Pro). Topical hormonal gels are comparable to patch HRT, but have fewer skin rashes [28]. The two products available are Estrasorb, an emulsion with estradiol, 0.05 mg, which is supplied in a foil packet that is applied to a leg daily, and EstroGel, a gel with estradiol, 0.75 mg, which is delivered by way of a nonaerosol meter-dose pump that is applied to one arm daily. A progestin should be added if the patient has an intact uterus. Although vaginal estrogens may not be useful for systemic symptoms, they are useful in women who have vaginal dryness and dyspareunia [29,30]. Intravaginal 25-mg 17-b-estradiol tablets and 1.25-mg CEE vaginal creams are equally effective in relieving vaginal symptoms. The vaginal tablets are preferred over intravaginal creams by most women [29]. Intravaginal therapy is safe. A Cochrane review showed no significant differences in rates of uterine hyperplasia between the types of intravaginal preparations [30]. This review also suggested that patients prefer the vaginal estrogen ring because of its ease of use and comfort. Available vaginal estrogen rings include 50- or 100-mg estradiol/d (Femring) and 2-mg estradiol/90 d (Estring). Both products are inserted in the vagina every 3 months. Other available intravaginal preparations include estradiol, 0.1 mg (Estrace Vaginal Cream); CEE, 0.625 mg (Premarin Vaginal Cream); and estradiol, 0.025 mg (Vagifem tablets). As a guide when choosing an estrogen therapy, Box 1 outlines approximate estrogen dose-equivalents of the available preparations. The relative potency equivalents were determined by uncommon analysis of variable data [31].
Box 1. Approximate estrogen dose equivalents 0.625 mg CEE (or esterified estrogen or estrone sulfate) 1 mg micronized 17-b-estradiol 1.25 mg piperazine estrone sulfate 5 to 10 mg ethinyl estradiol 50 mg/d 17-b-estradiol transdermal patch (or 0.50-mg estradiol patch) Data from Mashchak CA, Lobo RA, Dozono-Takano R, et al. Comparison of pharmacodynamic properties of various estrogen formulations. Am J Obstet Gynecol 1982;144:511–8.
HORMONE REPLACEMENT THERAPY
959
Testosterone Testosterone may be useful in place of progestins for treating decreased libido that often is associated with menopause [1,32]. This issue is controversial and has not been studied well. NAMS recommends that testosterone should be used only in the postmenopausal period to enhance sexual drive [32]. Transdermal and topical formulations are preferred. Contraindications include breast cancer, uterine cancer, the presence of cardiovascular disease, and liver disease. Testosterone therapy should be used at the lowest dose for the shortest time to relieve symptoms [32].
Alternatives to hormonal therapy Bioidenticals Given the negative press on HRT, many women are considering more ‘‘natural’’ therapies. ‘‘Bioidentical’’ botanical and dietary supplement products may be considered more ‘‘natural’’ to consumers, but they are not regulated nor standardized. One study of 500 perimenopausal and menopausal women found that 70% reported using some type of botanical or dietary supplement. This same study found that the same 70% did not report their use to their clinician [33]. Given such common usage, clinicians need to be familiar with the more commonly used products. Bioidentical hormones are treatments (eg, wild yam creams) that are individually compounded recipes of a variety of steroids in various dosage forms based on a person’s salivary hormone concentration. NAMS reviewed multiple studies and systemic reviews on ‘‘bioidentical’’ hormone therapy; they concluded that these products may decrease some menopause symptoms, but offer no advantages over conventional therapies and are not supported [34]. ACOG also does not support their use [18]. The National Institutes of Health (NIH) indicates that there are few good-quality data available on these compounds [1]. Botanicals Black cohosh is an herb that may decrease menopausal symptoms by 25% to 30% compared with placebo [35]. A small 6-month trial showed no significant effect on endometrial thickening, vaginal maturation index, or hormonal blood levels [36]. Given reports of possible estrogenic effects, adverse effects on liver function, and no controlled long-term trials, the German Commission E and ACOG recommend that the herb be used to alleviate menopausal symptoms for only up to 6 months [35,37]. Evening primrose has been used for many menstrual and menopause symptoms through the years; however, there are few controlled data to support its use. Reviews done by ACOG found evening primrose to be no better
