Association between mild hypercapnia and hospital mortality in patients admitted to the intensive care unit after cardiac arrest: A retrospective study.

Abstract

Mild hypercapnia may increase cerebral oxygenation and attenuate cerebral injury in post-cardiac arrest patients. However, its association with hospital mortality has not been evaluated. We conducted a retrospective multi-center study of prospectively collected data of all cardiac arrest patients admitted to the ICU between 2014 and 2015. Different kinds of arterial carbon dioxide tension (PaCO₂), including time-weighted mean PaCO₂, mean PaCO₂, admission PaCO₂ and proportion of time spent in four PaCO₂ categories (hypocapnia, normocapnia, mild hypercapnia, and severe hypercapnia) were used to explore the association with outcomes. Restricted cubic splines models were built to evaluate the association between PaCO₂ and odds ratio for hospital mortality in overall population and subgroups of different pH levels (acidosis, normal pH and alkalosis). 2783 post-cardiac arrest patients in 150 ICUs were included. 933 (33.5%) were classified into the hypocapnia (PaCO₂ < 35 mmHg), 1088 (39.1%) into the normocapnia (35-45 mmHg), 472 (17%) into the mild hypercapnia (45-55 mmHg) and 390 (10.4%) into the severe hypercapnia (>55 mmHg) group. Compared with normocapnia, mild hypercapnia was not associated with higher hospital survival probability (OR 1.08 [95% CI 0.84-1.38, p = 0.558]). Time spent in the normocapnia was associated with good outcome (OR 0.98 [95% CI 0.97-0.99, p < 0.001], for every 5 percentage point increase in time), but mild hypercapnia was not (OR 1 [95% CI 0.98-1.01, p = 0.542]). Cox-proportional hazards models supported these findings. Associations between PaCO₂ and hospital mortality were not statistically significant in normal pH and alkalosis subgroups. PaCO₂ has a U-shaped association with odds ratio for hospital mortality, with mild hypercapnia not having a higher hospital survival probability than normocapnia in post-cardiac arrest patients.