We aimed to compare regional cerebral oxygen saturation (rSO₂) levels during cardiopulmonary resuscitation (CPR), performed either manually or using a mechanical chest compression device (MCCD), in witnessed cardiac arrest cases in the emergency department (ED), and to evaluate the effects of both the CPR methods and perfusion levels on patient survival and neurological outcomes.

This single-center, randomized study recruited patients aged ≥18 years who had witnessed a cardiopulmonary arrest in the ED. According to the relevant guidelines, CPR was performed either manually or using an MCCD. Simultaneously, rSO₂ levels were continually measured with near-infrared spectroscopy.

Seventy-five cases were randomly distributed between the MCCD (n = 40) and manual CPR (n = 35) groups. No significant difference in mean rSO₂ levels was found between the MCCD and manual CPR groups (46.35 ± 14.04 and 46.60 ± 12.09, respectively; p = 0.541). However, a significant difference in rSO₂ levels was found between patients without return of spontaneous circulation (ROSC) and those with ROSC (40.35 ± 10.05 and 50.50 ± 13.44, respectively; p < 0.001). In predicting ROSC, rSO₂ levels ≥24% provided 100% sensitivity (95% confidence interval [CI] 92–100), and rSO₂ levels ≥64% provided 100% specificity (95% CI 88–100). The area under the curve for ROSC prediction using rSO₂ levels during CPR was 0.74 (95% CI 0.62–0.83).

A relationship between ROSC and high rSO₂ levels in witnessed cardiac arrests exists. Monitoring rSO₂ levels during CPR would be useful in CPR management and ROSC prediction. During CPR, MCCD or manual chest compression has no distinct effect on oxygen delivery to the brain.

Trial registration: clinicaltrials.gov identifier: NCT03238287.