Evaluation of treatment outcomes of patients with drug-resistant tuberculosis in Papua New Guinea: a retrospective cohort study (2017-2023)
Rendi Moke, Jennifer Koponia Banamu, Arnold Bainomugisa, John Kamakom, Lucy N John, Suzanne Gurra, Margaret Kal, Herolyn Nindil, Robin Yasi, Ken Witari, Miriam Boga, Israel Naraman, Mary Kaevakore, Simberi Jojoga, Dennis Tony, Chris Ope, Petronella Walando, Benson Meles, Martin Daimen, Weston Yambut, Kassi Ken, Bradley Aumora, Samantha Komoru, Sushil Pandey, Sarah Dunstan, Philipp du Cros, Lachlan J M CoinAbstract
Background and Aims
Papua New Guinea has a high burden of drug-resistant TB (DR-TB). Reported treatment success has been lower than the WHO-targets. This study aimed to evaluate factors associated with unfavourable treatment outcome of patients with DR-TB.
Methods
Clinical, demographic and programmatic data were collected from patients with DR-TB confirmed by phenotypic drug susceptibility testing (pDST), who were enrolled for treatment between 2017 and 2023 in PNG. Univariable analysis and multivariable Firth penalised logistic regression were performed to assess factors associated with unfavourable treatment outcomes. Firth-penalised Cox proportional hazards regression for time-to-event analysis, with Kaplan-Meier methods used to estimate event-free survival by treatment regimen.
Results
Of 348 pDST-confirmed DR-TB cases, clinical data were available for 249 patients, of whom 82% (n = 204) had documented treatment outcomes and 76.7% (n = 191) had complete SLD regimen and treatment outcomes. 32.4% (66/204) had unfavourable treatment outcome. In the primary regimen-outcome cohort (n = 191), SLD-injectable treatment was associated with higher odds of an unfavourable outcome in crude analysis (OR 2.36, 95% CI 1.27-4.46; p = 0.006), but this was attenuated after adjustment (aOR 1.92, 95% CI 0.98-3.82; p = 0.056). Western Province was associated with lower odds after adjustment (aOR 0.14, 95% CI 0.03-0.53; p = 0.004) and remained strongly associated with lower hazard (aHR 0.14, 95% CI 0.01-0.53; p = 0.002). Kaplan-Meier analysis showed significantly lower event-free survival among patients receiving SLD-injectable treatment (log-rank p = 0.017). Median pDST turnaround time (TAT) was 105 days (IQR 79–149), with 54% (134/249) exceeding 90 days, but showed no association with unfavourable treatment outcome.
Conclusions
Favourable treatment outcome was associated with residence in Western Province, while injectable-containing treatment was not independently associated with poorer treatment outcomes after adjustment. Strengthening access to timely complete drug resistance profile, scaling up all-oral regimens, treatment support and routine data quality are essential to improving DR-TB outcomes in PNG.