A Pilot Study of CPR Quality Comparing an Augmented Reality Application vs. a Standard Audio-Visual Feedback Manikin

Background: Guidelines-based cardiopulmonary resuscitation (CPR) during in-hospital cardiac arrest is a significant predictor of survival, yet the quality of healthcare provider (HCP) CPR (e.g., nurses, physicians etc.) has been shown to be poor. Studies have found that providing HCPs with simulated CPR refresher trainings can improve their CPR quality, however, no studies have compared the use of an augmented reality (AR) CPR refresher training with a standard audio-visual (AV) feedback manikin to improve HCP training. Objectives: In our pilot study, HCPs were randomized to a refresher CPR simulation training with either our AR CPR training application (CPReality) or a standard AV feedback manikin. All subjects completed 2 min of CPR on their respective CPR training modalities, followed by an additional 2 min post-simulation CPR evaluation with no feedback. We hypothesized that the AR CPR training application would confer improved CPR quality defined as chest compression rate and depth compared with the standard AV feedback training. Results: Between January 2019 and May 2019, 100 HCPs were enrolled (50 in the CPReality cohort and 50 in the standard AV manikin cohort). The mean chest compression (CC) rate for all subjects during the intervention was 118 ± 15 cpm, and CC depth was 50 ± 8; post-intervention the CC rate was 120 ± 13 and CC depth was 51 ± 8. The mean CC rate for those trained with CPReality was 121 ± 3 compared with the standard CPR manikin training which was 114 ± 1 cpm (p < 0.006); CC depth was 48 ± 1 mm vs. 52 ± 1 (p = 0.007), respectively. Post-simulation CPR quality with no feedback showed a mean CC rate for the CPReality application at 122 ± 15 cpm compared with the standard CPR manikin at 117 ± 11 cpm (p = 0.09); depth was 49 ± 8 mm vs. 52 ± 8 (p = 0.095), respectively. In the post-survey, 79% of CPReality subjects agreed that the AR application provided a realistic patient presence compared with 59% (p = 0.07) of subjects in the standard CPR manikin cohort. Conclusions: In a randomized trial of an AR CPR training application compared with a standard CPR manikin training, the AR CPR application did not improve the quality of CPR performed during a CPR refresher training compared with the standard training in HCPs. Future studies should investigate the use of this and other digital technologies for CPR training and education.