Abstract: Objective: To quantitatively evaluate the effectiveness of the “Active Tuberculosis Screening+Full-Course Hospitalization” model in the detection of tuberculosis patients in Hotan Prefecture of Xinjiang Uygur Autonomous Region, and to provide a scientific basis for the steady advancement of tuberculosis prevention and treatment in Xinjiang. Methods: The reported incidence data of tuberculosis patients in Hotan Prefecture of Xinjiang and other prefectures without full-course hospitalization strategy from January 2012 to December 2021 were collected. Joinpoint regression model was used to analyze the time trend of reported incidence of tuberculosis. Taking July 2018 as the policy intervention time point of the “Active Tuberculosis Screening+Full-Course Hospitalization” model, the interrupted time series (ITS) model and controlled interrupted time series (CITS) model were constructed according to whether the control area was set up or not, to analyze the effect of policy intervention. Results: The highest reported incidence of tuberculosis in Hotan Prefecture from 2012 to 2021 was 465.10/100000 (n=10278) in 2018, and the lowest incidence was 129.40/100000 (n=3241) in 2021, with an overall decreasing trend (AAPC=-4.5%, P<0.05); reported incidence of tuberculosis from 2012 to 2018 showed an increasing trend (APC=10.8%, P<0.05), while reported incidence of tuberculosis from 2018 to 2021 showed a rapid decreasing trend (APC=-29.0%, P<0.05). The ITS model analysis showed that one month after the implementation of the “Active Tuberculosis Screening+Full-Course Hospitalization” model in Hotan Prefecture (July, 2018), the reported incidence increased by 16.859/100000 (P=0.001), and the long-term effect of the implementation of the new model was a decreasing trend in the reported incidence of tuberculosis (β₃=-1.098, P<0.001). The CITS model analysis showed that reported incidence of tuberculosis in Hotan Prefecture increased significantly one month after the implementation of the new model (July, 2018), with an increase of 14.751/100000 (P<0.001) higher than that in the control area, and the long-term effect of the implementation of the new model was a decreasing trend in the reported incidence of tuberculosis, with an average monthly decrease of 0.815/100000 (β₅+β₇=-0.815, P<0.001), and the rate of decline was greater than that in the control area (β₇=-0.931, P<0.001). Conclusion: There is a dynamic causal relationship between the implementation of the “Active Tuberculosis Screening+Full-Course Hospitalization” model and reported incidence of tuberculosis in the Hotan Prefecture of Xinjiang, and the new model has contributed to a long-term trend of decreasing reported incidence after a short-lived increase. The implementation of this model has obvious advantages for controlling tuberculosis epidemics in areas with a high tuberculosis burden.

Key words: Tuberculosis, pulmonary, Health policy, Outcome assessment (Health Care), Models, Statistical