Frank J. Papa

Medical curriculum reform in North America, 1765 to the present: a cognitive science perspective.

Since 1765, five major curricular reform movements have catalyzed significant changes in North American medical education. This article describes each reform movement in terms of its underlying educational practices and principles, inherent instructional problems, and the innovations that were carried forward. When considering the motivating factors underlying these reform movements, a unifying theme gradually emerges: increasing interest in, attention to, and understanding of the knowledge-base structures and cognitive processes that characterize and distinguish medical experts and novices. Concurrent with this emerging theme is a growing realization that medical educators must call upon and utilize the literature, research methods, and theoretical perspectives of cognitive science if future curricular reform efforts are to move forward efficiently and effectively. The authors hope that the discussion and perspective offered herein will broaden, stimulate, and challenge educators as they strive to create the reform movements that will define 21st-century medical education.

Improving diagnostic capabilities of medical students via application of cognitive sciences-derived learning principles

Purpose  There is limited experimental evidence concerning how best to train students to perform differential diagnosis. We compared 2 different methods for training 2nd‐year medical students to perform differential diagnosis (DDX) of heart failure: a traditional classroom‐based lecture (control group) versus a cognitive sciences‐based approach to DDX instruction implemented through a computer‐based tutor (treatment group).

Further evidence of the relationship between case typicality and diagnostic performance: implications for medical education.

We have produced further evidence demonstrating that DDx performance is a function of a test cases typicality. Medical educators might consider exploring how cognitive scientists have used the typicality assumption to investigate and enhance the instruction and assessment of subjects engaged in other classification tasks. Further substantiation of the applicability and utility of the assumptions making up the abstraction and exemplar theories used to explain DDx performance could serve as the basis for effective and efficient curricular reforms in medical education.

Learning sciences principles that can inform the construction of new approaches to diagnostic training

Abstract The author suggests that the ill-defined nature of human diseases is a little appreciated, nonetheless important contributor to persistent and high levels of diagnostic error. Furthermore, medical education’s continued use of traditional, non-evidence based approaches to diagnostic training represents a systematic flaw likely perpetuating sub-optimal diagnostic performance in patients suffering from ill-defined diseases. This manuscript briefly describes how Learning Sciences findings elucidating how humans reason in the face of the uncertainty and complexity posed by ill-defined diseases might serve as guiding principles in the formulation of first steps towards a codified, 21st century approach to training and assessing the diagnostic capabilities of future health care providers.

Evidence of second‐order factor structure in a diagnostic problem space: Implications for medical education

Classification tasks form an integral part of diagnosing a patients medical problem. As experience is gained, knowledge of a medical problem space becomes represented by various disease classes. This knowledge and classification task ability for a medical problem space has been previously represented in an artificial intelligence setting using a knowledge‐based inference tool simulation. In this study, chest pain was defined as a specific medical problem space, and the various associated disease classes were modeled to define it. Results indicated a second‐order factor structure that related these various disease classes to chest pain. Findings suggest that chest pain is a multidimensional problem space. This has implications for how medical education and differential diagnostic problem‐solving instruction should be taught.

‘Disease Class-Specific, Typicality-Graded’ Test Case Vignettes: Towards ‘Construct-Referenced’ Assessments of Diagnostic Performance

The more precise our assessment measures and instructional feedback, the more likely both students and training institutions can take those steps necessary to developing and refining the knowledge base constructs underlying diagnostic competency. This paper provides a theoretical description of how literature elucidating the factors underlying diagnostic performance could be used to produce more valid assessments of diagnostic competency. More specifically, literature suggesting that diagnostic performance is disease class-specific, coupled with evidence of a positive relationship between a case’s typicality and diagnostic accuracy represent a potentially important breakthrough for medical educators. Use of disease class-specific, typicality-graded test items may make it possible to more directly link assessment measures with the constructs or latent traits responsible for observed performance. Such items make possible the development of construct-referenced as opposed to norm-referenced assessments of diagnostic performance. Construct-referenced assessments of diagnostic performance could prove to be more appropriate, meaningful and useful testing procedures for students, faculty and society.

The Influence of Neuro-Cognitive Modelling and Philosophy on Issues of Curriculum and Assessment in Medical Education

One major concern with medical education is the dichotomy between the goal of the educational process and the process itself. When the diagnostic process is understood as a categorization task, the goal becomes redefined as teaching diagnostic categorization skills within specific medical domains. The problem, then becomes apparent because the task being taught is a very specific, structured activity, and the methods used to teach it are often unstructured enough as to be considered amorphous. With the sole exception of the Exemplarist model taken to its extreme (no categories at all, just exemplars) the existence within mental representation of some form of categories is agreed upon in all major philosophical schools. If, then, medical decision making is a categorization task, the medical curriculum and instruction should be designed from day one around these categorization tasks. The lack of concern with philosophy by medical educators is disconcerting, given that the a philosophy of categorization has existed since Plato. This paper will discuss the problem of categorization in general and how the new fields of neuro-cognitive modelling and the Philosophy of Mind can offer important, fascinating and insightful contributions to the field of medical education and assessment.

Disease-Class Specific, Computer Adaptive Testing on the World Wide Web

Evidence of diagnostic competency is demonstrated by the physician’s ability to correctly diagnose a given disease despite the many different combinations of signs and symptoms with which the disease may manifest. Unfortunately, diagnostic performance on one case cannot be used to predict performance on another — the case-specificity phenomena. This later finding suggests that the assessment of diagnostic competency should be predicated upon a subject’s performance against a number of different case presentations for any given disease. Computer adaptive testing (CAT) appears to be a viable procedure for producing logistically feasible assessments of disease class-specific diagnostic capabilities. Disease class-class specific, CAT procedures designed for use on the World Wide Web (utilizing Netscape as a common interface across various hardware platforms) are currently undergoing field trials.