早期NLR、PLR、N/LP及其联合其他因素在预测重度创伤性颅脑损伤早期预后的研究

Study of early NLR,PLR,N/LP and combined of them with other indicators in prediction early outcomes of severe traumatic brain injury

目的探讨人外周血早期中性粒细胞-淋巴细胞比值(neutrophil to lymphocyte ratio,NLR)、血小板-淋巴细胞比值(platelet to lymphocyte ratio,PLR)和中性粒细胞-淋巴细胞和血小板比值(neutrophils to lymphocytes and platelets ratio,N/LP)在重度创伤性颅脑损伤(severe traumatic brain injury,sTBI)患者早期结果中的预测价值。方法回顾性分析2014年6月-2016年12月中国医科大学第七临床学院神经外科收治的95例sTBI患者的临床资料。比较预后良好组(n=36)和预后不良组(n=59)患者的早期NLR、PLR和N/LP差异,分析影响预后相关危险因素并绘制森林图,采用多因素logsitic回归确定独立危险因素并构建临床预测模型;绘制出受试者工作特征曲线(receiver operating characteristic curve,ROC),分析并比较NLR、PLR和N/LP单独或联合其他指标构建出的不同临床预测模型的差异。结果sTBI预后良好组和预后不良组早期NLR、PLR和N/LP比较,差异均有统计学意义(P<0.05);多因素logsitic回归分析提示,年龄、入院格拉斯哥昏迷评分(Glasgow coma scale,GCS)、NLR、PLR和N/LP是影响sTBI患者早期结果的独立危险因素,差异有统计学意义(P<0.05)。根据多因素logstic回归分析构建出19个临床预后预测模型,其中NLR+PLR+N/LP及其联合指标(年龄和GCS)模型曲线下面积(area under curve,AUC)均高于同组其他模型,分别为0.912、0.935、0.933和0.954;Age+GCS+NLR+PLR+N/LP预测模型在所有组别中AUC最大,表明该模型预测患者预后的价值最高。结论NLR、PLR和N/LP的升高与sTBI不良预后相关;早期NLR、PLR及N/LP联合年龄和GCS评分在sTBI早期结果预测中具有重要价值。

Objective To investigate the predictive value of early neutrophil-lymphocyte ratio(NLR),platelet-tolymphocyte ratio(PLR),neutrophil-lymphocyte-platelet ratio(N/LP)in human peripheral blood,and combination of them with other indicators in patients with severe traumatic brain injury(sTBI).Methods Clinical data of 95 sTBI patients admitted to department of neurosurgery,seventh clinical college of China Medical University from June 2014 to December 2016 were retrospectively analyzed.The differences in early NLR,PLR,and N/LP between the good prognosis group(n=36)and the poor prognosis group(n=59)were compared.The prognostic risk factors were analyzed and forest maps were drawn.The independent risk factors were determined by multi-factor logistic regression and clinical prediction models were established.The receiver operating characteristic curve(ROC)analysis was used and the differences of different clinical prediction models constructed by NLR,PLR,and N/LP alone or in combination with other indicators were compared.Results There were statistically differences in NLR,PLR,and N/LP between the good prognosis and poor prognosis(P<0.05).Multivariate logistic regression analysis showed that age,admission Glasgow coma scale(GCS)score,NLR,PLR,and N/LP were independent risk factors for early outcome of sTBI patients(P<0.05).Nineteen clinical prognostic models were constructed based on multi-factor logistic regression analysis.The area under curve(AUC)of NLR+PLR+N/LP and its combined with other indexes(age and GCS)were higher than those of other models,which were 0.912,0.935,0.933 and 0.954 respectively.Age+GCS+NLR+PLR+N/LP prediction model had the highest AUC in all groups,which indicated that the model had the highest prognostic value.Conclusion The elevation of NLR,PLR,and N/LP are associated with poor prognosis of sTBI.Model based on early NLR,PLR,and N/LP combined with age and GCS has great value in predicting early sTBI outcome.