Deciding on “termination of resuscitation” (TOR) is a daily dilemma for any physician facing cardiac arrest. Due to the lack of evidence-based criteria and the scarcity of the existing guidelines, crucial arbitration to interrupt the resuscitation remains at the discretion of the practitioner. The aim was to evaluate with a quantitative method the existence of a bias in decisional factors belonging to the clinician in the TOR.

From RéAC registry, we have extracted data relative to 5,144 OHCA managed between January 2013 and September 2021. We conducted a statistical analysis by comparing, through a bootstrap, generalized linear mixt and non-mixt models. The outcome variable was whether each OHCA leads to the TOR. Utstein data were considered as fixed parameters. As for the random effect, the physician was integrated into the model. The significance of this effect was estimated by pseudo-R² method.

5,144 OHCA were included, involving 173 physicians. Median age was 71 (IQ [58; 83]), among which women accounted for 62%. Analysis found a standard derivation of 0.388 ‒ statistically significant (p<0.001) ‒ of physician effect on the TOR, with a pseudo-R² imputable to this factor of about 7%, to be compared to the 62% of the model variance explained by the other variables.   (Figure)

These results show the existence of an influence related to the physician on the TOR, which is almost as important as the no-flow duration. They plead for the exploration and testing of tools and guidelines to guide physicians in their decision process.