RESUMEN
OBJECTIVE: To discuss the effects of the hematocrit (Hct) in patients with traumatic brain injury after decompressive craniectomy (DC). METHODS: Demographic data, inspection and treatment procedures, and 30-day prognosis were obtained for 158 patients with head injury who underwent unilateral DC in our hospital between January 2013 and June 2018. Uni- and multivariate logistic regression was applied to analyze independent risk factors for 30-day outcome. The quantitative analysis of postoperative Hct, ΔHct (postoperative Hct minus initial Hct), and their combination for the prognosis of patients with TBI was displayed graphically using receiver operating characteristic (ROC) curves. Multiple linear regression was used to explore factors influencing postoperative Hct and ΔHct. RESULTS: Short-term mortality was 29.7%. Uni- and multivariate logistic regression analysis showed that age (odds ratio [OR], 1.064; P = 0.024), Glasgow Coma Scale score (OR, 0.711; P = 0.027), Injury Severity Score (ISS) (OR, 1.156; P = 0.047), midline shift in millimeters (OR, 1.809; P <0.001), postoperative Hct (OR, 0.743; P = 0.001), and ΔHct (OR, 1.242; P =0.048) were independent risk factors for short-term death. In ROC curves, a combination of postoperative Hct and ΔHct showed the highest sensitivity (77.5%) and highest specificity (89.4%). When using this combination to predict prognosis, we could achieve an accuracy of 94.5%. ISS (ß = -0.172, P = 0.022), initial Hct (ß = 0.243, P = 0.001), principal hematoma location (ß = -2.628, P < 0.001), hours of operation (ß = -0.884, P = 0.048), and colloid quantity (ß = -0.002, P = 0.001) were independent contributing factors for ΔHct, which was similar to postoperative Hct. CONCLUSIONS: A combination of postoperative Hct and ΔHct could better predict short-term survival of patients with TBI. Developing an appropriate treatment strategy to increase postoperative Hct and reduce the ΔHct may be good for the short-term prognosis of patients with TBI after DC.