ABCD2评分对短暂性脑缺血后的中期脑梗死事件和死亡的预测能力

目的探讨ABCI2评分对中国人群短暂性脑缺血(TIA)后的中期脑梗死事件及死亡的预测能力，以便利用其识别高危TIA患者。 方法采用前瞻性设计，以广州医学院第二附属医院神经内科自2008年1月开始收治的TIA患者为研究对象，共收集到179例病例资料完整的患者。对就诊的TIA患者行ABCD2评分，追踪记录其TIA发作后18月内是否发生脑梗死事件，并根据是否发生脑梗死分为无脑梗死组和脑梗死组。以病程第18月为终点事件观察时间点，统计各种疾病导致的死亡情况，并根据是否临床死亡分为生存组和死亡组。采用统计学方法分析组间年龄、血压、临床特征、症状持续时间和糖尿病史等各指标的差异性，绘制ROC曲线。 结果179例TIA患者中无脑梗死组127例，脑梗死组52例。脑梗死组就诊时血压≥140/90 mm Hg(86.5％)、肢体无力(42.3％)、症状持续时间≥60 min(55.8％)、症状持续时间10～59 min(40.4％)和患糖尿病(80.8％)比例较无脑梗死组均明显增多，差异均有统计学意义(P＜0.05)。脑梗死组ABCD2评分比无脑梗死组高，差异有统计学意义(P＜0.05)。利用ROC曲线，以ABCD2评分来预测患者患脑梗死情况，其曲线下面积为0.874，P=0.000，95％可信区间为0.817～0.931,Cutoff值为4.5分。179例TIA患者中生存组144例，死亡组35例。死亡组ABCI2评分与生存组比较差异无统计学意义(P＞0.0.5)。利用ROC曲线，以ABCD2评分来预测患者死亡情况，其曲线下面积为0.492,P=0.889,95％可信区间为0.389～0.596。 结论ABCD2评分系统在判断TIA后中期脑梗死风险方面有一定预测价值，对TIA后中期死亡危险度无预测作用。

Predictive value of ABCD2 scale in cerebral infarction and death events after transient ischemic attack: a cohort medium-term follow-up study

Objective To evaluate the predictive value of ABCD2 scoring system in cerebral infarction and death events in the medium-term after transient ischemic attack (TIA) of Chinese population. Methods One hundred and seventy-nine patients with TIA having complete clinical data,admitted to our hospital from January 2008, were chosen in our perspective study. The ABCD2 scale was applied to all the patients and used to observe the 18-month cerebral infarction risk events; according to these data, patients after TIA were divided into cerebral infarction group and non-cerebral infarction group; To the end of 18 months, according to the death event, the patients were divided into survival group and death group. The age, level of blood pressure, clinical features, duration of symptoms and history of diabetes were collected and compared between each 2 groups; the cutofflevel of ABCI2 scale for anticipating cerebral infarction risk and mortality was evaluated according to receiver operating characteristic (ROC) curve. Results Among the studied 179 patients, 52 patients appeared cerebral infarctions within 18 months of TIA; patients in the cerebral infarction group had significantly higher percentage of patients with blood pressure ≥ 140/90 mm Hg (86.5％), unilateral weakness (42.3％),duration of symptoms ≥60 minutes (55.8％), or duration of symptoms 10 to 59 minutes (40.4％), and diabetes (80.8％) as compared with patients in the non-cerebral infarction group (P＜0.05). According to the ROC curve, area under curve was 0.874 of ABCD2 scale for anticipating cerebral infarction risk, 95％confidence interval (CI) was 0.817-0.931 and the cutoff level of ABCD2 scale for anticipating medium-term cerebral infarction risk was 4.5. Among the studied 179 patients, 35 patients died within 18months of TIA; no significant difference in the scores of ABCD2 scale was noted between survival group and death group (P＞0.05); according to ROC curve, area under curve was 0.492 of ABCD2 scale for anticipating mortality, and 95％ CI was 0.389-0.596. Conclusion ABCD2 scale has predictive value in cerebral infarction event, while not in death event in the medium-term of TIA.