Objective To determine whether acute traumatic coagulopathy (ATC) should be combined with the trauma and injury severity score (TRISS) to predict outcome in severe trauma patients and investigate effects of the change in coagulation state during early resuscitation on the actual survival rate.

Methods This was a retrospective study. Significant variables that affected 28-day mortality were analysed using multivariate logistic regression. Study patients were classified into three groups: no coagulopathy, mild coagulopathy or severe coagulopathy. Concordance between actual and predicted survival rates were compared for each group. The predicted survival rate was calculated using the TRISS method. The study also determined whether changes in the coagulation state during inhospital resuscitation affected the relationship between actual and predicted survival in patients who had rechecked coagulation profile within 12 h after presentation.

Results Data from a total of 336 patients were analysed. At presentation, 20.8% of the study patients had mild coagulopathy, whereas 7.7% had severe coagulopathy. Age, injury severity score, revised trauma score and presence of ATC at presentation were independently associated with 28-day mortality. Actual survival was significantly lower than predicted survival in the mild and severe coagulopathy groups. Aggravation of coagulation state from normal or mild to severe coagulopathy or persistent severe coagulopathy during inhospital resuscitation mainly contributed to the discrepancy between actual and predicted survival.

Conclusions ATC decreased actual survival more than expected. ATC should be combined with TRISS to predict trauma outcome in severely injured patients. Improvement in coagulopathy during resuscitation may reduce the incidence of preventable death after trauma.