Intra-arrest transport versus continued on-scene resuscitation in patients with out-of-hospital cardiac arrest by first documented rhythm: a nationwide cohort study using time-dependent propensity score sequential matching

Whether to transport patients during cardiac arrest or to prioritize on-scene resuscitation presents a critical dilemma. This study aims to evaluate the association between intra-arrest transport versus continued on-scene resuscitation, and survival after out-of-hospital cardiac arrest (OHCA) using time-dependent propensity score sequential matching. We consecutively collected data on adult OHCA between 2021 and 2022 using a nationwide, prospective, population-based registry of out-of-hospital cardiac arrest in Japan. Patients with OHCA were divided into two cohorts based on their first documented electrocardiographic rhythm: shockable and nonshockable. We performed a time-dependent propensity score-matching analysis to address resuscitation bias and differences in background characteristics. Patients undergoing intra-arrest transport were matched with patients at risk of intra-arrest transport (continued on-scene resuscitation) at the same time in each cohort. Then, outcomes between the matched groups were compared. The primary endpoint was survival at one month or discharge from the hospital within one month. A total of 13,826 patients with shockable rhythm and 187,036 patients with non-shockable rhythm were eligible for our analysis. After time-dependent propensity score matching, the matched cohorts were well-balanced. More than 90% of patients initially assigned to continued on-scene resuscitation later underwent intra-arrest transport. In the shockable cohort, the corrected risk ratio (RR) for the primary endpoint in the intra-arrest transport group compared with the continued on-scene resuscitation group was 1.03 (0.96–1.10, P = 0.443). In the nonshockable cohort, the corrected RR for the primary endpoint in the intra-arrest transport group compared with the continued on-scene resuscitation group was 0.84 (0.78–0.89, P < 0.001) (P for interaction: <0.001). Intra-arrest transport and continued on-scene resuscitation showed similar survival among patients with shockable rhythms, whereas continued on-scene resuscitation was associated with better survival among patients with nonshockable rhythms. Given the high rate and short delay of subsequent intra-arrest transport in the continued on-scene resuscitation group, our analysis more closely reflects a comparison of earlier versus later intra-arrest transport during ongoing on-scene resuscitation.