LAUSR

BACKGROUND It is well known that thrombocytopenia occurs in patients with traumatic brain injury (TBI), and its incidence increases with the severity of injury. We aimed to determine whether postoperative thrombocytopenia in patients with TBI is associated with poor clinical outcomes. METHODS This was a retrospective cohort study of a large international database called the Medical Information Mart for Intensive Care III (MIMIC-III), which included 1093 patients who underwent TBI surgery. Hospital mortality was the primary endpoint of this study. RESULTS Multivariate logistic regression analysis revealed non-thrombocytopenia was significantly associated with a decreased hospital mortality (adjusted odds ratio [OR] 0.49; 95% confidence interval [CI] 0.33-0.75; p = .01). In addition, platelet counts increased over time in both survivors and non-survivors, according to generalized additive mixed model (GAMM). However, the platelet count increased more noticeably in the survivors than in the non-survivors and the difference in platelet count between the two groups showed a trend toward increasing within 7 days after surgery. This difference increased by 7.97 per day on average. CONCLUSIONS Patients with TBI who experienced postoperative thrombocytopenia were more likely to have a poor short-term prognosis. In addition, we found that the rate of platelet growth over time varied significantly between the survival and non-survival groups. Patients with TBI who experienced a greater early increase in platelet count had a lower mortality rate.