列线图模型预测高龄急性脑梗死患者rt-PA溶栓近期预后的临床价值

Clinical Value of Nomogram Model in Predicting Short-term Prognosis of Elderly Patients with Acute Cerebral Infarction after Rt-PA Thrombolysis

目的观察列线图模型预测高龄急性脑梗死患者重组组织型纤溶酶原激活剂(rt-PA)溶栓近期预后的临床准确性和实用性。方法回顾性分析2018年9月1日—2022年1月1日收治的年龄≥80岁且使用rt-PA溶栓的急性脑梗死临床资料,根据患者出院后6个月改良Rankin量表(mRS)评分,分为预后良好组(mRS评分0~2分)和预后不良组(mRS评分≥3分),利用多因素Logistic回归分析导致高龄急性脑梗死rt-PA溶栓患者近期预后不良的危险因素,基于分析得出的危险因素利用R语言构建列线图模型,并利用受试者工作特征(ROC)曲线、Booststip方法及临床决策曲线验证该模型的准确性和临床决策的获益性。结果最终纳入预后良好组234例,预后不良组108例。多因素Logistic回归分析显示:入院时美国国立卫生研究院卒中量表评分、脑梗死部位、入院时收缩压、发病至溶栓时间、溶栓后并发脑出血、心房颤动史、入院时血糖高是高龄急性脑梗死rt-PA溶栓患者近期预后不良的危险因素(P<0.01)。ROC曲线验证列线图模型显示,构建的列线图模型预测高龄急性脑梗死rt-PA溶栓患者近期预后的准确性较高,其中曲线下面积为0.854,95%CI为0.71,0.92。Bootstrap方法重复抽样1000次验证列线图,发现校准曲线的平均绝对误差为0.017。临床决策曲线显示,当发生阈值为0.06~0.86时该模型的适用性最佳。结论列线图模型预测高龄急性脑梗死rt-PA溶栓患者近期预后的临床准确性较高,临床医生可借鉴该模型对高危人群进行积极干预,对临床决策制定具有较高的参考价值。

Objective To observe the clinical accuracy and practicability of nomogram model in predicting the short-term prognosis of elderly patients with acute cerebral infarction(ACI)after thrombolysis with recombinant tissue plasminogen activator(rt-PA).Methods The clinical data of ACI patients aged≥80 years who were treated with rt-PA thrombolysis from September 1,2018 to January 1,2022 were retrospectively analyzed.According to the modified Rankin scale(mRS)score at 6 months after discharge,the patients were divided into a good prognosis group(mRS score 0-2 points)and a poor prognosis group(mRS score≥3 points).Multivariate Logistic regression analysis was used to analyze the risk factors of poor short-term prognosis in elderly patients with ACI underoging rt-PA thrombolysis.The nomogram model of the analyzed risk factors was constructed using R language,and the receiver operating characteristic(ROC)curve,Booststip method and clinical decision curve were used to verify the accuracy of the model and the benefit of clinical decision-making.Results Ultimately,234 cases in the good prognosis group and 108 cases in the poor prognosis group were included.Multivariate Logistic regression analysis showed that national institutes of health stroke scale(NIHSS)score at admission,location of cerebral infarction,systolic blood pressure(SBP)at admission,time from onset to thrombolysis,post-thrombotic cerebral hemorrhage,history of atrial fibrillation,and high blood glucose at admission were risk factors for poor short-term prognosis in elderly patients with ACI undergoing rt-PA thrombolysis(P<0.01).The ROC curve of validation of the nomogram model showed that the constructed nomogram model had higher accuracy in predicting the short-term prognosis of elderly patients with ACI undergoing rt-PA thrombolysis,in which the area under the ROC curve was 0.854(95%CI:0.71,0.92).The Booststip method was used for repeated sampling for 1000 times to validate the nomogram and found that the mean absolute error of the calibration curve was 0.017.The clinical decision curve showed that when the occurrence threshold was 0.06-0.86,the model had the best applicability.Conclusion The nomogram model has high clinical accuracy in predicting the short-term prognosis of elderly patients with ACI undergoing rt-PA thrombolysis.Clinicians can learn from this model for active intervention in high-risk groups,which has a high reference value for clinical decision-making.