|
|
Continuity - 2021- Union - SBH |
|
Pearls of Evidence
What is the role of the clinical “pearl”? [see also Family Medicine p 619, Vol 36, No 9-2004 by William Crow: "Using 'Pearls' to Ensure That Learners Learn Key Information] Rajesh S. Mangrulkar MD 1, Sanjay Saint MD, MPH 1, Shelley Chu MD, PhD 1 and Lawrence M. Tierney Jr. MD 1
Medicine is undergoing rapid change. Science has altered the way in which we view disease, diagnostic procedures, and therapeutic options. Transmitting this information to practicing physicians, students, and residents remains one of the challenges facing medical educators. Typically, there are two methods of communicating these options, and these often appear to be in conflict. The clinical pearl is an often-anecdotal method of relaying information. Pearls offer clues about the patient at hand and are aphorisms worded to imply something absolute. The evidence-based medicine (EBM) approach has emerged as an alternative method to transmit information that is not yet common knowledge. The use of the EBM model depends on valid scientific studies. Yet applying such evidence to clinical practice still hinges on the sound judgment of clinicians. Typically, physicians and medical students rely primarily on only one heuristic model for their own education and decision making. Some critics consider the use of pearls and anecdotes to be a teaching method somewhat lacking in rigor and thus unscientific; others, however, consider EBM unnecessarily rigid and often impractical (1,2). It would seem that the practice of medicine is best suited by an amalgam of these two educational strategies, and indeed such a method may make medicine more enjoyable.
Most students and residents easily recall the precise clinical scenario when they first heard a particular pearl. The setting may be attending rounds, in a classroom, or at a patient’s bedside with a senior physician. A pearl involves information obtained from the patient’s history or examination; occasionally laboratory tests or other data form its basis. It provides insight into the care of a specific patient. The purpose of such a pearl may be to call attention to a serious but treatable disorder that should be considered, to modify the probability of a disease already under consideration, or to highlight treatment options. The following teaching scenario demonstrates some key features of the clinical pearl.
Several features of this statement are worth noting. First, the teaching point or pearl, although specific to the patient, can be generalized to other patients with a similar disorder. A clinical event has markedly altered the overall picture, and the primary care team is likely deeply concerned about his outcome. The learners may more readily retain information transmitted by the attending physician during this time of emotional and intellectual activation. Second, a respected and experienced teacher has provided the information, thereby lending validity to the information. Clinical experience, communication ability, and reputation comprise the components of this validity. Third, the pearl itself conveys information that is not common knowledge. If the content is widely known, uninteresting, or irrelevant to a specific patient, it is not a pearl but rather a “glass bead.” The preceding example can be taken to mean that serious metabolic disorders such as hypoglycemia can cause focal neurologic deficits, mimicking stroke or transient ischemic attack. The teacher has certainly seen examples of this problem and perhaps even missed this diagnosis at an earlier point in his or her career. He or she is motivated to make sure that future learners are aware of this entity. Finally, and most importantly, the delivery style is pithy, even catchy. The “50 of D50” is an easily recalled phrase, making it easy to pass on to the next generation of learners. Delivery style adds to the likelihood of recall. Indeed, most pearls imply certainty of content. The sentence frequently contains the words “never,” “always,” “everything,” or “nothing.” The absoluteness of the information is reassuring to a novice learner, providing order and structure in a science replete with uncertainty and overlap. Many of the characteristics of pearls are process oriented rather than content focused. There is a clear emphasis on ensuring that the information has an optimal chance of being retained by the learner, much as a mnemonic abets memorization of long lists of facts. However, a pearl differs significantly from a mnemonic in two important ways. First, the structure of a pearl is contained within a phrase or sentence, avoiding the need to recall an often-unrelated word such as an abbreviation, as commonly occurs with mnemonics; the phrases in pearls are more easily retained. A more crucial difference, however, is that the information in a pearl is always clinically important. Many abbreviations used as mnemonics frequently involve basic science knowledge (e.g., “SITS” to help remember the rotator cuff muscles) (3) or outdated information (e.g., in “MUDPLIERS” to memorize the differential diagnosis of an anion-gap acidosis, the “P” for paraldehyde being an irrelevancy today) (4). Six examples of clinical pearls drawn from the experience of one of the authors (LMT) are presented in Table 1. Two characteristics of this table are worth noting: the exact wording and the interpretive meaning are both provided, revealing the underlying richness of the information, as are the implications for diagnosis, treatment, or both. In the teaching vignette, the meaning of the attending physician’s pearl was similarly twofold: metabolic derangements can cause focal neurologic deficits (diagnosis), and administering a therapy, such as 50% dextrose solution (D50), may provide rapid therapeutic relief for this disorder (therapy). This categorization foreshadows a potential alliance between those who rely on the clinical pearl for teaching and those who predominantly use EBM instruction.
