Study on the prediction of early adverse outcomes after decompressive craniectomy in patients with severe traumatic brain injury by combined predictors
Objective To explore the influencing factors of early adverse outcomes in patients with severe traumatic brain injury(sTBI)after decompressive craniectomy(DC),and to construct combined predictors to predict early adverse outcomes.Methods The clinical data of 114 patients with sTBI who received DC treatment in our hospital from July 2019 to December 2022 were retrospectively analyzed.Univariate analysis was used to analyze the possible factors affecting the early adverse outcomes after DC in sTBI patients.After that,binary Logistic regression analysis was used to explore the independent influencing factors.The B value of independent influencing factors was used to construct combined predictors,and the receiver operating characteristic(ROC)curve was drawn to predict the early adverse outcomes.Results A total of 70 patients with Glasgow Outcome Scale(GOS)score of 4 points at discharge were included in the early good outcome group,a total of 44 patients with GOS score of 2-3 points were included in the early adverse outcome group.There were significant differences in age,pupil changes,Glasgow Coma Scale(GCS)score at admission,time from injury to operation and surgical methods between the two groups(P<0.05).Old age and long time from injury to operation were risk factors for early adverse outcomes in patients after DC,and high GCS score at admission was a protective factor for early adverse outcomes in patients after DC(P<0.05).The ROC curve analysis of the combined predictors showed that the areas under curve(AUC)was 0.944,the sensitivity was 95.45%,and the specificity was 88.57%.Conclusion The patients'age,GCS score at admission and time from injury to operation can be used as a combined predictor to predict the early adverse outcomes of sTBI patients after DC,which can provide help for clinical treatment.
severe traumatic brain injurydecompressive craniectomycombined predictor
万方数据
维普
