Back

Pretreatment Attrition, Treatment Delay and Treatment Outcome in Drug-Resistant Tuberculosis: A Mixed-Methods Cohort Study in Public and Private Hospitals in Indonesia

Azzumar, F.; Rhamadianti, A. F.; Anugrapaksi, E.; Wardoyo, M. P.; Hakiman, A. P. A.; Putri, L. Z.; Sawitri, N.; Kemalasari, K.; Sutarto, R.; Fuady, A.; Burhan, E.

2026-10-01 health systems and quality improvement
10.64898/2026.09.29.26364370 medRxiv
Show abstract

Background Drug-resistant tuberculosis (DR-TB) remains a major problem in Indonesia, with a significant low treatment initiation rate. This study identifies factors influencing treatment initiation and assess the impact of delay on patient outcomes. Methods We conducted a mixed-methods study using a parallel convergent design. The quantitative component analyzed the Tuberculosis Information System (TBIS) data from three hospitals in Jakarta and Bogor from January 2022 to December 2023. The qualitative component involved in-depth interviews and Focus Group Discussions (FGD) with DR-TB patients, healthcare workers, patient supporters, and program managers. Results Among 1,665 people with DR-TB included in this study, we found a pre-treatment attrition rate of 6.4%. We identified two factors of pre-treatment attrition: prior TB treatment (aOR=1.7, 95%CI 1.082-2.671) and distance to hospital (aOR=1.9, 95%CI: 1.136-3.180). Conversely, referred patients, treatment in suburban areas, and people with rifampicin-resistant (RR)-TB were 85%, 82%, and 63% less likely to experience attrition, respectively. Qualitative findings identify patient, social, and health system barriers to the treatment initiation, and alternative remedies emerged as a coping mechanism. Conclusion Strengthening referral systems, decentralizing care, and addressing socio-economic barriers are critical for ensuring timely DR-TB treatment initiation.

Matching journals

The top 2 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.