摘要： 目的 探讨入院时美国国立卫生研究院卒中量表（NIHSS）评分联合血清超敏C-反应蛋白（hs-CRP）对缺血性卒中预后的预测价值。方法 纳入2018年5月至2019年12月河北省唐山市工人医院收治的487例急性缺血性卒中患者，收集性别、年龄、高血压、糖尿病、冠心病、既往脑卒中病史、吸烟史和饮酒史等临床资料，测定血清总胆固醇、甘油三酯、低密度脂蛋白胆固醇、hs-CRP、同型半胱氨酸（Hcy）和空腹血糖（FBG），入院时采用NIHSS量表评估神经功能缺损程度，出院时采用改良Rankin量表（mRS）评估预后。单因素和多因素前进法Logistic回归分析筛查缺血性卒中患者预后不良危险因素，并绘制受试者工作特征（ROC）曲线，比较入院时NIHSS评分与血清hs-CRP单独或者联合应用对预后的预测价值。结果 根据mRS评分分为预后良好组（≤ 2分）和预后不良组（> 2分），预后不良组男性（χ²=4.729，P=0.030）、既往脑卒中病史（χ²=13.665，P=0.000）和饮酒史（χ²=8.326，P=0.004）比例，入院时NIHSS评分（Z=-9.316，P=0.000），血清FBG（Z=-3.234，P=0.001）、Hcy （Z=-3.652，P=0.000）和hs-CRP（Z=-8.780，P=0.000）水平均高于预后良好组。Logistic回归分析显示，入院时NIHSS评分增加（OR=1.316，95% CI：1.223~1.417；P=0.000）和血清hs-CRP升高（OR=1.045，95% CI：1.032~1.058；P=0.000）是缺血性卒中患者预后不良的危险因素。ROC曲线显示，入院时NIHSS评分、血清hs-CRP和二者联合的曲线下面积分别为0.818（95% CI：0.781~0.852，P=0.000）、0.806（95% CI：0.768~0.840，P=0.000）和0.903（95% CI：0.873~0.928，P=0.000），并且二者联合应用预测预后不良的价值优于单独入院时NIHSS评分（Z=3.187，P=0.001）和血清hs-CRP（Z=4.418，P=0.000）。结论 入院时NIHSS评分增加和血清hs-CRP水平升高是缺血性卒中患者预后不良的危险因素，二者联合应用可为临床预测缺血性卒中预后不良提供有效参考。

关键词: 卒中, 脑缺血, C反应蛋白质, 预后, Logistic模型

Abstract: Objective To investigate the predictive value of the National Institutes of Health Stroke Scale (NIHSS) score combined with serum high-sensitivity C-reactive protein (hs-CRP) on admission to ischemic stroke. Methods A total of 487 patients with acute ischemic stroke admitted to the Department of Neurology of Tangshan Gongren Hospital from May 2018 to December 2019 were selected. The clinical data of patients were collected such as gender, age, hypertension, diabetes, coronary heart disease, previous stroke history, smoking and drinking history, serum low-density lipoprotein cholesterol (LDL-C), total cholesterol (TC), triglyceride (TG), hs-CRP, homocysteine (Hcy), fasting blood glucose (FBG). The NIHSS scale was used to evaluate the degree of neurological deficit at admission, and the modified Rankin Scale (mRS) was used to evaluate the prognosis when the patient was discharged. The mRS score ≤ 2 was defined as good prognosis, and the mRS score > 2 was defined as poor prognosis. Univariate and multivariate forward Logistic regression analysis were used to screen risk factors for poor prognosis of ischemic stroke, and the receiver operating characteristic (ROC) curve was used to compare the predictive value of NIHSS and hs-CRP levels at admission alone or in combination for patient prognosis. Results Compared with the good prognosis group, the poor prognosis group had higher proportions of male patients (χ²=4.729, P=0.030), previous history of stroke (χ²=13.665, P=0.000), and drinking alcohol (χ²=8.326, P=0.004). The NIHSS score at admission (Z=-9.316, P=0.000), serum FBG (Z=-3.234, P=0.001), Hcy (Z=-3.652, P=0.000) and hs-CRP (Z=-8.780, P=0.000) level in poor prognosis group were higher than those in good prognosis group. Logistic regression analysis showed that the NIHSS score increased at admission (OR=1.316, 95%CI:1.223-1.417; P=0.000), and the serum hs-CRP level increased (OR=1.045, 95%CI:1.032-1.058; P=0.000) were risk factors for the prognosis of ischemic stroke. The area under the curve (AUC) predicted by the two combined (AUC=0.903; 95%CI:0.873-0.928, P=0.000) exceeded than that of the NIHSS score at admission alone (AUC=0.818; 95%CI:0.781-0.852, P=0.000) and the serum hs-CRP alone (AUC=0.806; 95%CI:0.768-0.840, P=0.000); pairwise comparison found that the combination of the two was better than the NIHSS score at admission alone (Z=3.187, P=0.001) and serum hs-CRP alone (Z=4.418, P=0.000). Conclusions NIHSS score and high level of serum hs-CRP at admission are risk factors for poor prognosis of ischemic stroke. Combining NIHSS score and serum hs-CRP can provide an effective reference for clinical prediction of poor prognosis of ischemic stroke.

Key words: Stroke, Brain ischemia, C-reactive protein, Prognosis, Logistic models