COMPUTER-BASED EXERCISES IN PATHOPHYSIOLOGIC DIAGNOSIS is a research project that aims to develop computer-assisted instructional tools for clinical science education in medicine. A computer system will be developed that consists of two related programs: DSS, a user-friendly decision support system, based on the concepts of clinical decision analysis and medical cost-effectiveness analysis; and PLAN-ALYZER, an interactive critiquing system based on the concepts of automated transition networks. Methods to be employed in the development of the system will arise from decision theory, artificial intelligence, and software engineering. The primary computer system that will be used for the project is the Apple Macintosh. The computer system will be applied to clinical problems in two content areas of second year medical education: the workup of anemias in hematology and the evaluation and management of suspected coronary artery disease in cardiology. Databases will be developed for each of these areas, drawing upon the resources of clinical medicine, community and family medicine, and laboratory medicine. The effect of the computer tool in these two areas will be tested through a randomized, controlled crossover trial in three medical school classes. Students will be stratified by sex and MCAT Verbal and Quantitative scores into two randomized groups. The first group will undergo the hematology intervention, and serve as a control group for the cardiology exercise. The other group will serve in reverse. Experimental subjects will undergo a laboratory exercise using the DSS and PLAN-ALYZER programs; control subjects will receive refined textual material that is equivalent in content. Performance will be measured by patient evaluation problems built into the examination mechanism of the medical school in the randomized design. Additionally, student performance will be measured over time, both within the second year intervention and through two repeat interventions in the clinical years of medical school. The project aims to investigate whether computer science research in basic medical education will assist the integration of discrete medical facts into effective clinical decision making, both in the classroom environment and through clinical training.
