Resuscitation From Out-of-Hospital Cardiac Arrest: When Is EtCO₂ Reliably Associated With ROSC?

Abstract

Exhaled end-tidal carbon dioxide (EtCO₂) trajectory is associated with out-of-hospital cardiac arrest (OHCA) outcomes. However, the minimum EtCO₂ monitoring duration needed to discriminate return of spontaneous circulation (ROSC) from non-ROSC remains unknown. We sought to determine the EtCO₂ trajectory observation time required to differentiate ROSC from non-ROSC patients. We performed a secondary analysis of the cluster-randomized Pragmatic Airway Resuscitation Trial (PART), which assessed endotracheal intubation or laryngeal tube strategies in OHCA resuscitation. We summarized mean EtCO₂ in 1-minute epochs over the resuscitation. Cases were stratified a priori by: (1) witnessed versus unwitnessed status, and (2) initial EtCO₂: low (≤30 mm Hg), moderate (31-49), and high (≥50). Within each stratum, group-based trajectory modeling (GBTM) was used to identify latent EtCO₂ trajectory classes, and patients were categorized into an upward or downward trajectory. To balance trajectory groups on baseline characteristics including age, sex, race, initial rhythm, location, and bystander CPR, we applied inverse probability of treatment weighting. We fit weighted pooled logistic regression models to estimate risk ratios (RRs) for ROSC comparing upward versus downward EtCO₂ trajectories. Within each stratum, we identified the earliest minute when CIs between upward versus downward EtCO₂ trajectories no longer overlapped. EtCO₂ data were available for 1168 patients: 452 (38.6%) witnessed and 716 (61.1%) unwitnessed. Patients were predominantly men (63.5%), with a median age of 65 years (Q1, Q3: 53-75), majority White race (51.3%), and presenting in a nonpublic setting (85.4%). Overall ROSC was 18.2%: 30.5% of witnessed and 10.5% of unwitnessed. Among witnessed arrests, 95% CI for upward versus downward EtCO₂ trajectories no longer overlapped at 8 minutes for low initial EtCO₂ (RR, 3.06; 95% CI, 1.49, 6.71), 12 minutes for moderate EtCO₂ (RR, 1.95; 95% CI, 1.23, 3.48), and 21 minutes for high EtCO₂ (RR, 2.12; 95% CI, 1.30, 3.73). Among unwitnessed arrests, nonoverlapping CIs were first observed at 7 minutes (RR, 3.56; 95% CI, 1.53, 10.37). Depending on witness status and initial EtCO₂, between 7 and 21 minutes of monitoring are needed to reliably differentiate upward from downward EtCO₂ trajectories during OHCA resuscitation. Dynamic EtCO₂ trajectory monitoring may provide early prognostic information to guide resuscitation.