Introduction: Prediction of subsequent mortality during the first 24 hours after resuscitation from out-of-hospital cardiac arrest (OHCA) remains challenging. We assessed markers of critical illness severity within the first 24 hours as predictors of inpatient mortality after OHCA.
Hypothesis: Organ failure severity scores, initial serum lactate, and vasopressor requirements will predict inpatient mortality after resuscitation from OHCA.
Methods: Historical cohort study of 279 patients receiving targeted temperature management at a single tertiary care hospital after OHCA from 12/2005 to 9/2016. Vasopressor requirements were quantified using the Vasoactive-Inotrope Score (VIS). Sequential Organ Failure Assessment, APACHE-III scores and APACHE-IV predicted mortality were automatically calculated. Hospital survivors and inpatient deaths were compared using Wilcoxon test. Area under the receiver-operator characteristic curve (AUROC) analysis was followed by multivariate logistic regression.
Results: Mean age was 62.6 years and 72% were male. Arrest rhythm was shockable in 87%, 70% required vasopressors and 73% underwent coronary angiography. Inpatient deaths (n =112, 40%) had significantly (all p <0.001) higher peak VIS (47.1 vs. 15.6), initial serum lactate (6.6 vs. 4.3), SOFA (9.2 vs. 7.6) and APACHE-III (112.2 vs. 98.5) scores and APACHE-IV predicted mortality (66.1% vs. 56.1%). These variables were all significant univariate predictors of mortality (p <0.001), with the highest AUROC values of 0.68 for initial serum lactate. Multivariate logistic regression identified inpatient coronary angiogram (OR 8.4, 95% CI 4.0-18.4; p <0.0001), peak VIS (OR 1.01, 95% CI 1.00-1.02; p = 0.01) and initial serum lactate (OR 1.22, 95% CI 1.11-1.35; p <0.0001) as independent predictors of mortality when correcting for age, shockable rhythm and APACHE-III score. SOFA, APACHE-III and APACHE-IV were not independent predictors of mortality (all p >0.1).
Conclusions: Peak vasopressor requirements and initial serum lactate independently predicted mortality in patients resuscitated from OHCA, while severity of illness scores did not. This emphasizes the need for disease-specific early mortality prediction scores for OHCA patients.