Introduction:
The act of conducting a research study could potentially improve quality of care by focusing providers on the study condition or detract from care by distracting providers with complex protocols. We hypothesized that odds of receiving bystander cardiopulmonary resuscitation (CPR) would be higher during a trial of dispatcher-assisted CPR instructions than during periods immediately before and after.
Methods:
The investigation was a cohort study of 8,626 adult subjects who suffered non-traumatic out-of-hospital cardiac arrest prior to emergency medical services (EMS) arrival in greater King County, Washington, between January 1, 1999, and December 31, 2011. Bystander CPR status was assessed through review of audio dispatch tapes and EMS reports to classify any bystander CPR (any B-CPR). Any B-CPR was further categorized as bystander CPR with dispatcher assistance (DA-CPR) or bystander CPR without dispatcher assistance (B-CPR, no DA). We used multivariable logistic regression to evaluate the odds of any B-CPR before, during, and after the Dispatcher Assisted Resuscitation Trial (DART), a randomized trial of dispatcher-assisted CPR instruction comparing chest compressions alone with compressions plus rescue breaths, which ran from June 1, 2004, to September 30, 2009. We also evaluated whether the odds varied by type-specific B-CPR.
Results:
Patient and arrest circumstances were similar across the study periods. Compared to the period before DART, odds of receiving any B-CPR were higher during DART (OR=1.35, 95% CI = 1.23-1.49), but no different after OR=1.11, 0.98-1.25). Similarly, compared to the before period, the odds of DA-CPR were higher during DART (OR=1.79, 1.59-2.02) but no different after (OR=0.94, 0.80-1.10) (Table 1).
Discussion:
Odds of bystander CPR were higher during the study compared to periods before and after. The increase seems to be largely related to higher likelihood of DA-CPR, suggesting a possible community-wide benefit from this research protocol.