On the Nature of "Expertise" or "Why Do We Teach Basic Science In a Professional Curriculum Anyway"?

L. E. Olson

When revising my lectures for VM601 last year, I started to ponder the nature of clinical expertise. I reasoned that as a non-clinician, I didn't have it, and if I understood it, I might be able to better craft the "basic science" lectures that I teach in the professional curriculum. A review of the literature turned up a surprising amount of information, which I will attempt to summarize briefly in this column. The take home message is that there are several theories regarding clinical expertise as compared to biomedical expertise, which have been evaluated with empirical research. Although each theory leads to a different conclusion regarding the "best way" to teach a professional curriculum, all suggest that biomedical knowledge needs to become connected to the clinical experience, and that clinical experience, perhaps as much as 10 years, is essential for developing clinical expertise.

There are two major theories of clinical expertise, which have recently been integrated in a network model. The first theory is referred to as the "two worlds" hypothesis, as articulated by V. Patel and colleagues (for example, see 1). Clinical medicine and biomedical sciences are viewed as representing two very different and incompatible "worlds", based on differences in the types of reasoning employed and the ways in which knowledge is structured. They show that physicians resort to basic science explanations of clinical cases only when they are unsure of their diagnoses. Based on their studies, they conclude that basic science does not facilitate clinical reasoning, per se, but rather enhances the ability of clinicians to explain and communicate clinical findings. Interestingly, these authors take issue with problem-based learning, claiming that although students learn some clinical medicine, the knowledge is "contextualized", meaning that they are unable to transfer the knowledge to a different situation.

An alternate theory, known as the "encapsulated knowledge" theory, was proposed by Schmidt and Boshuizen, and has been modified to incorporate the "two worlds" hypothesis2. They theorize that expertise is the ability to "package", or "chunk", or "encapsulate" knowledge in a way that permits accurate clinical diagnoses to be made without formulating a detailed causative explanation. (for example, see 2). Physicians were found to reach more accurate diagnoses but to provide significantly less detailed explanations for the decisions using fewer biomedical concepts, as compared to medical students. However, a more recent study, comparing how cardiologists, neurologists, and medical students dealt with a cardiology case, led to the conclusion that the theory is incomplete3. Interestingly, although the expert cardiologists reached the correct diagnosis more frequently than the others, their diagnostic accuracy was measured at only 73% of the maximum score.

In summary, there is a growing literature on the nature of clinical expertise, which has implications for both how material should be taught and how professional curricula should be designed. Personally, the following quote resonated with me, because it has implications for both student learning and for faculty teaching and test construction.

The critical observation, then, is that well-organized, coherent information is easier to remember than disjointed collections of facts. Thus, the ability to explain something, even idiosyncratically to oneself, is necessary if information is to be communicated effectively and retained in memory for further analysis and, more importantly, for learning. It is argued, then, that the role of basic science, besides providing the concepts and vocabulary required to formulate clinical problems, is to create a basis for establishing and assessing coherence in the explanation of biomedical phenomena. Basic science does not provide the axioms, the analogies, or the abstractions required to support clinical problem solving. Rather, it provides the principles that make it possible to organize observations that defy ready clinical classification and analysis. We also contend that, because clinical reasoning demands the coordination of multiple tasks and goals, the ability to organize and communicate observations is an absolute prerequisite for medical expertise. (Patel and Kaufman, page 42)


For my anatomy/histology colleagues. The research seems to show that the conclusions differ between anatomy/histology and other basic sciences, such as physiology, biochemistry etc. in that radiologists tend to use the anatomy/histology basic science information more directly. (see Lesgold, A.M. et al. Expertise in a complex skill: Diagnosing x-ray pictures. pp. 311-342 In M.T. H. Chi, R. Glaser, and M. Farr (eds) The Nature of Expertise Hillsdale, Lawrence Earlbaum, 1988.)


  1. Patel, N. L. and D. R. Kaufman Clinical reasoning and biomedical knowledge: Implications for teaching. pp. 33-44 In Clinical Reasoning in the Health Professions. J. Higgs and M. Jones (Eds). Oxford, Butterworth-Heinemann, 2000.
  2. van de Wiel, M. W. J., H. P. A. Boshuizen, and H. G. Schmidt. Knowledge restructuring in expertise development: evidence from pathophysiological representations of clinical cases by students and physicians. Euro. J. Cognitive Psychol. 12:323-355, 2000.
  3. Rikers, R. J. M. P., H. G. Schmidt, and H. P. A. Boshuizen. On the constraints of encapsulated knowledge: clinical case representations by medical experts and subexperts. Cognition and Instruction. 20:27-45, 2002.