颅脑创伤术后呼吸衰竭危险因素分析及相关模型的构建

Risk factors analysis of respiratory failure after craniocerebral trauma and construction of related models

目的探讨颅脑创伤术后呼吸衰竭的危险因素并构建相关模型。方法选择2016年1月至2019年7月于我院进行颅脑创伤术治疗的312例患者作为研究对象,以患者进行颅脑创伤术后是否发生呼吸衰竭作为分组依据,将发生呼吸衰竭的56例患者纳入发生组,将未发生呼吸衰竭256例患者纳入未发生组,对两组患者的临床资料进行单因素和多因素Logistic回归分析,并建立预测模型,用灵敏度、特异度、ROC曲线下面积及预测正确率评价模型的预测效果。结果单因素结果表明:两组患者的术后采用机械通气治疗、合并肺部感染、入院时GCS评分、CRP、PCT相比较,差异有统计学意义(P<0.05);多因素Logistic回归分析结果显示:合并肺部感染、入院时GCS评分、CRP、PCT是颅脑创伤术后呼吸衰竭的危险因素(P<0.05);颅脑创伤术后呼吸衰竭的概率预测模型为Prob=1/[1+e^(39.211-3.066*合并肺部感染-0.609*入院时GCS评分-0.166*CRP-0.093*PCT)],模型的灵敏度为96.20%,特异度为96.70%,ROC曲线下面积为0.983,预测正确率为84.62%。结论合并肺部感染、入院时GCS评分、CRP、PCT是颅脑创伤术后呼吸衰竭的危险因素,根据颅脑创伤术后呼吸衰竭的独立危险因素构建的预测模型,能有效评估术后呼吸衰竭的风险。

Objective To explore the risk factors of respiratory failure after craniocerebral trauma and to construct related models.Methods A total of 312 patients undergoing craniocerebral trauma in our hospital from January 2016 to July 2019 were selected as study subjects.Based on whether patients had respiratory failure after traumatic brain injury or not,56 patients with respiratory failure were included in the occurrence group,and 256 patients without respiratory failure were included in the non-occurrence group.Univariate and multivariate Logistic regression was used to analyze the clinical data of the two groups,and a prediction model was established.Sensitivity,specificity,area under the ROC curve and diagnostic accuracy rate were used to evaluate the prediction effect of the model.Results The univariate results showed that there was a statistically significant difference in pulmonary infection,GCS score,CRP,and PCT at admission between the two groups(P<0.05).Multivariate logistic regression analysis showed that postoperative mechanical ventilation,combined with pulmonary infection,GCS score at admission,CRP,and PCT were risk factors for respiratory failure after traumatic brain injury(P<0.05).Probability prediction model for respiratory failure after craniocerebral trauma was Prob=1/[1+e^(39.211-3.066*with pulmonary infection-0.609*GCS score at admission-0.166*CRP-0.093*PCT)].The sensitivity of the model was 96.20%,the area under the ROC curve was 0.983,and the diagnostic accuracy was 84.62%.Conclusion Combined with pulmonary infection,GCS score,CRP,and PCT at admission are risk factors for respiratory failure after traumatic brain injury.A predictive model based on independent risk factors for respiratory failure after craniocerebral trauma can effectively assess the risk of respiratory failure after surgery.