摘要Objective To investigate the effects of propofol and sevoflurane on neurological recovery of traumatic brain injury(TBI)patients in the early postoperative stage.Methods We retrospectively analyzed the clinical data of TBI patients who underwent craniotomy or decompressive craniectomy.Generalized additive mixed model(GAMM)was used to analyze effects of propofol and sevoflurane on Glasgow Coma Scale(GCS)on postoperative days 1,3,and 7.Multivariate regression analysis was used to analyze effects of the two anesthetics on Glasgow Outcome Scale(GOS)at discharge.Results A total of 340 TBI patients were enrolled in this study.There were 110 TBI patients who underwent craniotomy including 75 in the propofol group and 35 in the sevoflurane group,and 134 patients who underwent decompressive craniectomy including 63 in the propofol group and 71 in the sevoflurane group.It showed no significant difference in GCS at admission between the propofol and the sevoflurane groups among craniotomy patients(β = 0.75,95%CI:-0.55 to 2.05,P = 0.260).However,elevation in GCS from baseline was 1.73 points(95%CI:-2.81 to-0.66,P = 0.002)less in the sevoflurane group than that in the propofol group on postoperative day 1,2.03 points(95%CI:-3.14 to-0.91,P＜0.001)less on day 3,and 1.31 points(95%CI:-2.43 to-0.19,P = 0.022)less on day 7.The risk of unfavorable GOS(GOS 1,2,and 3)at discharge was higher in the sevoflurane group(OR = 4.93,95%CI:1.05 to 23.03,P = 0.043).No significant difference was observed among two-group decompressive craniectomy patients in GCS and GOS.Conclusions Compared to propofol,sevoflurane was associated with worse neurological recovery during the hospital stay in TBI patients undergoing craniotomy.This difference was not detected in TBI patients undergoing decompressive craniectomy.