早期改良美国国立卫生研究院卒中量表评分对缺血性卒中预后的预测作用

Predictive value of the early modified National Institutes of Health Stroke Scale for the prognosis of ischemic stroke

目的探索卒中发生早期不同时间段改良美国国立卫生研究院卒中量表（mNIHSS）评分对缺血性卒中后6个月预后的预测价值。方法收集前循环缺血性卒中病例并记录卒中发生后3d内、4—6d、7～9d的mNIHSS评分及卒中后6个月的改良Rankin评分（mRS）。使用Spearman秩相关系数及受试者工作特征（ROC）曲线进行统计学分析。结果共收集161例患者。发病3d内、4—6d、7—9d的mNIHSS评分和卒中后6个月时的mRS评分呈正相关，相关系数-分别为0．592、0．597、0．595，均P〈0．01。发病6个月时曲线下面积（AUC）在卒中后3d内、4～6d、7～9d分别为0．789（95％CI0．708～0．856）、0．792（95％CI0．712—0．857）、0．799（95％CI0．721～0．864）。发病6个月时不同时间段AUC两两比较，3d内和4—6d（z=0．93，P=0．14）、3d内和7—9d（z=0．94，P=0．52）、4—6d和7～9d（z=0．96，P=0．27）差异无统计学意义。预测6个月预后的最佳界值在3d内、4～6d和7—9d分别为8、7和6分。在相应时间段评估mNIHSS时高于最佳界值则预后不良，低于或等于最佳界值则预后良好。不同时间段mNIHSS预测卒中6个月预后的阴性预测值由3d内的0．847渐增至第7—9天的0．867，而阳性预测值由3d内的0．601渐减至第7～9天的0．533。总体的一致率由3d内的0．753渐减至第7—9天的0．709。结论卒中发生9d内mNIHSS评分对卒中6个月预后有良好的预测价值。但在卒中早期的不同时间段使用mNIHSS评分预测6个月预后的最佳界值不同。

Objective To investigate the predictive value of the modified National Institutes of Health Stroke Scale （mNIHSS） , measured within 9 days after stroke, for the outcome in terms of modified Rankin Scale （mRS） at 6 months after stroke. Methods All 161 patients with acute anterior circulation ischemic stroke were recruited consecutively from July 2010 until November 2010. The mNIHSS score was assessed within 3 days, 4 to 6 days and 7 to 9 days after stroke, and the mRS score at 6 months after stroke was assessed by a neurology resident. Spearman rank correlation and ROC curve were used for statistic analysis. Results One hundred and sixty-one patients were assessed within 3 days, at 4 to 6 days and 7 to 9 days after stroke. Significant Spearman rank correlation coefficients were found between mRS at 6 months and mNIHSS scores within 3 days （ rs = 0. 592, P 〈 0.01 ） , 4 to 6 days （ rs = 0. 597, P 〈 0. 01 ） and 7 to 9 days （rs =0. 595, P 〈 0. 01 ）. At 6 months after stroke, the AUC ranged from 0. 789 （95% CI 0. 708- 0. 856） for measurements within 3 days to 0. 792 （95% CI 0. 712-0. 857） and 0. 799 （95% CI 0. 721- 0. 864） for 4 to 6 days and 7 to 9 days, respectively. There was no significant difference for the AUC between them. The optimal cut-off values for the prediction of prognosis at 6 months were 8, 7 and 6 points of mNIHSS score within 3 days, 4 to 6 days and 7 to 9 days, respectively. The negative predictive value gradually increased from 0. 847 for assessment within 3 days to 0. 867 （ 95% CI 0. 578-0. 765 ） for 7 to 9 days, whereas positive predictive value declined from 0. 601 for assessment within 3 days to 0. 533 for 7 to 9 days. The overall accuracy of predictions decreased from 0. 753 for assessment within 3 days to 0. 709 for 7 to 9 days. Conclusions When measured within 9 days, the mNIHSS has a good predictive value for final outcome in terms of mRS at 6 months after stroke. However, the optimal cut-off values for the prediction of prognosis at 6 months are different when early mNIHSS scores were assessed at different periods.