Simulation intervention with manikin-based objective metrics improves CPR instructor chest compression performance skills without improvement in chest compression assessment skills.

INTRODUCTION: Cardiopulmonary resuscitation (CPR) instructor/coordinator (CPR-I/C) adherence to published guidelines during resuscitation and learner assessment for basic life support (BLS)/CPR skills has not been experimentally studied. Investigators sought to (1) determine the quality of CPR-I/C chest compression and the accuracy of CPR-I/C chest compression assessment, and (2) improve CPR-I/C compression and assessment skills through cardiac arrest simulations with objective in-scenario performance feedback.

METHODS: Thirty CPR-I/Cs (median, 20 years [range, 4-40 years] of BLS provider experience; 6 years [range 1-40 years] of BLS instructor experience) were randomized to control or experimental group. Each subject performed compressions during a 2-minute simulation, then reviewed 6 videos of simulated CPR performances (featuring prespecified chest compression parameters) for scoring as "pass" or "needs remediation." Subjects participated in a second simulation with or without real-time manikin compression feedback, then reviewed 6 additional videos. Primary outcome variables were the proportion of subjects with more than 80% (American Heart Association regional criteria) or more than 23 of 30 (ie, 77%; American Heart Association instructor manual criteria) correct compressions and subjects' accuracy of "pass"/"needs remediation" assessment for videos. The secondary outcome variable was correlation between subjects' correctness of chest compressions and their assessment accuracy for simulated CPR compression performance.

RESULTS: All CPR-I/C subjects compressed suboptimally at baseline; real-time manikin feedback improved the proportion of subjects with more than 77% correct compressions to 0.53 (P < 0.01). Video review data revealed persistently low CPR-I/C assessment accuracy. Correlation between subjects' correctness of compressions and their assessment accuracy remained poor regardless of interventions.

CONCLUSIONS: Real-time compression feedback during simulation improved CPR-I/C's chest compression performance skills without comparable improvement in chest compression assessment skills.