Abstract:

Objective: To analyze the occurrence of catastrophic household cost among patients with rifampicin-susceptible pulmonary tuberculosis in Chongqing and to develop a risk prediction model, so as to provide evidence for identifying and supporting high-risk households during treatment management. Methods: A stratified multistage sampling method was used to select 355 patients with rifampicin-susceptible pulmonary tuberculosis who had received treatment for more than two weeks and were managed by tuberculosis designated hospitals in five counties/districts of Chongqing from January 2023 to December 2024. The participants were randomly divided into a training set and a validation set at a ratio of 7∶3, with 249 cases in the training set and 106 cases in the validation set. LASSO regression was used to screen potential predictors, followed by multivariable logistic regression analysis. A nomogram prediction model was constructed based on the results. The discrimination, calibration, and clinical utility of the model were evaluated using ROC curve, calibration curve, and decision curve analysis. Results: Among the 355 patients with rifampicin-susceptible pulmonary tuberculosis, the incidence of catastrophic household cost was 58.59% (208/355). Multivariable logistic regression analysis showed that hospitalization (OR=5.906, 95%CI: 2.752-13.279), occupational change (OR=12.074, 95%CI: 4.844-33.425), being the main household income earner (OR=2.388, 95%CI: 1.174-4.985), and comorbidity with chronic diseases (OR=3.304, 95%CI: 1.485-7.686) were predictors of catastrophic household cost. AUC of the risk prediction model based on these variables was 0.883 (95%CI: 0.843-0.924) in the training set and 0.867 (95%CI: 0.800-0.935) in the validation set. The calibration curves of both the training and validation sets were generally consistent with the ideal curve. Decision curve analysis showed that the model had a high net benefit when the threshold probability ranged from 0.20 to 0.96. Conclusion: The incidence of catastrophic household cost among patients with rifampicin-susceptible pulmonary tuberculosis in Chongqing was relatively high. The nomogram-based risk prediction model developed in this study showed good predictive performance and may provide a scientific basis for early identification of high-risk households and targeted assistance during tuberculosis treatment management.

Key words: Tuberculosis, pulmonary, Rifampin, Fees, medical, Models, economic, Nomograms