Live or video? A prospective comparison of non-technical skills assessment in in-situ resuscitation simulation

Introduction: Non-technical skills (NTS) are central to cardiopulmonary resuscitation, yet real-time assessment during in-situ simulation is challenging. Video review offers a practical alternative, but its comparability with live observation for team-based NTS assessment remains uncertain. This study compared live and video-based Team Emergency Assessment Measure (TEAM) ratings in in-situ CPR simulations. Methods: In this prospective observational study, adult in-situ CPR simulations in a tertiary emergency department were rated live by one of three trained raters and later re-rated from video by the same rater after a ≥ 30-day washout. Agreement was assessed using intra class correlation coefficients (ICC), Pearson correlations, Bland-Altman analysis, and standard error of measurement. Secondary analyses examined paired score differences, effect sizes, and internal consistency. Results: Of 48 recorded simulations, 37 had complete paired assessments. Differences between modalities were small. No paired comparison reached the Bonferroni-corrected sensitivity threshold of p < 0.01; Leadership showed only a small nominal difference under unadjusted exploratory testing. Agreement was moderate to good (ICC 0.59-0.75), Pearson correlations ranged from 0.60 to 0.76, and Bland-Altman analysis showed small mean differences without proportional bias but relatively wide limits of agreement. Discussion: Video-based TEAM assessment produced broadly similar group-level results to live observation. Video review may support formative feedback, faculty calibration, and programme-level evaluation, but should not be treated as fully interchangeable with live ratings for individual or high-stakes judgments.