960
DULL
than placebo in controlling menopausal symptoms [37]. The NIH did not find enough data to support or discourage its use [1]. Dong quai, a Chinese root that has been used for many gynecologic complaints for centuries, has little effect on menopausal symptoms. It also contains coumarins, which may potentate the anticoagulant effect of warfarin [38]. It contains safrole, an oil that has been shown to be carcinogenic [38]. ACOG did not find enough evidence to support its use and expressed concern over potential toxicity [37]. The NIH found little evidence to support its use [1]. Dietary supplements Interest in soy and other phytoestrogenic isoflavones began after epidemiologic studies demonstrated that Asian women reported fewer vasomotor symptoms; had lower rates of cardiovascular disease, breast cancer, and uterine cancer; and had much higher soy content in their diet [39]. Soy products are not standardized and differ in composition between trials, however. Many clinical reviews of trials have shown varying effects on menopause symptoms. There is no significant effect on menopause symptoms from soy foods, soy extracts, or red clover extracts [40]. Other studies have found minimal or no effect of soy extracts on menopausal symptoms and have noted concerns about the wide variation of product composition and dosing in clinical trials [41]. The NIH position is that the soy extracts may have minimal, if any, effect on symptom relief [1].
‘‘Off-label’’ therapy Patients with contraindications to the use of estrogens (Box 2) may seek prescription treatments that are ‘‘off-label’’ and not approved by the FDA for the primary treatment of menopause [24]. Clonidine has shown mild effectiveness against hot flushes; however, its blood pressure lowering effect
Box 2. Contraindications to hormone replacement therapy Undiagnosed abnormal genital bleeding Known, suspected, or history of breast cancer Known or suspected estrogen-dependent neoplasia Active deep vein thrombosis, pulmonary embolism, or history of either Active or recent (within past year) arterial thrombosis (eg, stroke or myocardial infarction) Liver disease or dysfunction Data from http://www.fda.gov. Accessed December 14, 2005.
961
HORMONE REPLACEMENT THERAPY
Table 1 Summary of evidence-based recommendations Recommendation Do not use HRT routinely to prevent chronic disease. HRT may increase breast cancer, but it is unclear if duration, formulation, or dosage may be the variable. Lower doses of hormones can relieve many women’s symptoms with minimal side effects. Use the lowest dose of HRT for the shortest time needed to control symptoms. Alternatives to hormone therapy may be helpful and considered. Black cohosh may be recommended for up to 6 mo to relieve postmenopausal symptoms. ‘‘Bioidenticals’’, such as wild yam creams, are not recommended for postmenopausal symptoms. Paroxetine, venlafaxine, and gabapentin are moderately effective against hot flushes.
Strength of recommendation A B B A C C C A
and central nervous system side effects limit its use [42]. Although centrally active belladonna products are no longer available in the United States, researchers have turned to antidepressants and gabapentin. Initial pilot and larger placebo-controlled trials showed that paroxetine, venlafaxine, and gabapentin were moderately effective in preventing hot flushes and were tolerated well [42]. Other antidepressants have had mixed results in small pilot studies [42].
Summary HRT should not be used for the prevention or treatment of chronic disease (eg, heart disease, osteoporosis, dementia). HRT is effective in alleviating moderate to severe menopausal symptoms. Clinicians must be aware of the risks and benefits of HRT and discuss them thoroughly with their patients. As new forms and lower doses of hormones become available, additional studies will be needed to compare therapeutic strategies adequately. Studies of younger perimenopausal and menopausal women are needed. Whatever regimen is chosen, the lowest allowable estrogen dose to relieve symptoms for the shortest time is recommended. Alternatives to hormonal therapy are being examined and some may be considered. Table 1 provides an evidence-based summary of current recommendations.
References [1] NIH State-of-the-Science Panel. National Institutes of Health state-of-the-science conference statement: management of menopause-related symptoms. Ann Intern Med 2005; 142(12) (Part 1):1003–13.
