Establishment And Evaluation Of Hypertension Risk Prediction Model For Community Physical Examination Population In Hangzhou Road

Objective: Based on the data of health examination in Hangzhou Road community of Urumqi City,this paper discusses the prevalence of hypertension and its risk factors in Hangzhou Road community residents,and constructs and evaluates the prediction model of hypertension in Hangzhou Road community,it can provide reference for community volunteers to identify and manage the high risk population of hypertension in the early stage.Methods: The subjects were 2021 residents of Hangzhou Road community in Urumqi who underwent physical examinations from January to December in the fifth affiliated hospital of Xinjiang Medical University,the demographic characteristics,behavioral life style,anthropometric indexes and laboratory test data were collected.Univariate logistic regression analysis was used to screen the possible predictors of outcome events,and the factor set of hypertension prediction model was selected by Lasso regression,multivariate logistic regression analysis was used to explore independent predictive factors,and all samples were randomly divided into training set(70%)and validation set(30%)by R language sample function,the model was validated by 1000 bootstrap sampling,and the validation set was used to validate the model.The clinical application value of the model was evaluated by area under ROC curve(AUC),correction curve and decision curve analysis(DCA),quantifies the net gain within the threshold probability range.Results: 1)a total of 2143 residents were selected for the study,including 919 men(42.9%)and 1224 women(57.1%),with an average age of(66.7.13.9)years.2)Lasso regression combined with multivariate logistic regression analysis showed that age,diet habit,diabetes mellitus,waist circumference,BMI and smoking were the risk factors of hypertension.3)The predictive model was constructed according to the above independent risk factors,and the results of 1000 bootstrap sampling were verified.The AUC of the predictive model of hypertension was 0.813(95% CI: 0.719,0.834),the correction curve shows that the predicted values in the prediction model are in good agreement with the observed results.The AUC of the validation set was 0.823(95% CI:0.791,0.855).The decision curve showed that when the threshold probability of hypertension was between 0.01 and 0.98,the net benefit of using the nomogram model to predict the risk of hypertension was better.Conclusion: 1)Compared with the national average level,the prevalence rate of hypertension was higher in the residents who took part in the health examination in Hangzhou Road community of Urumqi City.2)Age,diabetes mellitus,waist circumference,BMI,diet habit and smoking were the risk factors of hypertension among residents participating in health examination in Hangzhou Road community.3)The prediction model of high blood pressure nomogram constructed by this research has been proved to be good in discrimination,calibration and practicability by internal validation,to help the community medical staff to identify the high-risk groups of hypertension early,and to develop targeted interventions for the corresponding population.