Early hyperoxemia is not associated with cardiac arrest outcome

Abstract Aim Studies suggest that hyperoxemiaincreases short-term mortality after cardiopulmonary resuscitation (CPR), but the effect of hyperoxemiaon long-term outcomes is unclear. We determined the prevalence of early hyperoxemiaafter CPR and its association with long-term neurological outcome and mortality. Methods We analysed data from adult cardiac arrest patients treated after CPR in tertiary ICUs during 2005–2013. We retrieved data from the resuscitation and the first arterial blood samplecollected after returnof spontaneous circulation(ROSC) (severe hyperoxemiadefined as PaO 2  > 40 kPa and moderate as PaO 2 16–40 kPa). We inspected two outcomes, neurological performance at one year after resuscitation according to the Cerebral Performance Category and one-year mortality. We used logistic regression to test associations between hyperoxemiaand the outcome and interaction analyses to test the effect of hyperoxemiaexposure on the outcomes in smaller subgroups. Results Of 1110 patients 11% had severe hyperoxemia, prevalence was 10% for out-of-hospital arrests, 13% for in-hospital arrests and 9% for in-ICU arrests. In total 585(53%) patients had an unfavourable neurological outcome. Compared to normoxemia, severe (Odds ratio [OR] 0.81, 95% confidence interval [CI] 0.50–1.30) and moderate hyperoxemia(OR 0.94 95%CI 0.69–1.27) did not associate with neurological outcome. Additionally, hyperoxemiahad no association with mortality. In subgroup analyses there were no significant associations between severe hyperoxemiaand outcomes regardless of cardiac arrest location, initial rhythm or time-to-ROSC. Conclusion We found no association between early post-arrest hyperoxemiaand unfavourable outcome. Subgroup analysis found no differential effect depending on arrest location, initial rhythm or time-to-ROSC.