962
DULL
[2] Rossouw JE, Anderson GL, Prentice RL, et al, for the Writing Group for the Women’s Health Initiative Investigators. Risks and benefits of estrogen plus progestin in healthy postmenopausal women: principal results from the Women’s Health Initiative randomized controlled trial. JAMA 2002;288:321–33. [3] Grady D, Herrington D, Bittner V, et al. Cardiovascular disease outcomes during 6.8 years of hormone therapy: Heart and Estrogen/progestin Replacement Study follow-up (HERS II). JAMA 2002;288:49–57. [4] Hulley S, Furberg C, Barrett-Connor E, et al. Noncardiovascular disease outcomes during 6.8 years of hormone therapy: Heart and Estrogen/progestin Replacement Study followup (HERS II). JAMA 2002;288:58–66. [5] Utian WH. Managing menopause after HERS II and WHI: coping with the aftermath. Menopause Manage 2002;14:6–7. [6] Hulley SB, Grady D. The WHI estrogen-alone trialddo things look any better? JAMA 2004; 291:1769–71. [7] Anderson GL, Limacher M, Assaf AR, for the Women’s Health Initiative Steering Committee. Effects of conjugated equine estrogen in postmenopausal women with hysterectomy: The Women’s Health Initiative randomized controlled trial. JAMA 2004;291: 1701–12. [8] Mcintosh J, Blalock S. Effects of media coverage of women’s health initiative study on attitudes and behavior of women receiving hormone replacement therapy. Am J Health Syst Pharm 2005;62(1):69–74. [9] Shumaker SA, Legault C, Rapp SR, et al. Estrogen plus progestin and the incidence of dementia and mild cognitive impairment in postmenopausal women: the Women’s Health Initiative Memory Study: a randomized controlled trial. JAMA 2003;289(20):2651–62. [10] Shumaker SA, Legault C, Kuller L, et al. Conjugated equine estrogens and incidence of probable dementia and mild cognitive impairment in postmenopausal women: Women’s Health Initiative Memory Study. JAMA 2004;291(24):2947–58. [11] Beral V, Million Women Study Collaborators. Breast cancer and hormone-replacement therapy in the million women study. Lancet 2003;362(9382):419–27. [12] Beral V, Bull D, Reeves G, et al. Endometrial cancer and hormone-replacement therapy in the million women study. Lancet 2005;365(9470):1543–51. [13] Faber A, Bouvy ML, Loskamp L, et al. Dramatic change in prescribing of hormone replacement therapy in the Netherlands after publication of the million women study: a follow-up study. Br J Clin Pharmacol 2005;60(6):641–7. [14] Chleboweki RT, Hendrix SL, Langer RD, et al. Influence of estrogen plus progestin on breast cancer and mammography in healthy postmenopausal women. The Women’s Health Initiative Randomized Trial. JAMA 2003;289:3254–63. [15] Speroff L. Postmenopausal hormone therapy and the risk of breast cancer: a clinician’s view. Maturitas 2004;49:51–7. [16] Sendag F, Terek MC, Ozsener S. Mammographic density changes during different postmenopausal hormone replacement therapies. Fertil Steril 2001;76:445–50. [17] US Preventive Task Force. Hormone therapy for the prevention of chronic conditions in postmenopausal women: recommendations from the US Preventive Services Task Force. Ann Intern Med 2005;142:855–60. [18] American College of Obstetricians and Gynecologists. ACOG issues state-of-the-art guide to hormone therapy. September 30, 2004. Available at: www.acog.org/from_home/ publications/press_releases/nr09–30–04–2.cfm. Accessed November 22, 2005. [19] North American Menopause Society. Treatment of menopause-associated vasomotor symptoms: position statement of the North American menopause society. Menopause 2004;11(1): 11–33. [20] The North American Menopause Society. Recommendations for estrogen and progestogen use in peri- and postmenopausal women: October 2004 position statement of The North American Menopause Society. Menopause 2004;11:589–600.
