Determinants of willingness to accept Tuberculosis Preventive Treatment (TPT) among close contacts with Tuberculosis Infection (TBI) in Xinjiang, China

Tuberculosis preventive treatment (TPT) is a critical public health tool for tuberculosis (TB) control; however, willingness to accept TPT among eligible individuals remains suboptimal in high-burden settings. While demographic correlates of willingness to accept TPT have been identified in prior studies, few have developed clinically applicable integrated prediction frameworks in high-burden Chinese settings. This study aims to identify multidimensional determinants of willingness to accept TPT among close contacts with tuberculosis infection (TBI) in Xinjiang, China, and to construct and internally validate a predictive nomogram for individualized prediction of this willingness and for guiding targeted counseling interventions for frontline healthcare workers. A cross-sectional study was conducted from January 1 to December 31, 2023, enrolling 460 TBI-positive close contacts of 87 persons with bacteriologically confirmed pulmonary TB. Willingness to accept TPT — defined as a self-reported affirmative response to a standardized survey question assessed through face-to-face structured interview — was the primary outcome, representing stated intention rather than actual treatment initiation. Significant variables were screened using least absolute shrinkage and selection operator (LASSO) regression with 10-fold cross-validation. Multivariable logistic regression identified independent predictors. A predictive nomogram was developed and evaluated using the AUC, calibration plots, Brier score, sensitivity, specificity, and decision curve analysis (DCA). All analyses were conducted using R (version 4.3.2; R Core Team, 2023) and the rms package (Harrell, 2023). Of the 460 eligible participants, 372 (80.9%) expressed willingness to accept TPT upon structured interview — the primary study outcome. Multivariable analysis identified seven independent predictors of TPT acceptance: high TB-related knowledge (aOR = 6.00; 95% CI: 3.30–10.91), health education by healthcare workers (aOR = 3.65; 95% CI: 1.90–7.00), history of smoking (aOR = 3.59; 95% CI: 1.43–9.06), alcohol consumption (aOR = 3.10; 95% CI: 1.02–9.41), and cohabitation with a person with TB (aOR = 2.09; 95% CI: 1.14–3.84). Conversely, higher annual household income (aOR = 0.076; 95% CI: 0.030–0.193) and frequent exercise were associated with lower acceptance; small subgroup sizes substantially limit the numerical precision of these estimates. The nomogram demonstrated good discrimination (AUC = 0.861; 95% CI: 0.817–0.902), a Brier score of 0.099, and satisfactory calibration (Hosmer–Lemeshow χ² = 4.41, P = 0.819). TPT acceptance in this high-burden setting is relatively high, yet significantly influenced by socioeconomic factors and TB-related knowledge. TB-related knowledge and professional health education by healthcare workers are the strongest modifiable drivers. The counterintuitive associations of smoking and alcohol history with higher acceptance warrant cautious interpretation, as these may reflect perceived health vulnerability or residual confounding rather than direct causal pathways. The nomogram offers a structured visual prediction aid with good discriminative performance and calibration; external validation in independent Chinese settings is required before clinical implementation. From a public health perspective, prioritizing TB-related knowledge enhancement programs and structured healthcare worker counseling represents the most scalable evidence-informed strategy for improving TPT coverage.