Purpose of the study: To explore whether variation in in-hospital cardiac arrest (IHCA) survival can be explained by differences in resuscitation service provision across UK acute hospitals.
Methods: We linked information on key clinical practices with patient data of adults who had a cardiac arrest on a general hospital ward or emergency admissions unit in 2016/17. We used multi-level Bayesian models to explore associations between system quality indicators (number of resuscitation officers, audits time to first shock, review unexpected non-survivors, arrest team meets at handover, hot debrief, cold debrief, real-time audio-visual feedback, frequency of mock arrest provision) and adjusted hospital survival.
Results: We received survey responses from 110 out of 180 eligible hospitals (response rate 61%) relating to 12,285 cardiac arrest cases. Variation across trusts was observed in the number of resuscitation officers (median 0.7 (interquartile range 0.5, 0.9) per 750 clinical staff employed. Key system quality indicators were undertaken infrequently: audit of time to first shock (44.7%), arrest team meeting at handover (28.9%), mock arrests ≥ monthly (22.4%), and use of CPR feedback devices (18.4%). The probability that the system quality indicators had a positive effect on hospital survival ranged from 10% to 89%. However, there was uncertainty in the estimated odds ratios and we cannot exclude the possibility of a clinical benefit. Findings were consistent across secondary outcomes.
Conclusion: In this study, we identified variation in implementation of system quality indicators. Amongst hospitals that responded to our survey, the probability that individual factors increase the odds of hospital survival ranges from 10 to 89%.
Keywords: Cardiac arrest; Cardiopulmonary resuscitation; Quality of health care.
Copyright © 2020. Published by Elsevier B.V.