Predictive analysis of all-cause mortality of previously untreated pulmonary tuberculosis patients complicated by hypertension.

Objective: To investigate the risk factors for all-cause mortality of previously untreated pulmonary tuberculosis patients complicated by hypertension and construct a predictive model.

Methods: We retrospectively analyzed the clinical data of inpatients with previously untreated pulmonary tuberculosis complicated by hypertension from 2019 to 2021 in Changsha Central Hospital. Patients' survival status and cardiovascular events were collected through telephone follow-up. LASSO regression was utilized to screen predictive variables, and binary logistic regression identified mortality risk factors. A predictive nomogram model was developed using R software, and its precision and reliability were verified.

Results: Among the 1,014 patients, there were 100 (9.86%) deaths and 82 (8.09%) cardiovascular events. LASSO regression screened out 13 predictive variables. Multivariate logistic regression analysis revealed that smoking history, sputum bacteriology, pleural effusion, coronary heart disease, and chronic kidney disease were independent risk factors. Based on the training set data, a nomogram prognostic model was developed, showing an AUC of 0.712 (95% CI: 0.777-0.847), with 50.0% sensitivity and 84.3% specificity. The model's fit was confirmed through internal and external validations.

Conclusion: The prediction model constructed in this study has high predictive ability and satisfactory clinical efficacy, and can provide an effective individualized prediction tool for assessing all-cause mortality risk in patients with previously untreated pulmonary tuberculosis complicated by hypertension.