ABCD~3模型预测短暂性脑缺血发作后早期卒中风险

发布时间：2018-04-18 02:14

本文选题：脑缺血发作 + 短暂性　；参考：《郑州大学》2014年硕士论文

【摘要】：背景和目的短暂性脑缺血发作是重要的脑卒中危险因素，尤其是在短暂性脑缺血发作后的早期阶段。2010年Merwick等提出ABCD3模型预测早期卒中发生，该模型在西方人群中具有较好的预测作用，然而在中国人群中的应用价值尚需前瞻性研究进行验证。本研究拟通过前瞻性纳入连续入住郑州大学第一附属医院神经内科的短暂性脑缺血发作患者，验证此模型在中国人群中的应用价值。方法前瞻性连续纳入2010年10月至2012年9月郑州大学第一附属医院神经内科住院TIA患者。TIA采用WHO传统定义，即症状持续时间㩳24小时，不遗留神经功能缺损。所有患者均由统一培训的两名或两名以上神经内科医师独立诊断为TIA，进行评分并完成CRF表登记，记录所有纳入患者的人口学特征及临床资料。并完成TIA发作后第2、7、28及90天的随访，记录在随访期内是否发生脑卒中或TIA的复发。使用SPSS16.0对资料进行统计分析。定量资料用均数±标准差(x±s)表示,定性资料采用χ2检验。采用Logistic回归，分析影响TIA后早期卒中发生的危险因素，绘制ABCD2及ABCD3的ROC曲线，计算曲线下面积，并对其曲线下面积进行比较。先进行单因素回归分析(P0.1)，筛选出有统计学意义的变量，再用前进法进行多因素回归分析（P㩳0.05）。结果 1.共纳入符合条件TIA患者320例,其中6例失访。年龄18-86岁，平均年龄为56.97±13.399岁，其中女性129（40.3%）例，男性191（59.7%）例。90d后46例发生卒中，其中男性22例，女性24例，无死亡患者，无出血性卒中的发生。 2.随着评分的增加，发生卒中的风险逐渐增加，2天、7天、28天及90天发生卒中的风险分别为：5.6%(3.1%-8.1%)，8.8%(5.7%-11.9%)，10.9%(7.5%-14.3%)，14.4%(10.6%-18.3%)，趋势性检验结果显示卒中发生率与卒中是否发生有统计学意义，随着评分的增加，卒中发生风险越大。 3.应用预先设定敏感度法，得出ABCD3最佳截断点为5分；2天，7天，28天，90天ABCD3模型ROC曲线下面积分别为：AUROCC=0.757（95%CI0.706-0.786），AUROCC=0.722（95%CI0.669-0.770），AUROCC=0.738（95%CI0.686-0.786），AUROCC=0.717（95%CI0.663-0.766）。2天，28天，90天ROC曲线下面积与ABCD2比较，差异均有统计学意义，而7天时差异无统计学意义。 4.Logistic回归结果，2天时多因素回归分析结果发现：ABCD3评分（OR=4.726，95%CI1.177-18.976，P=0.029）有统计学意义；7天时多因素分析示：ABCD3评分（OR=2.620，95%CI1.107-6.201，P=0.028）及脑梗死病史(OR=3.995，95%CI1.465-10.896，P=0.007)有统计学意义，与7天内卒中发生相关（P0.05）；28天时多因素分析结果：ABCD3评分（OR=3.017，95%CI1.280-7.111，P=0.012）,脑梗死病史(OR=5.288，95%CI1.823-15.339，P=0.002)及7天内TIA发作超过2次（OR=3.358，，95%CI1.132-9.958，P=0.029）有统计学意义；90天时分析结果示：ABCD3评分（OR=2.638，95%CI1.243-5.598，P=0.012）,脑梗死病史(OR=2.808，95%CI1.100-7.172，P=0.031)，7天内TIA发作超过2次（OR=3.050，95%CI1.146-8.116，P=0.026），是各时间段进展至脑梗死的危险因素。结论 ABCD3评分模型对住院的中国TIA人群具有良好的预测价值，较ABCD2能更好的预测早期卒中发生，为TIA后卒中预防提供良好的预测工具。
[Abstract]:Background and purpose
Transient ischemic attack is the most important risk factor for stroke, especially in transient ischemic attack in early stage after.2010 Merwick raised the ABCD3 model to predict early stroke, the model has a good prediction of the population in the west, however, the application value in the crowd China prospective studies are needed to verify the purpose of this study was to prospectively. Patients in the First Affiliated Hospital of Zhengzhou University in the continuous transient ischemic attack, to verify the application value of this model in the Chinese population.
Method
Prospective consecutive First Affiliated Hospital of Zhengzhou University from October 2010 to September 2012 in the Department of neurology patients with TIA.TIA WHO using the traditional definition, namely the duration of symptoms? 24 hours, no neurological deficit. Two or more than two neurologists and all patients were trained independently diagnosed as TIA score and complete the CRF registration form that documented clinical data and demographic characteristics of patients. And complete the TIA after the onset of 2,7,28 and 90 days of follow-up, relapse records the occurrence of cerebral stroke or TIA during the follow-up period.
