Democratizing Virtual Patient Case Creation: A Proof-of-concept Technical Framework for Clinicians

ObjectiveVirtual patient cases are a scalable and engaging tool for training medical professionals. Strategies and frameworks for their implementation in teaching and training settings are few, technically complicated and/or expensive. We developed and evaluated open source and free virtual patient cases to test knowledge acquisition during an echocardiography training program for internal medicine trainees in Haiti. The objective of this paper is to describe the technical aspects of the GMENEcho virtual patient cases implementation and motivate similar work by resource-constrained teams. MethodsWe used an open source engine for text-based games (Twine) since it provides the necessary interaction mechanics and is usable out-of-the-box. The case code was written in SugarCube 2.30.0 notation and the tweego-generated .html file was hosted on Github Pages for continuous integration and deployment, making iterations by the clinical team seamless. Data from completed tests were reported back via email through a third party integration. ResultsThe technical work was completed in two weeks by a team member with a clinical background and minimal computer programming experience. The virtual patient cases were deployed for a pretest (November 2023) and a second time unaltered for a posttest (June 2024) after the interim hands-on and theoretical training had been completed. Qualitative feedback was positive or neutral. The overall score in the posttest was significantly higher with a large effect size (mean absolute improvement 15.26%, p < 0.001; Cohens d: 1.398), similarly to the diagnostic score (mean absolute difference 16.09%, p < 0.001; Cohens d: 1.402). Management performance missed statistical significance by a small margin. The System Usability Scale (SUS) score was 74.6 ("Excellent").There was reduced inter-trainee variability across metrics in the posttest, including the SUS score. DiscussionThis proof-of-concept methodology can be applied to create clinical patient cases for use within a class or a clinical training setting, through a friendly graphical user interface. A more complex software stack can allow for remote or larger scale implementations with additional features. ConclusionThe rapid development time and positive qualitative and quantitative feedback highlight the potential of this approach for clinical education in resource-constrained settings. It can serve as a template for more streamlined adaptations of case-based learning in diverse healthcare settings.