Effectiveness of a serious game addressing guideline adherence: cohort study with 1.5-year follow-up

Abstract

Background Patients presenting with acute shortness of breath and chest pain should be managed according to guideline recommendations. Serious games can be used to train clinical reasoning. However, only few studies have used outcomes beyond student satisfaction, and most of the published evidence is based on short-term follow-up. This study investigated the effectiveness of a digital simulation of an emergency ward regarding appropriate clinical decision-making. Methods In this prospective trial that ran from summer 2017 to winter 2018/19 at Göttingen Medical University Centre, a total of 178 students enrolled in either the fourth or the fifth year of undergraduate medical education took six 90-min sessions of playing a serious game (‘training phase’) in which they managed virtual patients presenting with various conditions. Learning outcome was assessed by analysing log-files of in-game activity (including choice of diagnostic methods, differential diagnosis and treatment initiation) with regard to history taking and patient management in three virtual patient cases: Non-ST segment elevation myocardial infarction (NSTEMI), pulmonary embolism (PE) and hypertensive crisis. Fourth-year students were followed up for 1.5\u2009years, and their final performance was compared to the performance of students who had never been exposed to the game but had otherwise taken the same five-year undergraduate course. Results During the training phase, overall performance scores increased from 57.6\u2009±\u20091.1% to 65.5\u2009±\u20091.2% ( p \u2009<\u20090.001; effect size 0.656). Performance remained stable over 1.5\u2009years, and the final assessment revealed a strong impact of ever-exposure to the game on management scores (72.6\u2009±\u20091.2% vs. 63.5\u2009±\u20092.1%, p \xa0<\u20090.001; effect size 0.811). Pre-exposed students were more than twice as likely to correctly diagnose NSTEMI and PE and showed significantly greater adherence to guideline recommendations (e.g., troponin measurement and D-dimer testing in suspected PE). Conclusions The considerable difference observed between previously exposed and unexposed students suggests a long-term effect of using the game although retention of specific virtual patient cases rather than general principles might partially account for this effect. Thus, the game may foster the implementation of guideline recommendations.