Use SPSS16.0 for statistical analysis. The mean and standard deviation for quantitative data (x + s) said that the qualitative data using 2 test. Using Logistic regression analysis of risk factors for early stroke effect of TIA, ABCD2 and ROC curves of ABCD3, calculate the area under the curve, and the area under the curve were compared. The single factor regression analysis (P0.1), screened statistically significant variables, then forward multiple regression analyses (P? 0.05).
Result
1. a total of 320 eligible TIA patients were enrolled. 6 of them were lost. The average age was 56.97 + 13.399 years, including 129 (40.3%) cases, male 191 (59.7%), and.90d after 46 cases of stroke, including male 46 cases, female cases, no deaths and no hemorrhagic stroke.
2. with the score increased the risk of stroke increased gradually, 2 days, 7 days, 28 days and 90 day risk of stroke was 5.6% (3.1%-8.1%), 8.8% (5.7%-11.9%), 10.9% (7.5%-14.3%), 14.4% (10.6%-18.3%), the trend of the test results showed that the incidence of stroke and stroke occurrence statistical significance, with the score increases, the greater the risk of stroke.
3. application of preset sensitivity method, obtains the cut-off point of ABCD3 was 5; 2 days, 7 days, 28 days, 90 days ABCD3 model area under the ROC curve were: AUROCC=0.757 (95%CI0.706-0.786), AUROCC=0.722 (95%CI0.669-0.770), AUROCC=0.738 (95%CI0.686-0.786), AUROCC=0.717 (95%CI0.663-0.766).2 days, 28 days, compared the area with ABCD2 for 90 days under the ROC curve, the differences were statistically significant, but on the 7 day when the difference was not statistically significant.
The results of 4.Logistic regression, day 2 multivariate regression analysis showed that: the ABCD3 score (OR=4.726,95%CI1.177-18.976, P=0.029) were statistically significant; 7 days when multivariate analysis showed: ABCD3 score (OR=2.620,95%CI1.107-6.201, P=0.028) and cerebral infarction (OR=3.995,95% CI1.465-10.896, P=0.007) had statistical significance, associated with stroke occurred within 7 days (P0.05); multiple factor analysis for 28 days: ABCD3 score (OR=3.017,95%CI1.280-7.111, P=0.012), cerebral infarction (OR=5.288,95%CI1.823-15.339, P=0.002) and within 7 days of the onset of TIA more than 2 times (OR=3.358,95%CI1.132-9.958, P=0.029) had statistical significance; the 90 day analysis results showed: ABCD3 score (OR=2.638,95%CI1.243-5.598, P=0.012), cerebral infarction (OR=2.808,95%CI1.100-7.172, P=0.031). Within 7 days of the onset of TIA more than 2 times (OR=3.050,95% CI1.146-8.116, P=0.026), is the time to progress to the brain A risk factor for infarction.
conclusion
The ABCD3 scoring model has a good predictive value for the hospitalized Chinese TIA population. Compared with ABCD2, it can better predict the occurrence of early stroke, and provide a good prediction tool for stroke prevention after TIA.

【学位授予单位】：郑州大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：R743.3

【参考文献】