医学
置信区间
内科学
前瞻性队列研究
接收机工作特性
曲线下面积
逻辑回归
单变量分析
多元分析
心脏病学
作者
Michael N. Cocchi,Justin D. Salciccioli,Tuyen T. Yankama,Maureen Chase,Parth V. Patel,Xiaowen Liu,Timothy J. Mader,Michael William Donnino
标识
DOI:10.1177/0885066619873315
摘要
Background: Outcome prediction after out-of-hospital cardiac arrest (OHCA) is difficult. We hypothesized that lactate and need for vasopressors would predict outcome, and that addition of a mitochondrial biomarker would enhance performance of the tool. Methods: Prospective observational study of OHCA patients presenting to an academic medical center September 2008 to April 2016. We conducted univariate and multivariate logistic regressions. Results: Patients were divided based on 2 variables: vasopressor status and initial lactate (<5 mmol/L, 5-10, ≥10). Three hundred fifty-two patients were evaluated; 249 had a lactate within 3 hours and were included. Patients on vasopressors had higher mortality (74% vs 40%; P < .001). A stepwise increase in mortality is associated with increasing lactate (45% lactate <5, 66% 5-10, and 83% ≥10; P < 001). Multivariable models with lactate group and vasopressors as predictors demonstrated excellent discrimination (area under the receiver operating curve [AUC]: 0.73 [95% confidence interval, CI: 0.66-0.79]; adjusted for additional covariates: AUC: 0.81 [95% CI: 0.75-0.86]). Thirty-six patients had cytochrome c levels available; among these 36, when comparing models with and without cytochrome c, there was no difference (AUC: 0.88 [95% CI: 0.76-1.00] vs AUC: 0.85 [95% CI: 0.73-0.98], respectively; P = .30). Conclusion: In this prospective validation, the combination of lactate and vasopressors in the immediate postarrest period is predictive of mortality. Cytochrome c offered minimal additional predictive power.
科研通智能强力驱动
Strongly Powered by AbleSci AI