Changing trend analysis on early detection of rifampicin resistant tuberculosis patients in southwestern area of China, 2016–2020

There were no data about prevention and control status of RR-TB in a poor area with high burden of TB in China. In order to develop evidence-based RR-TB response strategies and improve enrollment of RR-TB patients in Yunnan province, China, this study was aimed at analyzing the changing trends in the detection and enrollment of RR-TB patients and examining the factors that may have implication on enrollment in treatment. Data, which includes demographics, screening and testing, and treatment enrollment, was collected from the TB Management Information System. Retrospective data analysis and factors analysis were applied. Descriptive statistics, Chi-square test, Rank sum test and logistic regression analysis were used. From 2016 and 2018, the province had been challenged by low levels of screening, detection and enrollment of RR-TB. During the period between 2019 and 2020, a comprehensive model of RR-TB prevention and control was established in Yunnan, characterized by a robust patient-centered approach for RR-TB care and multiple, targeted interventions through the cascade of care from detection to treatment. In 2020, 93.8% of the bacteriologically positive TB patients were screened for RR-TB, which had been significantly increased by 146.9% from 38.0% in 2016. The interval from initial consultation at RR-TB facility to diagnosis (inter-quartile range) was reduced from 29.5 (1-118) days in 2016 to 0 (0-7) days in 2020. Despite the increasing rates of enrollment of RR-TB patients over the years, non-enrollment of those detected was still high (32.3%) in 2020. The main reasons for non-enrollment identified were refusal of treatment due to financial difficulties, loss to follow-up or death before starting treatment. Multivariate analysis showed that the elderly patients aged 65 or above (OR = 2.7, CI: 1.997-3.614), new patients (OR = 0.7, CI: 0.607-0.867), conventional DST used for confirmatory diagnosis of RR-TB (OR = 1.9, CI: 1.620-2.344) and diagnosis of RR-TB being conducted by the RR-TB care facilities at the prefecture and municipal level (OR = 4.4, CI: 3.608-5.250) have implications on RR-TB non-enrollment. As a comprehensive RR-TB model was implemented in Yunnan with scaled up use of molecular test for rapid detection of RR-TB, initial screening of RR-TB were decentralized to the county- and district-level to strengthen rapid, early detection of RR-TB, achieving a higher coverage of screening in the end. However, there remains a major gap in enrollmen