Medical interns’ performance on screen-based virtual patient assessments before and after an emergency medicine internship

Emergency medicine (EM) internships are central to acquiring acute-care competencies, yet their educational impact is usually judged by self-assessment or supervisor ratings that correlate weakly with objective performance. Screen-based virtual patient (VP) simulation provides a standardized setting for assessing clinical performance in simulated scenarios. In this study we examined changes in medical interns’ performance on a screen-based VP assessment before and after an eight-week EM internship. In this single-center, single-group, prospective, pre–post cohort study, 86 sixth-year medical interns completed screen-based VP assessments (Full Code) before and after an eight-week EM internship. Each participant managed six randomly assigned cases (one per category: cardiac, pulmonary, abdominal, central nervous system, internal medicine, and other emergencies). The platform generated an overall case-management score, stage-specific scores, time to disposition, and binary indicators of correct final diagnosis and disposition. Continuous outcomes were analysed at the participant level using paired-samples t-tests or Wilcoxon signed-rank tests; binary outcomes were analysed with generalized estimating equations accounting for clustering of cases within participants. Secondary comparisons were Holm-adjusted and are reported as exploratory. The overall case-management score increased from 69.2 ± 9.4 to 83.4 ± 5.4 (Δ=+14.2, 95% CI 12.4 to 15.9; p < 0.001). Correct final diagnosis rose from 85.7% to 98.4% and correct disposition from 70.5% to 93.0% (both p < 0.001), while time to disposition decreased from 399 to 336 s ( p < 0.001). Five of six stage scores improved significantly, with the largest gains in intervention (Δ=+17.9) and stabilization (Δ=+9.6); review of systems did not change significantly (Δ=+2.7; p = 0.087). All six presentation categories improved, greatest for cardiac (Δ=+18.5) and smallest for CNS emergencies (Δ=+8.8). Medical interns demonstrated improved performance on screen-based VP assessments following an EM internship, although review of systems and intervention scores remained comparatively lower than those of other domains. Given the single-group, pre-post design and use of identical cases at both time points, these findings should be interpreted as temporal improvement in simulation-based performance rather than definitive evidence of internship effectiveness or transfer of learning to clinical practice. Nevertheless, screen-based VP simulation may provide standardized assessment data that could contribute to a broader programmatic assessment approach in undergraduate EM education.