Cardiopulmonary Resuscitation Feedback: A Comparison of Device-Measured and Self-Assessed Chest Compression Quality

Abstract

BackgroundHigh-quality cardiopulmonary resuscitation is the foundation of cardiac arrest care. Guidelines specify chest compression depth, recoil, and rate, but providers often fail to achieve these targets. Furthermore, providers are largely unable assess the quality of their own or other peoples’ chest compressions. Chest compression feedback devices can improve chest compression quality; their use is endorsed internationally, but they remain largely absent in clinical care.This article analyzes preclinical data collected during a quality improvement project. It describes provider demographics and perceptions about their chest compression quality and correlates them to measured chest compression quality, compares clinician perception of chest compressions to objective measures, and describes the effect of feedback on compression quality. MethodsClinicians were recruited from 2 metropolitan emergency departments. A questionnaire was used to assess participants’ levels of training and experience. A before-and-after assessment of chest compression quality was performed using a Laerdal CPRmeter 2 and a CPR mannequin. Pretest measures of chest compression quality were made by covering the device screen thereby blinding providers to feedback; repeat measures were then collected from the same participants but unblinded to feedback. Provider charecteristic were collected by survey. Correlations between blinded chest compression quality and provider charecteristics; the reliability of providers estimated compared to measured quality; and the effects of feedback on chest compression quality were assessed using Pearsons correlations, Cohens κ, and paired t testing. Results84 participants were assessed. The mean years of certification were 11.74. Ninty-five percent of the providers self-assessed as more experienced than novice and 81% reported performing cardiopulmonary resuscitation at least occasionally. The frequency of performing chest compressions was correlated with self-assessed skill (r = 0.58, P < .001). However, self-assessed skill was only weakly correlated with chest compression quality (r = 0.29, P = .01) and not at all with the frequency of performing chest compressions or years of certification. There was no agreement between self-assessed and device-measured chest compression depth (κ = −0.10, P = 0.11), recoil (κ = −0.14, P = .03), or rate (κ = 0.06, P =.30). The overall quality of compressions improved by 16.9%; the percentage of chest compressions achieving target depth by 3.58%; recoil by 22.82%; and rate by 23.66% with feedback. A total of 97.6% of the staff rated chest compression feedback helpful. ConclusionsOur findings suggest that participants’ demographics were not correlated with chest compression quality and that providers cannot reliably assess chest compression quality. The data also demonstrate that with minimal training, feedback can significantly improve chest compression quality.