Early prediction of acute traumatic coagulopathy: a validation of the COAST score using the German Trauma Registry

Abstract

Background Early identification of trauma patients at risk of developing acute traumatic coagulopathy (ATC) is important for initiating appropriate, coagulopathy-focused treatment. A clinical ATC prediction tool is a quick, simple method to evaluate risk. The COAST score was developed and validated in Australia but is yet to be validated on a European population. We validated the ability of the COAST score to predict coagulopathy and adverse bleeding-related outcomes on a large European trauma population.

Methods The COAST score was modified and applied to a retrospective cohort of trauma patients from the German Trauma Registry (TR-DGU). The primary outcome was coagulopathy defined as INR > 1.5 or aPTT > 60 s. Secondary outcomes were massive transfusion, blood product requirements, urgent surgery and mortality. The cohort included adult trauma patients with Injury Severity Score > 15 treated in Germany/Austria in 2012–2016.

Results 15,370 cases were included, of which 10.9% were coagulopathic. The COAST score performed with sensitivity 21.6% and specificity 94.2% at a threshold of COAST ≥ 3. The AUROC was 0.625 (95% CI 0.61–0.64). The COAST score also identified patients who had more massive transfusions (15.3% v 1.6%), more emergency surgery (49.6% v 28.2%), and higher early (21.7% v 5.4%) and total in-hospital mortality (38.1% v 14.5%).

Conclusion This large retrospective study demonstrated that the modified COAST score predicts coagulopathy with low sensitivity but high specificity. A positive COAST score identified a group of patients with bleeding-related adverse outcomes. This score appears adequate to act as an inclusion criterion for clinical trials targeting ATC.