Evaluating the Efficacy of Longitudinal Coagulation Biomarkers to Predict in-Hospital Death in Comatose TBI Patients: A Time-Dependent Roc Curve Approach.

Background: Coagulopathy is a significant risk factor for in-hospital mortality in patients with traumatic brain injury (TBI). Timely monitoring of coagulation biomarkers (CB) may be crucial for predicting TBI patient prognosis. The objective of this study was to assess the time-dependent accuracy of conventional CB in predicting TBI patient outcomes using an ROC curve based on a joint longitudinal and survival data analysis.

Methods: A retrospective cohort study was conducted on 132 out of 604 TBI patients admitted to Shohada-E-Tajrish Hospital in Tehran, Iran, between 2018 and 2020. CB (INR, PT, and aPTT) every 24 hours, and patient outcomes (death or survival) up to 10 days after admission were recorded. After choosing the best joint model longitudinal and survival data, the time-dependent ROC curve analysis was used, and the AUC, sensitivity, and specificity values were calculated in R 4.1.0.

Results: In this study, non-surviving TBI patients had significantly higher mean aPTT, PT, and INR values than survivors. In addition, aPTT demonstrated the strongest time-dependent predictive power (AUC>0.7) for up to the fourth day after admission, while PT and INR had AUC>0.7 for up to the third day after admission.

Conclusion: This study highlights that conventional coagulation biomarkers (aPTT, PT, and INR) demonstrate significant time-dependent efficacy in predicting in-hospital mortality in comatose TBI patients within the first three days of admission. Early recognition and proper management of coagulopathy are essential to improve outcomes and reduce morbidity and mortality in these patients during three days of admission.