Abstract:

Objective: To compare the application value of the tuberculin skin test (TST) and interferon-gamma release assays (IGRA) in screening for latent tuberculosis infection (LTBI) among close contacts of tuberculosis patients in school, explore influencing factors, and evaluate the performance of the corresponding prediction models. Methods: A retrospective analysis was conducted. Data of 6893 close contacts (including staff and students) from 58 schools and childcare institutions in Haizhu District, Guangzhou, who participated in tuberculosis screening between January 2022 and December 2024, were collected from the “Tuberculosis Information Management System,” a subsystem of the “China Information System for Disease Control and Prevention”. The data included general information (demographics, exposure characteristics, school type, student or teacher) and screening data (standardized symptom inquiry, infection examination method, chest X-ray result, index case identity, and etiological results). Among them, 2714 individuals underwent TST, and 4179 underwent IGRA. Univariate and multivariate logistic regression models were used to analyze factors associated with positive LTBI screening results. The discriminatory performance of prediction models based on the two test results was evaluated using the area under the receiver operating characteristic curve (AUC). Results: The crude positive rate in the TST group was significantly higher than that in the IGRA group (22.62% (614/2714) vs. 8.33% (348/4179), χ²=280.041, P<0.001). After adjusting for confounders, the positive rates for IGRA and TST were 7.04% and 21.74%, respectively. The likelihood of a positive result with IGRA was only 29.20% of that with TST (aOR=0.292, 95%CI: 0.250-0.341, P<0.001). Multivariate logistic regression analysis revealed that for TST, female gender and increasing age were risk factors (aOR=1.549, 95%CI: 1.277-1.879; aOR=1.037, 95%CI: 1.015-1.061). For IGRA, increasing age was associated with increased LTBI positivity (aOR=1.094, 95%CI: 1.067-1.122), while having an index case being a teacher or other staff member, or the contact being a staff member, were protective factors (aOR=0.585, 95%CI: 0.415-0.826; aOR=0.305, 95%CI: 0.106-0.880; aOR=0.244, 95%CI: 0.115-0.520). The prediction model based on IGRA results showed moderate discriminatory power (AUC=0.716), while the model based on TST results showed lower discriminatory performance (AUC=0.641). Conclusion: In populations with widespread BCG vaccination, TST may overestimate the LTBI burden. IGRA demonstrates higher specificity, and its positive results show a more consistent association with known epidemiological risk factors. The model based on IGRA shows better discriminatory ability. It is recommended that in practical screening, IGRA or a “TST initial screening plus IGRA confirmation” strategy be prioritized to to improve screening efficiency and accurately identify targets for intervention.

Key words: Tuberculosis, Schools, Tuberculin test, Interferons, Contact tracing, Factor analysis, statistical