Construction and Validation of a Predictive Model for Poor Control of Hypertension in Community Patients

Sun Hongye; Tian Zhifeng

doi:10.3969/j.issn.1004-5775.2025.03.002

Abstract：Objective:To investigate the risk factors associated with poor hypertension control among community patients and to construct a Nomogram predictive model for assessing the risk of poor hypertension control.Methods:832 patients with hypertension were divided into well-controlled group and poorly controlled group according to their blood pressure control.Basic data,lifestyle and laboratory indicators of the two groups were investigated.With poor blood pressure control as the dependent variable,Lasso regression and Logistic regression were used to screen independent risk factors for poor blood pressure control in communities.Nomogram prediction model was constructed with R language,and the constructed model was evaluated.Results:Among the 832 hypertensive residents,399 cases(47.96％)had poor blood pressure control.Logistic regression screened out 6 risk factors for poor blood pressure control in the community,including BMI＞24 kg/m2,smoking,drinking,insufficient physical activity,static behavior time＞5 h and FBG.Based on these factors,a Nomogra prediction model was established,and the evaluation of the NOMoGRA prediction model showed that the C-index value was 0.865,indicating good differentiation.The AUC calculated by ROC curve was 0.822(95％CI:0.792～0.845),which indicates that the model was accurate.The calibration curve was basically close to the diagonal dotted line with slope=1,which proved that the prediction ability of the model was reasonable.Conclusion:The Nomogram predictive model constructed in this study for poor hypertension control among community patients shows good discrimination and accuracy.