Setting: Papua New Guinea (PNG) has one of the world's highest TB incidence rates. It is difficult for patients to access TB care in remote provinces due to insufficient infrastructure and challenging terrain, making varied, targeted delivery models for treating TB necessary.

Objective: To assess treatment outcomes using self-administered treatment (SAT), family-supported treatment and community-based directly observed therapy (DOT) via treatment supporter (TS) in the PNG context.

Design: A retrospective, descriptive analysis of routinely collected data from 360 patients at two sites in 2019-2020. All patients were assigned a treatment model based on risk factors (adherence or default) and offered patient education and counselling (PEC), family counselling and transportation fees. End-of-treatment outcomes were assessed for each model.

Results: Treatment success rates among drug-susceptible TB (DS-TB) were good overall: 91.1% for SAT, 81.4% for family-supported treatment and 77% for DOT patients. SAT was strongly associated with favourable outcomes (OR 5.7, 95% CI 1.7-19.3), as were PEC sessions (OR 4.3, 95% CI 2.5-7.2).

Conclusion: By considering risk factors when determining their treatment delivery model, strong outcomes were seen in all three groups. Multiple modes of treatment administration, tailored to individuals' needs and risk factors, is a feasible, effective, patient-centred care model for hard-to-reach, resource-limited settings.