Background

Virtual patients (VPs) can be used to assess clinical reasoning. Grading metrics applied to VP exams may include scoring for appropriate differential diagnoses, proposed therapy, and the learner’s approach to the case. The learner’s inquiry of medical history, physical exam, and lab/ancillary tests utilized during the exam can all be graded.  However, the best grading metric used to assess clinical reasoning for VP examination is unresolved. 

Summary of Work

Results from two groups of students assessed by VP-based examination (n>300) were used as a basis to evaluate different grading metrics. These grading models were also compared with results of other traditional student examination performance. 

Summary of Results

Each method of grading had both pronounced advantages and disadvantages with none considered ideal. However one grading metric was perceived to perform slightly better. None of the scoring methods had a direct correlation with four traditional exam formats to which they were compared. 

Conclusions

Each grading metric used in this study had advantages and disadvantages. Medical school exams employing VP-based exams need to define what should be assessed for reliable utilization.

Take-home Messages

Objectives of VP-based examination are essential to measure learner competency in an appropriate context. Traditional exams do not necessarily measure the same aptitude that VP-based exams measure.