HORMONE REPLACEMENT THERAPY
963
[21] Utian WH, Shoupe D, Bachmann G, et al. Relief of vasomotor symptoms and vaginal atrophy with lower doses of conjugated equine estrogens and medroxyprogesterone acetate. Fertil Steril 2001;75:1065–79. [22] Archer DF, Dorin M, Lewis V, et al. Effects of lower doses of conjugated equine estrogens and medroxyprogesterone acetate on endometrial bleeding. Fertil Steril 2001;75: 1080–7. [23] Pickar JH, Yeh IT, Wheeler JE, et al. Endometrial effects of lower doses of conjugated equine estrogens and medroxyprogesterone acetate: two-year substudy results. Fertil Steril 2003;80: 1234–40. [24] US Food and Drug Administration. ‘‘Angeliq’’. Available at http://www.fda.gov. Accessed December 14, 2005. [25] The Writing Group for the PEPI Trial. Effects of hormone replacement therapy on endometrial histology in postmenopausal women. JAMA 1996;275(5):370–5. [26] Modena MG, Sismondi P, Mueck AO, et al, The TREAT Collaborative Study Group. New evidence regarding hormone replacement therapies is urgently required: transdermal postmenopausal hormone therapy differs from oral hormone therapy in risks and benefits. Maturitas 2005;52(1):1–10. [27] Nelson HD. Commonly used types of postmenopausal estrogen for treatment of hot flashes: scientific review. JAMA 2004;291:1610–20. [28] Samsioe G. Transdermal hormone therapy: gels and patches. Climacteric 2004;7(4):347–56. [29] Rioux JE, Devlin C, Gelfand MM, et al. 17beta-estradiol vaginal tablet versus conjugated equine estrogen vaginal cream to relieve menopausal atrophic vaginitis. Menopause 2000; 7(3):156–61. [30] Suckling J, Lethaby A, Kennedy R. Local oestrogen for vaginal atrophy in postmenopausal women. Cochrane Database Syst Rev 2003;(4):CD001500. [31] Mashchak CA, Lobo RA, Dozono-Takano R, et al. Comparison of pharmacodynamic properties of various estrogen formulations. Am J Obstet Gynecol 1982;144:511–8. [32] The North American Menopause Society. The role of testosterone therapy in postmenopausal women: position statement of the North American Menopause Society. Menopause 2005;12(5):497–511. [33] Mahaday GB, Parrot J, Lee C, et al. Botanical dietary supplement use in peri- and postmenopausal women. Menopause 2003;10:65. [34] Boothby LA, Doering PL, Kipersztok S. Bioidentical hormone therapy: a review. Menopause 2004;11(3):356–67. [35] Pepping J. Black cohosh: Cimicifuga racemosa. Am J Health Syst Pharm 1999;56:1400–2. [36] Liske E, Hanggi W, Henneicke-von Zepelin HH, et al. Physiological investigation of a unique extract of black cohosh (Cimicifugae racemosae rhizome): a 6-month clinical study demonstrates no systemic estrogenic effect. J Womens Heath Gend Based Med 2002;11: 163–74. [37] American College of Obstetricians and Gynecologists Committee on Practice Bulletinsd Gynecology. ACOG Practice Bulletin Clinical Management Guidelines for ObstetricianGynecologists. Use of botanicals for management of menopausal symptoms. Obstet Gynecol 2001;97(Suppl):1–11. [38] Taylor M. Botanicals: medicines and menopause. Clin Obstet Gynecol 2001;44:853–63. [39] Tham DM, Gardnes CD, Haskell WL. Clinical review 97: Potential health benefits of dietary phytoestrogens: a review of the clinical, epidemiological, and mechanistic evidence. J Clin Endocrinol Metab 1998;83:2223–5. [40] Krebs EE, Ensrud KE, MacDonald R, et al. Phytoestrogens for treatment of menopausal symptoms: a systemic review. Obstet Gynecol 2004;104:824–36. [41] Dog TL. Menopause: a review of botanical dietary supplements. Am J Med 2005;188: 98S–108S. [42] Loprinzi CL, Stearns V, Barton D. Centrally active nonhormonal hot flash therapies. Am J Med 2005;118(12)(Suppl 2):118S–23S.