Determining prognostic indicator for anticoagulant therapy in sepsis-induced disseminated intravascular coagulation

Abstract

Abstract Background There is no reliable indicator that can assess the treatment effect of anticoagulant therapy for sepsis-associated disseminated intravascular coagulation (DIC) in the short term. The aim of this study is to develop and validate a prognostic index identifying 28-day mortality in septic DIC patients treated with antithrombin concentrate after a 3-day treatment. Methods The cohort for derivation was established utilizing the dataset from post-marketing surveys, while the cohort for validation was acquired from Japan’s nationwide sepsis registry data. Through univariate and multivariate analyses, variables that were independently associated with 28-day mortality were identified within the derivation cohort. Risk variables were then assigned a weighted score based on the risk prediction function, leading to the development of a composite index. Subsequently, the area under the receiver operating characteristic curve (AUROC). 28-day survival was compared by Kaplan–Meier analysis. Results In the derivation cohort, 252 (16.9%) of the 1492 patients deceased within 28 days. Multivariable analysis identified DIC resolution (hazard ratio [HR]: 0.31, 95% confidence interval [CI]: 0.22–0.45, P < 0.0001) and rate of Sequential Organ Failure Assessment (SOFA) score change (HR: 0.42, 95% CI: 0.36–0.50, P < 0.0001) were identified as independent predictors of death. The composite prognostic index (CPI) was constructed as DIC resolution (yes: 1, no: 0) + rate of SOFA score change (Day 0 SOFA score–Day 3 SOFA score/Day 0 SOFA score). When the CPI is higher than 0.19, the patients are judged to survive. Concerning the derivation cohort, AUROC for survival was 0.76. As for the validation cohort, AUROC was 0.71. Conclusion CPI can predict the 28-day survival of septic patients with DIC who have undergone antithrombin treatment. It is simple and easy to calculate and will be useful in practice.