BACKGROUND: An important aspect of virtual patients (VPs), which are interactive computer-based patient scenarios, is authenticity. This includes design aspects, but also how a VP collection represents a patient population and how a patient is presented in a VP scenario. Therefore, our aim was to analyze VP scenarios integrated into the combined internal medicine and surgery curriculum at the University of Munich (LMU) and compare the results with data from the population in Germany.

METHOD: We developed a coding framework with four main categories: patient data, patient representation, diagnoses, and setting. Based on the framework we analyzed 66 VP and compared the results with data from the German healthcare system.

RESULTS: Especially in the categories of patient data and patient representation, the VPs presented an unrealistic image of the real world; topics such as unemployment, disability, or migration background were almost non-existent. The diagnoses of the VPs and the onset of diseases were comparable with the healthcare data.

CONCLUSIONS: An explanation for the lack of representativeness of the patient data and representation might be a trend to create VPs based on fictional patient stories with VP authors trying to minimize complexity and cognitive load for the students. We suggest raising awareness among VP authors concerning personalized representations of patients without overwhelming their students. Our framework can support educators to assess the authenticity and diversity of a VP collection.