By Joshua Moss, MD

Associate Professor of Clinical Medicine, Cardiac Electrophysiology, Division of Cardiology, University of California, San Francisco

Dr. Moss reports he is a consultant for Abbott, Boston Scientific, and Medtronic.

SYNOPSIS: In a recent analysis, using any CPR was associated with significant improvement in 30-day survival, with slightly better outcomes associated with standard CPR over compression-only CPR.

SOURCE: Riva G, Ringh M, Jonsson M, et al. Survival in out-of-hospital cardiac arrest after standard cardiopulmonary resuscitation or chest compressions only before arrival of emergency medical services. Circulation 2019; Apr 1. doi: 10.1161/CIRCULATIONAHA.118.038179. [Epub ahead of print].

In 2000, the American Heart Association first recommended compression-only CPR (CO-CPR) when bystander rescuers were unwilling or unable to perform mouth-to-mouth rescue breathing. By 2010, efficacy of CO-CPR was supported by two randomized studies. The guideline update that year reoriented the universal sequence of assessment and care from A-B-C (Airway, Breathing, Compressions) to C-A-B. Untrained rescuers were advised to provide CO-CPR or follow EMS dispatcher instructions. In the 2015 guidelines, dispatchers were advised to provide CO-CPR instructions to callers for adults with suspected cardiac arrest. The Swedish national CPR guidelines were modified similarly in 2006, 2011, and 2016. Riva et al analyzed data from three guideline “eras,” 2000-2005, 2006-2010, and 2011-2017, to evaluate frequency and type of CPR delivered and association with 30-day survival. The authors used data from the Swedish Register for Cardiopulmonary Resuscitation, a national quality registry of out-of-hospital cardiac arrest (OHCA) events to which all EMS organizations in Sweden report. A total of 30,445 bystanders witnessed OHCA events between 2000 and 2017, with registry information on type of CPR performed and 30-day survival included in the analysis. Cases were excluded if they were unwitnessed, EMS-witnessed, or treated with rescue breath-only CPR. Otherwise, type of CPR was classified as standard (S-CPR, including rescue breathing), CO-CPR, or NO-CPR (none delivered before EMS arrival).

Patients were more likely to receive NO-CPR or CO-CPR when EMS response time was shorter (median response time seven minutes in the NO-CPR group, eight minutes in the CO-CPR group, and 10 minutes in the S-CPR group). Responders delivering S-CPR were more likely to be medically educated (25.1% of the S-CPR group vs. 15.3% of the CO-CPR group). Over the three eras analyzed, rates of CPR received before EMS arrival increased from about 41% (2000-2005) to 59% (2006-2010) to 68% (2011-2017). The proportion of patients receiving CO-CPR also increased, from 13% of CPR recipients (2000-2005) to 24% (2006-2010) to 44% (2011-2017).

Interestingly, survival at 30 days, always poor for OHCA, nearly doubled over the study period, regardless of whether CPR was delivered or what type of CPR was used. In the most recent era (2011-2017), 30-day survival was best for patients who received standard CPR (16.2% vs. 14.3% for CO-CPR and 7.1% for NO-CPR). When data were adjusted for patient age, sex, cause of arrest, location, EMS response time, and year, the odds of survival were 2.6 times higher with S-CPR and two times higher with CO-CPR compared with NO-CPR. The authors concluded that future CPR guidelines should continue to endorse CO-CPR as an option.


The current data, while limited by their observational nature, provide some important insights into CPR patterns and efficacy. Bystander CPR delivery increased dramatically between 2000 and 2017. Well over half of CPR delivered was CO-CPR between 2014 and 2017 (compared to less than 20% of the CPR delivered in 2000). Those observations suggest one of two things: either the emphasis on CO-CPR has contributed to an increase in willingness and/or ability to perform CPR, or that willingness and ability increased independently, and people are simply following guideline recommendations. We can only speculate about which explanation accounts for these findings, but data from other studies certainly suggest that bystanders are more likely to participate when CO-CPR is accepted and expected.

For witnesses of OHCA without medical training, the publication of data such as these hopefully will encourage even more to participate in providing CPR until EMS arrives (in addition to ensuring the rapid arrival of EMS). However, for cardiologists and other medically trained responders, a few take-home messages should not be lost. S-CPR, including rescue breathing, was associated with improved 30-day survival compared with CO-CPR, with an adjusted odds ratio of 1.2 (95% confidence interval, 1.1-1.4). S-CPR appeared slightly superior to CO-CPR (and both were significantly superior to NO-CPR) across all subgroups analyzed, except when EMS response time was longer than 10 minutes. With an EMS response time between 10 and 14 minutes, only S-CPR was associated with improved survival over NO-CPR. No form of CPR was associated with improved survival when EMS response time was > 14 minutes. While not proof of direct causation, the data suggest that CO-CPR without rescue breathing could become inadequate before S-CPR. The prognosis for prolonged arrest without EMS support is unsurprisingly grim.

Importantly, there are insufficient data to support CO-CPR for children and infants (for whom asphyxia may be the cause of their cardiac arrest), and rescue breathing still is recommended. From these data, it is unknown whether patients who are alive at 30 days after S-CPR and CO-CPR experience equivalent neurologic outcomes.

Further, it is unknown whether CO-CPR is similarly effective for victims of unwitnessed cardiac arrest. Overall, it is unknown whether this robust Swedish data set can be completely generalized to all populations, considering the potential for variations in causes of OHCA, response time and efficacy of EMS systems, and education and awareness of bystanders. With these points in mind, it is notable that the 2015 Guidelines for Basic Life Support, the most recent update, still recommend chest compressions and rescue breaths in a ratio of 30:2 when administered by a healthcare provider or a trained lay rescuer (if able). Emphasis should be placed on timely and effective chest compressions, but healthcare providers should not forget to breathe.