It is not coincident that pearls can have diagnostic and therapeutic implications. These appear to be two of the most common types of questions that physicians ask (5). It is in the search and use of external sources of information to answer questions that influence medical decision making that the proponents of EBM have sought to make explicit what is typically implicit. This was once thought to result in a dramatically different approach to the teaching of medical decision making, although many practitioners are familiar with the techniques of searching MEDLINE and critical appraisal. Comments such as “EBM is boring” or “EBM is impractical to use in my busy clinical practice” are often heard at workshops designed to teach these skills. An analysis of the content of EBM instruction may help identify the barriers to its current use. EBM has been described as the “conscientious, explicit and judicious use of current best evidence in making decisions about the care of individual patients” (6). Its historical roots originate in the movement of “Medecine d’Observation”, led by the father of clinical epidemiology, P.C.A. Louis, in the 1830s. He is credited with first documenting the use of systematic observation and record keeping to determine the relative efficacy of two forms of therapy (early vs. late bloodletting with leeches) for treatment of a disorder (pneumonia) (7). Through statistical analysis, he called the magnitude of the benefit of leech therapy into question, arguing that early treatment resulted in increased mortality. In addition, he introduced the concepts of randomization, confounding factors, and control groups. The adverse reaction by some clinicians to the current teachings of EBM is at least analogous to the backlash that met Louis and his followers in the 19th century (8). Nonetheless, EBM has been adopted in training programs internationally. As one of the original developers of this paradigm, Sackett and his colleagues have emphasized certain skills that clinicians can employ to answer clinical dilemmas: formulating sound clinical questions, searching and retrieving the best evidence to answer the question, appraising the evidence for validity, and deciding how the results may impact the care of patients (9). Categorizing and formulating the question Questions asked by clinicians can be classified into several types ( Table 2). Each reflects the type of information sought and the setting in which the question was asked. This step accomplishes two objectives: it identifies information a priori in which the learner is interested and it helps map the path to retrieving the information by identifying key search terms.
Identifying the best source and searching it A clearly constructed question also helps point the learner to an information resource likely to provide an answer. For example, an inquiry into the effectiveness of selective serotonergic reuptake inhibitors in the treatment of vasodepressor syncope (a therapy question) may be best answered by the Cochrane database (10), a source that contains systematic reviews of therapeutic interventions. Similarly, a question on screening for prostate cancer can be addressed through a clinical practice guideline, available on the National Guideline Clearinghouse Internet site (11). Appraising the evidence for the bottom line Once the learner has searched a resource and selected a reference, he or she then appraises the content (including methodology, results, and conclusions), determining both the internal validity of the study and generalizability of the study’s conclusions. The majority of published EBM curricula have focused on this step (12). Typically, learners participate in a journal club, discussing articles using the Users Guide to the Medical Literature series for guidance (13). The impact of these interventions as teaching strategies, however, appears to be minimal (12). Deciding how the results may impact the patient Here, EBM returns to the bedside, and it is perhaps where the greatest deficiencies in EBM lie. Many experts have argued that EBM considers groups of patients (those enrolled in studies) but ignores the individual from whom the practitioner generated the question (14). Application of the results of an article goes far beyond determining whether the patient would have been included in the study, calculation of pre- and post-test probabilities of diagnosis or treatment, or establishing test-treat thresholds. Clinicians consider many other issues in the decision-making process, including patients’ values, compliance, social support, and models of illness. EBM is usually silent on these topics.
A summary of the strengths and weaknesses of both teaching strategies is presented in Table 3. A pattern emerges on inspection of this table: many of the deficiencies of clinical pearls are exactly contained within the strengths of EBM and vice versa. The advantages for a teacher using clinical pearls include specificity to the individual patient, and the emphasis on easy retention and transmission of information. Evidence-based instruction is often deficient in these matters. By contrast, EBM ideally provides the learner a systematic, unbiased approach to clinical questions, emphasizing scientific rigor with the use of numbers, such as absolute risk reductions and likelihood ratios. Both of these characteristics are lacking when pearls are used to teach.
