Predictive Value Of The ABCD~2Score In Chinese: A Meta-analysis

Predicting the risk of cerebral infarction after TIA has the important clinicalsignificance.The ABCD~2rule is a clinical prediction score system (clinical prediction rule,CPR), which is mainly used to predict the risk of acute cerebral infarction after TIA.Theinternational community has a number of countries applying it to the disease prevention andthe treatment guidelines. The derivation and the development of the model mainly camefrom Europe and the United States population, so the incidence, risk factors, geneticbackground, and medical conditions are different from Chinese population. In recent years anumber of validation studies have evaluate its effectiveness, but there is no systematicreview to evaluate whether it suited to Chinese population.In this study, we use the method of meta-analysis to determine the predictive value ofthe ABCD~2at2and7days across three strata of risk in Chinese population. We retrieval thestudies which have vadiated the prodictive value of the ABCD~2score in Chinese populationafter TIA. Review Manager5software is used to perform the meta-analysis and presents asthe risk ratios（RRs） with95%confidence intervals (CIs).We compared the nember ofstroke predicted across the three strata with the observed nember of stroke in each of thesubsequent validation studies. At last we proved that the prodictive value of the ABCD~2inChinese(A RR score of1represents accurate prediction by the ABCD~2rule,<1representsunder-prediction and>1over-prediction).We included in a total of22studies.8studies report2-day stroke risk (n=1214).TheABCD~2rule predicts occurrence of stroke at2days across all three risk strata: low[RR=0.43,95%CI（0.17–1.10）, I~2=0%], moderate [RR=0.42,95%CI (0.26–0.67), I~2=0%]and high [RR=0.32,95%CI (0.21–0.48), I~2=0%].21studies report7-day stroke risk(n=3823),low[RR=0.25,95%CI (0.16–0.39), I~2=0%], moderate[RR=0.27,95%CI (0.22–0.34), I~2=0%]，and high[RR=0.22,95%CI (0.18–0.27), I~2=28%]. A chi-squared test fortrend confirmed that there is an increasing trend in risk of stroke across the three ABCD~2riskcategories: within two days （χ~2=106.52,P<0.001）,seven days（χ~2=570.34, P<0.001）.At Last we get the following conclusions by the results: The predictive value of theABCD~2score in Chinese population within2days after TIA is slightly higher than thepatients within7days, but both of them under-predict the risk; As the score increasing across the three strata,the risk of the stroke within2days and7days is higher;The ABCD~2scorecan be used to predict short-term risk of stroke after TIA across the three strata in Chinesepopulation, but comparing with its derived model in Europe and the United States it mayunder-predict the risk in Chinese population. The predictive value of this model in Chineseneed further evidence based assessment.