Abstract: Objective To investigate the risk factors of deep venous thrombosis of lower extremity in severe neurological patiens and to construct the relevant nomogram prediction model based on this analysis. Methods The relevant clinical data of 75 severe neurological patients admitted to the neurosurgical intensive care unit from January 2019 to March 2022 were analyzed retrospectively.According to the Results of the lower limb color Doppler ultrasound examination,they were divided into thrombotic group and non-thrombotic group.Comparing the related clinical data between the two groups,univariate analysis and multivariate logistic regression analysis were used to determine the risk factors of deep venous thrombosis in these neurosevere patients.In addition,the Nomogram model for the risk prediction of deep venous thrombosis of lower limbs was constructed.The reliability of the prediction model was also evaluated using the receiver operator characteristic curve (ROC). Results The univariate analysis showed:there was not significant difference in the data of gender,history of hypertension,cerebral hernia during the course and platelet count.The age of 65 years,muscle strength of lower limb<3,GCS score≤8,lower limb braking and deep venipuncture and plasma D-dimer> 0.5 mg/L are the factors affecting deep venous thrombosis of lower extremity in severe neurological patients ; Further multivariate logistic regression analysis showed that aged 65 years,muscle strength of lower limb <3,lower limb braking and deep venipuncture and plasma D-dimer> 0.5 mg/L were independent risk factors for deep venous thrombosis of lower extremity in severe neurologic patients.The Nomogram prediction model of lower limb DVT formation was established after the assignment of these risk factors indicators,Its ROC showed AUC=0.831 (95% CI:0.837～0.975,P<0.001),The predictive sensitivity was 82.9%,The predicted specificity was 88.2%. Conclusion The nomogram model of risk prediction of deep venous thrombosis of lower extremity in severe neurological patients is simple,easy to master and has good calibration,which is of great significance for the prediction of DVT,reducing the risk and improving the prognosis of patients in the neuro-critical patients.

Key words: Severe neurological patients, Deep venous thrombosis of lower extremity, Nomogram model

摘要： 目的 探讨神经重症患者下肢深静脉血栓形成的危险因素,以构建下肢深静脉血栓形成的风险预测的Nomogram模型。方法 回顾性选取我院2019年1月至2022年3月期间神经外科重症病房收治的75例患者的相关临床资料,按照下肢彩色超声检查结果将其分为血栓组和非血栓组,比较分析两组患者的相关临床资料,采用单因素分析和多因素Logistic回归分析判断神经重症患者发生下肢深静脉血栓形成的危险因素,并以此构建下肢深静脉血栓形成的风险预测的Nomogram模型,同时采用受试者工作特征曲线(ROC)评估预测模型的可靠性。结果 单因素分析结果显示:患者性别、有无高血压病史、病程中有无脑疝以及患者的血小板计数两组间无统计学意义,而年龄≥65岁、下肢肌力<3级、GCS评分<8分、下肢制动和有下肢深静脉穿刺以及血浆D-二聚体＞0.5 mg/L是神经重症患者发生下肢深静脉血栓形成的影响因素;在此基础上再进行多因素Logistic回归分析得出以下结果:年龄≥65岁、下肢肌力<3级、下肢制动和有下肢深静脉穿刺以及血浆D-二聚体＞0.5 mg/L为神经重症患者发生下肢深静脉血栓形成的独立危险因素,再以这些危险因素指标赋值后建立下肢DVT形成的Nomogram预测模型,其ROC显示AUC=0.831(95% CI:0.837～0.975,P＜0.001),预测敏感度为82.9%,预测特异度为88.2%。结论 建立于临床资料基础上的神经重症患者下肢深静脉血栓形成的风险预测nomogram模型简单、易于掌握,具有较好的校准度,对于临床神经重症患者发生DVT的预判、降低其发生风险和改善患者的预后有较大的意义。

关键词: 神经重症患者, 下肢深静脉血栓, Nomogram模型