Back

Implementation of enhanced 99DOTS for TB treatment supervision in Uganda: An interrupted time series analysis

Crowder, R.; Nakasendwa, S.; Kityamuwesi, A.; Musoke, M.; Nannozi, J.; Waswa, J.; Nakate, A. S.; Lamunu, M.; Twinamasiko, A.; Kunihira Tinka, L.; Oyuku, D.; Babirye, D.; Berger, C.; Thompson, R.; Turyahabwe, S.; Dowdy, D.; Katamba, A.; Cattamanchi, A.; Kiwanuka, N.

2024-01-23 infectious diseases
10.1101/2024.01.22.24300949 medRxiv
Show abstract

BackgroundDigital adherence technologies like 99DOTS are being scaled-up for tuberculosis (TB) treatment despite limited evidence of their effectiveness and concerns about accessibility. We evaluated the reach and effectiveness of 99DOTS during its scale-up in Uganda. MethodsWe included all adults initiating drug-susceptible pulmonary TB treatment between August 2019 and June 2021 at 12 99DOTS-naive health facilities (n=3,526). Using an interrupted time series design, we compared the proportion with treatment success (primary outcome) and enrolled on 99DOTS in the 9 months before and 12 months after implementing an enhanced 99DOTS intervention. This included providing low-cost phones to people with TB when needed, task shifting monitoring and support to community health workers, and automated task lists. Treatment data were derived from routine TB treatment registers. ResultsThe proportion enrolled on 99DOTS post-intervention was high (87.2%) and had a slight upward trend (proportion ratio 1.01, 95% CI 1.00-1.01). Treatment completion remained similar across periods (89.8% post-intervention vs. 87.1% pre-intervention). There was no immediate level or slope change in treatment success following the intervention. ConclusionsWith appropriate implementation supports, 99DOTS can have high uptake without negatively impacting treatment outcomes. Equity in access should be prioritized during implementation

Matching journals

The top 7 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.