Rather than thinking of these two teaching tools as mutually exclusive and from competing camps, the complementary nature of their strengths and weaknesses offers a solution where both can be used during medical instruction. What follows is a description of a teaching exercise capitalizing on the important features of both. Table 4 summarizes the results of performing the steps in this exercise for six pearls previously presented.
Returning to the vignette described earlier, the pearl, “A stroke is never a stroke unless it’s had 50 of D50” had just been delivered to the residents.
This dialogue illustrates the process of deriving meaning from the pearl. Some clarification is usually necessary, because the statements are often poetic and potentially subject to various interpretations. The resident now understands that the attending is teaching that metabolic disorders may cause focal neurologic deficits and that correction of the disorder may reverse the deficit. By the end of the dialogue, the learner has generated two searchable clinical questions. Using the EBM model ( Table 2), the first can be characterized as either a differential diagnosis question (“How likely is hypoglycemia a cause for the symptoms compared to a thromboembolic cause?”) or a risk factor question (“What is the likelihood that hypoglycemia is the cause of the neurological deficit?”). The resident’s second concern is clearly therapeutic (“What will be the efficacy of hypertonic glucose?”) The ideal resource to answer the two questions would be a synthesis of published literature, either of cohort studies (to answer the risk factor component) or of controlled clinical trials (to answer the therapy question). The Cochrane Collaboration database may provide an excellent place to start. Other potential sources include ACP Journal Club (15), a MEDLINE controlled trial or cohort study, a MEDLINE review article, a neurology textbook, or individual case reports. Embarking on a search, the resident returns to the attending physician later that afternoon with her results:
There are several advantages of this exercise. First, it brings evidence to the bedside. Because the pearl is always delivered within a clinical context, the subsequent search for evidence will never be out of this context. Second, the resident has engaged in higher-order thinking according to Bloom’s taxonomy of hierarchical cognitive processes (16) because she has been forced to synthesize published opinions and anecdotes to arrive at a value judgment on a teaching point. Third, the teacher and student are partners in this learning endeavor. Both sides must be willing to accept the results of the search, balancing it with the clinical experience of the participants. Finally, the resident may be more likely to appreciate both teaching techniques, incorporating them into her own “toolbox” when the time comes that she is functioning as a teacher. To test the practicality of this educational exercise, we systematically searched for the evidence behind the other five pearls presented ( Table 4). Some important observations were forthcoming. First, generation of the questions after clarification of the meaning is straightforward. Specifying important outcomes at the outset was crucial to reducing the scope of the search. In addition, the ideal sources were also easy to identify. However, we experienced on average only a 50% successful retrieval rate of information (an asterisk next to resource denotes a successful search). Next, most of the questions were not answered by controlled trials but by case reports and uncontrolled case series. Compared with the number of questions clinicians pose daily, the number of internally valid, controlled studies to answer these questions is small (5). Finally, most of the pearls remained reasonably valid when subjected to the rigors of evidence-based searches. Most often, modifications provided exceptions to the certainty with which the pearl was presented. For example, the incidence of multiple sclerosis over the age of 50 is not 0% (“diagnose something else”; see pearl 2, Table 4), but multiple sclerosis is unusual in patients over 50 years of age. Also, the incidence of splenomegaly as a finding in multiple myeloma is probably not 0% (“NO splenomegaly”; see pearl 6, Table 4), but is nonetheless very low (about 5%). The absoluteness of the pearl’s wording clearly helps its retention by the learner, but at the cost of some accuracy. The evidence-based search both confirms the pearl’s validity and gives it precision.
