A critical emerging issue in medicine is the definition of core competencies and standards for their achievement while the relevant descriptive epidemiology of cancer in the United States, common clinical problems seen in routine medical practice, and successful interventions in cancer medicine have been defined, comprehensive methods for assessment of the essential clinical skills needed by all practitioners based on this knowledge are limited. We have used this knowledge base, and data from a survey of primary care faculty physicians to define a working list of 15 case areas in cancer medicine, and we have begun to develop rigorous performance assessments in objective structured clinical examination (OSCE) format for each.
The specific aims of the work proposed here are: 1. To define case areas and develop OSCE stations testing core competencies in several performance dimensions 2. To determine the psychometric characteristics of each station including validity an( reliability in a major pilot study with late third year University of Wisconsin medical students [N=1O0] 3. To use Judgment Policy Capturing and an Extended Angoff procedure to set performance standards for performan( assessments using a panel of clinician experts .4. To develop a one to two hour multi-station exam to assess an describe the performance of, and to assess the incremental validity of the exam in three groups: i. fourth year Medical College of Wisconsin medical school students [N=lOO], ii. incoming University of Wisconsin and Medical College of Wisconsin primary care resident trainees [N=loo], and ili. regional non-academic primary care physicians [N=lOO]; and to assess the predictive validity of this multi-station exam for subsequent clinical performance by resident trainees 5. To disseminate the standardized multi-station exam widely, with general instructions and instructional feedback and parallel station creation protocols for each station. This set of specify aims addresses several recommendations about research in performance assessment by groups of experts (van der Vleuten and Swanson, 1990; Colliver and Williams, 1993) in the areas of content, psychometric research on reliability and validity, standard setting, and case dissemination.
