Base On The Meta-analysis Of Risk Of Stroke In Chinese Hypertensive Patients By Risk Appraisal Model

Objective:The combined risk values of risk factors in Hypertensive patients complicated with stroke were obtained by meta-analysis,and the health risk assessment model was established to predict the risk of stroke in the hypertensive population.Then the high risk population and low risk population were differentiated and performed personalized health education and health promotion for different risk groups,helps to reduce the risk of stroke in people with hypertension.Methods:This study was divided into three parts.First,retrospective analysis of literature about stroke risk factors in hypertension patients,the RR value and the average exposure level(Pi)of risk factors in hypertensive patients complicated with stroke were obtained by meta-analysis.Second,Based on the coefficients obtained in the first step,the Rothman-Keller model and Logistic regression model was constructed and the risk of stroke in individuals with hypertension was predicted.Then,a group of random data sets is generated through binomial distribution function method,determining the dividing value of risk assessment grading,and to draw a diagram of risk evaluation.Third,using the actual morbidity population data and prediction results of Rothman-Keller model and Logistic regression model,respectively,drew the ROC curve for verifying.Results:(1)The result of meta-analysis:Thirty-one studies are included in total.In this meta-analysis,fifteen risk factors of Chinese hypertensive patients complicated with stroke were included,including systolic blood pressure,diastolic blood pressure,low-density lipoprotein cholesterol,high-density lipoprotein cholesterol,homocysteine,total cholesterol,triglyceride,obesity,hyperuricemia,impaired fasting glucose/diabetes,family history of stroke,carotid plaque,regular drug taking,physical exercise,and smoke,and their pooled RR were1.06,1.02,1.69,0.75,3.05,1.74,1.21,1.18,5.63,1.94,1.94,3.81,2,0.42,1.81,respectively.(2)The determination of the model parameters and the risk assessment levelAccording to the trend of the risk of joint distribution map,the risk of disease was 0.0653,0.0989 as the Rothman-Keller model cutoff value to respectively predict the risk of low middle and high risk of stroke in the hypertensive population.Logistic regression mode:PLogit)((28)-5.152+0.0581 x +0.0202x-0.2883 x +0.5254 x +1.1155 x +1.7286x+0.6637 x +0.6638 x +1.3389 x +0.69310x-0.86811 x +0.59312 x.According to the trend of the probability of joint distribution map,the risk of disease was 0.0661,0.1570 as the Logistic regression model cutoff value to respectively predict the risk of low middle and high risk of stroke in the hypertensive population.(3)Actual data validation:The area under the ROC curve of actual population incidence and the forecasted results of the Rothman-Keller model and 95%CI is 0.784(0.718,0.849),P=0.000.The area under the ROC curve of actual population incidence and the forecasted results of the Logistic regression model and 95%CI is 0.725(0.651,0.799),P=0.000.Conclusion:In this study,the risk assessment model of Chinese hypertensive patients with stroke was established by the method of meta-analysis quantitative synthesis,combined Rothman-Keller model and Logistic regression model.The Rothman-Keller model may be superior to the Logistic regression model.The results of this study can be used to differentiate the different risk levels of stroke in the hypertensive population and to provide the basis for the individual health education and health promotion,and performed personalized health education and health promotion for different risk groups.And also provide basis for intervention effect of the Health education and health promotion.To provide reference basis for the prediction of stroke risk in hypertensive population in China,but also provides a new ideas to establish the risk prediction model of other chronic diseases.