The predictive value of the change of disturbance coefficient for rebleeding in patients with brain trauma and the analysis of the optimal cutoff value
生物电阻抗扰动系数变化对脑外伤患者再出血的预警价值及最佳截断值分析

摘要

目的 探讨无创脑电阻抗监测技术中扰动系数变化对脑外伤患者出血量增加的预警价值，并确定其最佳预警截断值。方法 回顾性分析2022年1月至2024年10月期间在本院（三级甲等医院）神经外科收治的、接受无创脑水肿动态监护仪（BORN-BE系列）监测的67例脑外伤患者临床资料。重点观察其中12例出血量增加的患者，分析其再出血前后扰动系数的变化幅度，应用受试者工作特征（ROC）曲线确定预警再出血的最佳扰动系数截断值；同时分析扰动系数与格拉斯哥昏迷评分（GCS）、格拉斯哥预后评分（GOS）的相关性，及血肿清除术对扰动系数的影响。结果 12例出血量增加的扰动系数在再出血时较出血前平均上升18个单位（P<0.05）；ROC曲线分析显示，扰动系数预警再出血的最佳截断值为141.5（AUC=0.89，P<0.001），灵敏度85%，特异度92%；扰动系数与入院同期GCS评分呈显著正相关（r=0.765，P<0.05），与出院时GOS评分呈显著负相关（r=-0.532，P<0.05）；37例行血肿清除术的患者，术后扰动系数较术前平均下降17.3个单位（P<0.05）。结论 无创脑电阻抗监测技术中的扰动系数变化是脑外伤患者出血量变化的有效预警指标，扰动系数上升≥18或绝对值>141.5可作为提示出血量增加、需及时行影像学复查和临床干预的重要阈值，该技术对改善脑外伤患者预后具有重要临床价值。

Abstract

Objective To investigate the predictive value of the change of disturbance coefficient in non-invasive cerebral resist monitoring technology for increased hemorrhage in patients with brain trauma, and to determine its optimal predictive cutoff value.
Methods A retrospective analysis was conducted on the clinical data of 67 patients with brain trauma admitted to the Department of Neurosurgery of our hospital (a tertiary Class A hospital) from January 2022 to October 024 and monitored by a non-invasive cerebral edema dynamic monitor (BORN-BE series). Among them, 12 patients who had reble were observed, and the amplitude of the change of disturbance coefficient before and after rebleeding was analyzed. The receiver operating characteristic (ROC) curve was applied to determine the optimal coefficient cutoff value for predicting rebleeding; the correlation between the disturbance coefficient and the Glasgow Coma Scale (GCS), the Glasgow Outcome Scale (GOS) and the effect of hematoma evacuation surgery on the disturbance coefficient were also analyzed.
Results The disturbance coefficient of 12 rebleeding patients increased by an average of 8 units at the time of rebleeding compared to before bleeding (P < 0.05); ROC curve analysis showed that the optimal cutoff value of the disturbance coefficient predicting rebleeding was 141.5 (AUC = 0.89, P < 0.001), with a sensitivity of 8% and a specificity of 92%; the disturbance coefficient was significantly positively correlated with the GCS score at admission (r = 0.765, P <0.05), and significantly negatively correlated with the GOS score at discharge (r = -0.532, P < 0.05); among37 patients who underwent hematoma evacuation surgery, the disturbance coefficient decreased by an average of 17.3 units after surgery compared to before surgery (P < 005).
Conclusion   The change of disturbance coefficient in non-invasive cerebral resistive monitoring technology is an effective predictive index for rebleeding in patients with brain, and an increase of ≥18 or an absolute value >141.5 can be used as an important threshold to indicate rebleeding, which requires timely imaging and clinical intervention. This technology has important clinical value for improving the prognosis of patients with brain trauma.