ABSTRACT
Objective@#To establish a hypertension risk assessment model among the middle-aged and elderly populations based on residents' electronic healthcare records of the basic public health service program, so as to provide insights into prevention of hypertension.@*Methods@#Demographic features and physical examinations were collected among residents at ages of 40 years and older from residents' electronic healthcare records of the basic public health service program in a county of Zhejiang Province from 2019 to 2020. The risk factors of hypertension were identified using a multivariable logistic regression model, and the odds ratio (OR) for each risk factor was transformed into approximate relative risk (RR), which was included in the formula for calculation of the disease risk proposed by Harvard School of Public Health to create a hypertension risk assessment model. The predictive value of the model was evaluated using a receiver operator characteristic (ROC) curve.@*Results@#Totally 7 275 subjects were enrolled, with a mean age of (66.15±7.91) years, and the participants included 3 189 males and 4 086 females, with a male-to-female ratio of 0.78∶1. There were 190 cases with new-onset hypertension (2.61%). Multivariable logistic regression analysis revealed that overweight, obesity, central obesity, borderline high triacylglycerol (TG), elevated TG, abnormal fasting plasma glucose (FPG), prehypertension and family history of hypertension were included in the hypertension risk assessment model, with approximate RR values of 1.66, 1.96, 1.54, 1.17, 1.64, 1.45, 1.69 and 1.11. The area under the ROC curve (AUC) of the model was 0.678 (95%CI: 0.641-0.715, P<0.001), and the optimal positive cut-off was 0.899. The model predicted 139 subjects with RR>0.899 for hypertension, with a sensitivity of 73.16% and specificity of 55.79%.@*Conclusions@#The hypertension risk assessment model created in this study is feasible to predict the RR for developing hypertension among the middle-aged and elderly populations, which has a predictive value in healthcare management.