Clinical pearls and EBM are powerful teaching tools that can be mutually enhancing. To use both, however, it is necessary to abandon the rhetorical debate between empiricism and science. Instead, we need to adjust our expectations and interpretations of both. Assumption 1: “Anecdotal experience is not rigorous.” Although less scientific, a systematic review of a collection of anecdotes can often provide valid conclusions about disease processes and therapy. In addition, anecdotes provide insights into other relevant aspects of patient care, such as the social context, the expression of patient values, or the subtle variations in the ways that patients present. Assumption 2: “Pearls fail to convey the uncertainty inherent in medical decisions.” This may be a valid criticism of not only pearls, but of EBM as well. Although EBM attempts to deal with statistical uncertainty (by using P values and confidence intervals), to date neither technique provides important methods to help physicians sharpen medical judgment in the face of clinical uncertainty. Assumption 1: “EBM mandates the use of randomized-controlled trials above all others.” Because the “ideal” study is not available to support many clinical decisions (17), expectations for the quality of evidence that we seek should now be modified to accept the “best available” evidence. Assumption 2: “The use of EBM ignores the caregiver’s clinical judgment and the patient’s values.” In fact, implicit in the EBM skill set is the presence of both constructs. The first step (formulating questions) mandates that the physician discover what outcomes are important to him or her and to the patient and is strongly influenced by prior experiences of both parties. The last step (applying to the patient) forces the clinician to decide if the results apply to the specific patient and to consider the patient’s values and characteristics that might influence the success of the management of his disease. Teachers of EBM should further explore techniques to help practitioners develop this set of skills. Both the use of clinical pearls and EBM as teaching techniques improve the care of patients. Pearls are exciting and efficient vehicles for the delivery of information while simultaneously serving as fodder for the generation of important searchable (and researchable) clinical questions. The principles of EBM can then be used to lend validity to the pearl, rendering it worthy of further transmission or identifying it as merely a “glass bead.” As clinical educators, we should fully embrace both the clinical pearl and EBM because they serve valuable and complementary roles in medical education. References 1 Paauw D.S. (1999) Did we learn evidence-based medicine in medical school? Some common medical mythology. J Am Board Fam Prac 12:143-149 2 Maynard A. (1997) Evidence-based medicine: an incomplete method for informing treatment choices. Lancet 349:126-128 MEDLINE 3 (2000) Shoulder pain. In: Frances C. , Bent S. and Saint S. (Eds.) (2000) Saint-Frances Guide to Inpatient Medicine(2000) (pp. 356-358) Baltimore, MD: Lippincott, Williams and Wilkins 4 (1997) Approach to acid base disorders. In: Saint S. and Frances C. (Eds.) (1997) Saint-Frances Guide to Inpatient Medicine(1997) (pp. 236-238) Baltimore, MD: Lippincott, Williams and Wilkins 5 Ely J.W. , Osheroff J.A. and Ebell M.H. (1999) Analysis of questions asked by family doctors regarding patient care. BMJ 319:358-361 MEDLINE 6 Sackett D.L. , Rosenberg W.M.C. and Gray J.A.M. et al. (1996) Evidence-based medicine: what it is and what it isn’t. BMJ 312:71-72 MEDLINE 7 Morabia A. (1996) PCA Louis and the birth of clinical epidemiology. J Clin Epidemiol 49:1327-1333 MEDLINE 8 Vandenbroucke J.P. (1996) Evidence-based medicine and “medecine d’observation”. J Clin Epidemiol 49:1335-1338 MEDLINE 9 (2000) Introduction to EBM. In: Sackett D.L. , Straus S.E. and Richardson W.S. et al. (Eds.) (2000) Evidence-based Medicine: How to Practice and Teach EBM(2000) (2nd ed) (pp. 3-4) New York, NY: Churchill Livingston 10 The Cochrane Collaboration http://www.cochrane.org (http://www.cochrane.org) Accessed October 4, 2002 11 National Guideline Clearinghouse http://www.guidelines.gov (http://www.guidelines.gov) Accessed January 15, 2002 12 Green M.L. (1999) Graduate medical education training in clinical epidemiology, critical appraisal, and evidence-based medicine: a critical review of curricula. Acad Med 74:686-694 MEDLINE 13 Oxman A.D. , Sackett D.L. and Guyatt G.H. (1993) Users guide to the medical literature I: how to get started. JAMA 270:2093-2095 MEDLINE 14 Feinstein A.R. and Horwitz R.I. (1997) Problems in the “evidence” of “evidence-based medicine”. Am J Med 103:529-535 MEDLINE 15 ACP Journal Club. Available for subscription at http://www.acponline.org. Accessed October 4, 2002 16 Krathwohl D. and Bloom B. (1994) Taxonomy of Educational Objectives: Cognitive Domain. New York, NY: Addison-Wesley 17 Howes N. , Chagla L. , Thorpe M. and McCulloch P. (1997) Surgical practice is evidence-based. British J Surg 84:1220-1223 Footnotes 1 Dr. Mangrulkar is from the Department of Internal Medicine at the University of Michigan Medical School. Dr. Saint is from the Ann Arbor VA Medical Center. Dr. Chu is from the University of California at San Francisco School of Medicine. Dr. Tierney is from the San Francisco VA Medical Center.
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
