基底核区高血压脑出血患者神经内镜术后再出血的影响因素及预测模型构建

The influencing factors and predictive model construction of postoperative rebleeding in patients with hypertensive basal ganglia hemorrhage in the basal ganglia region after neuroendoscopic surgery

目的分析基底核区高血压脑出血(hypertensive basalganglia hemorrhage,HBGH)神经内镜术后再出血的影响因素,并构建预测模型。方法回顾性分析210例行神经内镜手术的HBGH病例资料,依据术后再出血情况分为再出血组(n=38)与对照组(n=172)。比较再出血组与对照组资料,建立回归模型,分析HBGH神经内镜术后再出血的影响因素。基于回归模型分析得到的主要影响因素,构建列线图预测模型,并采用校准曲线、决策曲线评价模型预测价值。结果入院时GCS评分低、血肿量大、发病至手术时间短、术前纤维蛋白原(Fibrinogen,FIB)水平高是HBGH神经内镜术后再出血发生的危险因素(P<0.05);基于上述危险因素构建列线图模型发现,入院时GCS评分低、血肿量大、发病至手术时间短、术前FIB水平高的HBGH患者发生术后出血的风险较大;使用Bootstrap进行内部验证,绘制标准曲线,校准曲线和Y-X直线相近,C指数为0.988,决策曲线净受益率最大值为0.181,模型准确度、区分度良好。结论入院时GCS评分低、血肿量大、发病至手术时间短、术前FIB水平高可能是HBGH神经内镜术后再出血的危险因素,基于上述危险因素构建的预测模型对于HBGH神经内镜术后再出血的发生有较高预测价值。

Objective To analyze the influencing factors of postoperative rebleeding in patients with hypertensive basal ganglia hemorrhage(HBGH)in the basal ganglia region and construct a predictive model.Methods A retrospective analysis was conducted on 210 HBGH patients who underwent neurosurgery.The patients were divided into the rebleeding group(n=38)and the control group(n=172)based on the postoperative rebleeding.The data of the two groups were compared,and a regression model was established to analyze the influencing factors of rebleeding after neurosurgery for HBGH.Based on the main factors obtained from the regression model analysis,a column chart prediction model was constructed,and calibration curve and decision curve were used to evaluate the predictive value of the model.Results Low GCS score at admission,large hematoma volume,short time from onset to surgery,and high preoperative fibrinogen(FIB)level were identified as risk factors for rebleeding after neurosurgery for HBGH(P<0.05).A ROC curve was constructed based on these risk factors,and it was found that patients with low GCS score,large hematoma volume,and short time from onset to surgery,and high preoperative FIB level had a higher risk of postoperative bleeding.The model was validated using bootstrap and internal validation,and a standard curve,calibration curve,and Y-X straight line were drawn.The C index was 0.988,and the net benefit rate of the decision curve was the maximum at 0.181.The model had good accuracy and discriminatory ability.Conclusions Low GCS score,large hematoma volume,short onset to surgery time,and high preoperative FIB level at admission may be risk factors for postoperative rebleeding in HBGH patients undergoing neuroendoscopy.A predictive model constructed based on these risk factors has high predictive value for the occurrence of postoperative rebleeding in HBGH patients undergoing neuroendoscopy.