应用logistic回归模型及重复测量方差分析法探讨缺血性中风证候要素与近期预后的关系

目的：探讨缺血性中风病证候要素与近期预后的关系。方法：动态采集缺血性中风病患者发病第1、7和14天的“缺血性中风证候要素诊断量表”（内风、内火、痰湿、血瘀、气虚、阴虚）评分及美国国立卫生研究院卒中量表（National Institutes of Health Stroke Scale，NIHSS）评分。以发病第14天的NIHSS评分作为缺血性中风病患者近期预后的评价指标，并依据评分进行分组（发病第14天NIHSS评分≥7作为近期预后不良组，NIHSS评分〈7分为预后良好组），应用logistic回归模型探讨缺血性中风病急性期证候要素与近期预后的关系并利用受试者操作特征（receiver operating characteristic，ROC）曲线来评价模型的预测能力。根据logistic回归模型得到的结果，进一步应用两样本重复测量数据方差分析比较血瘀证组与非血瘀证组发病第1、7和14天的NIHSS评分。结果：发病第14天的血瘀证（比值比为2．924，95％置信区间为1．231～6．946，P=0．015）和发病第1天的神经功能缺损程度（比值比为1．956，95％置信区间为1．701～2．250，P=0．000）是预测缺血性中风病近期预后的独立危险因素。Logistic回归模型的ROC曲线下面积为0．95。从重复测量均数变化趋势图可看出，血瘀证患者的神经功能缺损程度较非血瘀证患者重，且其NIHSS分值在疾病过程中持续高于非血瘀证组。结论：血瘀证为缺血性中风近期预后的独立危险因素，临床上积极采取活血化瘀治疗将有助于改善疾病的近期预后。本研究还为实施缺血性中风病急性期中西医结合综合治疗方案动态证候干预提供了循证依据及数据支持。

Study on the correlation between traditional Chinese medicine syndrome and short-term prognosis of ischemic stroke using logistic regression model and repeated-measures analysis of variance

OBJECTIVE： To investigate the correlation between traditional Chinese medicine （TCM） syndrome and shortterm prognosis of ischemic stroke. METHODS： TCM syndrome factors and the neurological deficit degree of 464 patients with ischemic stroke were assessed using the Ischemic Stroke TCM Syndrome Factor Diagnostic Scale （ISTSFDS） and National Institutes of Health Stroke Scale （NIHSS） on the 1st, 7th and 14thday from the onset of ischemic stroke. Patients were assigned to the favorable short-term prognosis group and the unfavorable short-term prognosis group, depending on the NIHSS score on the 14th day after onset of stroke. The correlation between TCM syndrome factors and the short-term prognosis of ischemic stroke was studied using the logistic regression model. Then, the receiver operating characteristic （ROC） curves were used to evaluate the predicting capacity of logistic regression mode. Lastly, according to the results of the logistic regression model, the authors investigated the correlation between blood stasis syndrome and the neurological deficit degree of stroke with application of repeated-measures analysis of variance. RESULTS： Blood stasis syndrome （odds ratio=2. 924, 95% confidence interval from 1. 231 to 6. 946, P= 0.015） on the 14th day and NlSSS score （odds ratio= 1. 956, 95% confidence interval from 1. 701 to 2. 250, P=0.000） on the 1st day after onset of stroke were risk factors that could predict short-term prognosis of ischemic stroke. The area under the ROC curves of the logistic regression model was 0.95. There was a tendency for stroke patients with blood stasis syndrome to have higher NIHSS scores than patients without blood stasis syndrome, and there was a decreased NIHSS score with time points delay. CONCLUSION： Blood stasis syndrome is a risk whose diagnosis could predict short-term prognosis of ischemic stroke. Clinically, the application of treatment focusing on activating blood and resolving stasis can improve the short-term prognosis of stroke patients. This study provides an evidence base for dynamic intervention of a comprehensive integrative medical treatment program based on syndrome differentiation for ischemic stroke.