Healthcare Provider Perceptions of Cardiopulmonary Resuscitation Quality During Simulation Training.

OBJECTIVES: To assess the relationship between quantitative and perceived cardiopulmonary resuscitation performance when healthcare providers have access to and familiarity with audiovisual feedback devices. DESIGN: Prospective observational study. SETTING: In situ simulation events throughout a pediatric quaternary care center where the use of continuous audiovisual feedback devices during cardiopulmonary resuscitation is standard. SUBJECTS: Healthcare providers who serve as first responders to in-hospital cardiopulmonary arrest. INTERVENTIONS: High-fidelity simulation of resuscitation with continuous audiovisual feedback. MEASUREMENTS AND MAIN RESULTS: Objective data was collected using accelerometer-based measurements from a cardiopulmonary resuscitation defibrillator/monitor. After the simulation event but before any debriefing, participants completed self-evaluation forms to assess whether they believed the cardiopulmonary resuscitation performed met the American Heart Association guidelines for chest compression rate, chest compression depth, chest compression fraction, chest compression in target, and duration of preshock pause and postshock pause. An association coefficient (kappa) was calculated to determine degree of agreement between perceived performance and the quantitative performance data that was collected from the CPR defibrillator/monitor. Data from 27 mock codes and 236 participants was analyzed. Average cardiopulmonary resuscitation performance was chest compression rate 106 ± 10 compressions per minute; chest compression depth 2.05 ± 0.6 in; chest compression fraction 74% ± 10%; chest compression in target 22% ± 21%; preshock pause 8.6 ± 7.2 seconds; and postshock pause 6.4 ± 8.9 seconds. When all healthcare providers were analyzed, the association coefficient (κ) for chest compression rate (κ = 0.078), chest compression depth (κ = 0.092), chest compression fraction (κ = 0.004), preshock pause (κ = 0.321), and postshock pause (κ = 0.40) was low, with no variable achieving moderate agreement (κ > 0.4). CONCLUSIONS: Cardiopulmonary resuscitation performance during mock codes does not meet the American Heart Association's quality recommendations. Healthcare providers have poor insight into the quality of cardiopulmonary resuscitation during mock codes despite access to and familiarity with continuous audiovisual feedback.