Effectiveness of digital resuscitation training in improving knowledge and skills: A systematic review and meta-analysis of randomised controlled trials

This review aims to evaluate the effectiveness of digital resuscitation training in improving knowledge and skill compared with standard resuscitation training. We searched through the CINAHL, Cochrane Library, EMBASE, ERIC, ProQuest Dissertations and Thesis, PsycINFO, PubMed and Scopus from inception of our review until 5 March 2018. The quality of individual and overall evidence was evaluated according to the risk of bias, Medical Education Research Study Quality Instrument (MERSQI) and Grade of Recommendation, Assessment, Development and Evaluation (GRADE) system, respectively. Meta-analyses were performed with the Review Manger software. Z-statistics were used to evaluate the overall effect of training, and I2 test was used to assess heterogeneity. Sensitivity and subgroup analyses were used for additional meta-analyses. Amongst the 15,528 studies retrieved, 20 randomised controlled trials (RCTs) were selected from 13 countries across different ethnicities. More than half (52%) of the trials had a low risk of bias, and MERSQI scores ranged from 13.5 to 15.5. The overall quality of evidence was very low according to GRADE criteria. Meta-analyses revealed that trainees in digital resuscitation training had better knowledge scores but poorer chest compression rates than that of trainees in standard resuscitation training. Digital resuscitation trainings were non-inferior to standard resuscitation trainings in skill performance scores. Subgroup analyses suggested that digital resuscitation training might consider using blended learning approach with virtual patient, computer-screen based, learning theories and video-recorded assessment, especially for basic life support trainings amongst health professionals. Despite the wide variation in digital resuscitation trainings, evidence suggesting the use of digital resuscitation training for improving knowledge and skills is inadequate. Well-designed non-inferiority RCTs in multiple settings with follow-up data and large sample size are needed to ensure the robustness of the evidence.