Designing a touchless physical examination for a virtual Objective Structured Clinical Examination

Given the COVID-19 pandemic, many Objective Structured Clinical Examinations (OSCEs) have been adapted to virtual formats without addressing whether physical examination maneuvers can or should be assessed virtually. In response, we developed a novel touchless physical examination station for a virtual OSCE and gathered validity evidence for its use.

We used a touchless physical examination OSCE station pilot-tested in a virtual OSCE in which Internal Medicine residents had to verbalize their approach to the physical examination, interpret images and videos of findings provided upon request, and make a diagnosis. We explored differences in performance by training year using ANOVA. In addition, we analyzed data using elements of Bloom's taxonomy of learning, i.e. knowledge, understanding, and synthesis.

Sixty-seven residents (PGY1-3) participated in the OSCE. Scores on the pilot station were significantly different between training levels (F=3.936, p = 0.024, η_p² = 0.11). The pilot station-total correlation (STC) was r = 0.558, and the item-station correlations ranged from r = 0.115-0.571, with the most discriminating items being those that assessed application of knowledge (interpretation and synthesis) rather than recall.

This touchless physical examination station was feasible, had acceptable psychometric characteristics, and discriminated between residents at different levels of training.