Admission Arterial Partial Pressure of Carbon Dioxide (PaCO₂) and Outcomes After Out-of-Hospital Cardiac Arrest: A Single-Centre Retrospective Cohort Study

Abstract

Background:

The prognostic utility of arterial carbon dioxide tension (PaCO₂) following out-of-hospital cardiac arrest (OHCA) remains poorly defined. We hypothesized that elevated admission PaCO₂ is independently associated with increased one-month mortality and worse neurological outcomes in OHCA patients. This study aimed to evaluate the association between admission PaCO₂ and one-month survival and neurological recovery after OHCA.

Material and Methods:

In this single-center retrospective cohort study, 617 adult OHCA patients (2014–2023) were categorized into three PaCO₂ groups: 60 mmHg or less, 61 to 100 mmHg, and greater than 100 mmHg. The primary outcomes were one-month mortality and neurological outcome, classified as favorable (Cerebral Performance Category 1 or 2) or unfavorable (Cerebral Performance Category 3 to 5). Multivariable logistic regression was used to assess the independent association between PaCO₂ (analyzed as both a continuous and categorical variable) and the outcomes, adjusting for potential confounders.

Results:

One-month mortality was 73.6% (454/617) and rose across PaCO₂ groups (49.7%, 81.1%, and 90.7%; p < 0.001). Each 10-mmHg increase in PaCO₂ was associated with higher mortality after adjustment for measured confounders (adjusted odds ratio [aOR] 1.45, 95% CI 1.24-1.70; p < 0.001), with a concordant gradient for the ordinal CPC outcome (common aOR 1.53, 95% CI 1.31-1.79) and among survivors (aOR 1.91, 95% CI 1.25-2.91). The graded association was robust across categorization schemes but was attenuated after additional adjustment for concurrent acid-base and gas-exchange markers (lactate, base excess, PaO₂; aOR 0.80, 95% CI 0.64-1.01; p = 0.063), consistent with strong collinearity between PaCO₂ and acid-base status. Adding PaCO₂ to a clinical model produced a minimal incremental improvement in discrimination of uncertain clinical utility (AUC 0.794 to 0.816; ΔAUC 0.022, 95% CI 0.002-0.043; optimism-corrected AUC 0.802). Calibration was assessed via bootstrap validation (calibration slope: 0.98, 95% CI 0.89-1.07; calibration-in-the-large: -0.04, 95% CI -0.21 to 0.13), indicating adequate agreement between predicted and observed risks.

Conclusions: Higher admission PaCO₂ is independently associated with increased one-month mortality and worse neurological outcomes after OHCA. Admission PaCO₂ is a strong prognostic marker; whether its early correction improves outcomes warrants further study.