Comparison Of Difefrent Scores In Predictive Value Of Cerebral Infarction Developed By Transient Ischemic Attack In Chinese Population

Comparison of different risk scores in predictive value of cerebralinfarction developed by transient ischemic attack in Chinese PopulationTransient ischemic attack (TIA) is considered to be the warning signal ofcerebral infarction, TIA patients with a high risk of early onset of cerebral infarction,the risk assessment can be seen on the TIA after cerebral infarction predictivemodels for clinical work is essential. There are a number of countries in theinternational clinical prediction rule (CPR) included in cerebrovascular diseaseprevention and treatment guidelines. The development of the predictive instrumentsmainly came from Europe and the United States population, so the races, risk factors,medical conditions, the incidence of severity, genetic background, are different fromChinese population. A large number of related validation studies have evaluatedthem effectiveness in recent years. However, there is no systemetic comparation toevaluate the predictive value of the four scoring models whether they suited toChinese population.In this study, a retrospective cohort analysis method, draw the ROC curve,observed area under the curve,completed four kinds of models to predict thereliability of TIA after cerebral infarction incidence analysis,calculate the relativerisk (RRs) and95%confidence intervals (CIs).Compare different risk stratificationmodel to predict cerebral infarction rate and the actual incidence;Relative riskcalculated as an indicator, the predictive value of the model of the Chinesepopulation (RR <1: indicate lack of risk assessment, RR=1: indicate accurate riskprediction, RR>1: indicate risk assessment the excessive). All data is analyzed byusing SPSS17.0statistical software. Balancing test using the chi-square test, ROCarea under the curve estimated using non-parametric tests (Mann-Whitney method).In this study,260patients with cervical artery TIA patients were studiedretrospectively25cases were lost to follow, and finally included235patients withTIA. Observed7days progress to cerebral infarction,22cases (9.36%),1year progressed to cerebral infarction,48cases (20.43%).7days of the ABCD score AUC0.70[95%CI (0.60-0.81), P <0.05]; the The ABCD2score AUC of0.74[95%CI(0.64-0.84), P <0.05]; SPI-II score AUC was0.67[95%CI (0.56-0.80), P <0.05];The ESSEN rated AUC was0.62[95%CI (0.50-0.74, P>0.05)].1year ABCD scoreAUC was0.62[95%CI (0.54-0.71)]; the ABCD2score AUC was0.62[95%CI(0.54-0.71)]; SPI-II score AUC of0.64[95%CI (0.56-0.73),]; ESSEN score AUC0.65[95%CI (0.57-0.74)], four prediction score P value less than0.05. Comparedwith the original model, TIA ABCD score within7days after the low-risk group [RR=0.09,95%CI (0.01-0.67)], the risk group [RR=0.92,95%CI (0.30-3.37)],high-risk group [RR=0.72,95%CI (0.15-3.54)].7days after TIA ABCD2scorelow-risk group [RR=0.49,95%CI (0.04-5.56)], in the risk group [RR=0.59,95%CI (0.24-1.48)], high-risk group [RR=0.65,95%CI (0.18-2.37)]. TIA after1yearSPI-â…¡ score low risk group [RR=0.58,95%CI (0.18-1.21)], in the risk group [RR=0.87,95%CI (0.42-1.80)], the high-risk group [RR=0.55,95%CI (0.16-1.92)].1year after TIA ESSEN rated low risk group [RR=0.11,95%CI (0.03-0.38)], the riskgroup [RR=0.18,95%CI (0.07-0.57)], high-risk group [RR=0.55,95%CI(0.00-2.73)].Finally we obtain conclusions by the results as follows: ABCD, ABCD2,SPI-â…¡ andESSEN score models can be used to predict the risk of cerebralinfarction in the Chinese population after TIA. ABCD2score is more suitable forshort-term risk prediction, ESSEN score is more suitable for long-term riskprediction.But they may under-predict the risk in Chinese population comparing withtheir derived models in Europe and the Uinted States population. The predictivevalue of these models in Chinese population need further evidence based assessment.