Assessing the effect of health system resources on HIV and tuberculosis programmes in Malawi: a modelling study

Tara D Mangal,Sakshi Mohan,Timothy Colbourn,Joseph H Collins,Mathew Graham,Andreas Jahn,Eva Janoušková,Ines Li Lin,Robert Manning Smith,Emmanuel Mnjowe,Margherita Molaro,Tisungane E Mwenyenkulu,Dominic Nkhoma,Bingling She,Asif Tamuri,Paul Revill,Andrew N Phillips,Joseph Mfutso-Bengo,Timothy B Hallett

DOI: https://doi.org/10.1016/S2214-109X(24)00259-6

Abstract:Background: Malawi is progressing towards UNAIDS and WHO End TB Strategy targets to eliminate HIV/AIDS and tuberculosis. We aimed to assess the prospective effect of achieving these goals on the health and health system of the country and the influence of consumable constraints. Methods: In this modelling study, we used the Thanzi la Onse (Health for All) model, which is an individual-based multi-disease simulation model that simulates HIV and tuberculosis transmission, alongside other diseases (eg, malaria, non-communicable diseases, and maternal diseases), and gates access to essential medicines according to empirical estimates of availability. The model integrates dynamic disease modelling with health system engagement behaviour, health system use, and capabilities (ie, personnel and consumables). We used 2018 data on the availability of HIV and tuberculosis consumables (for testing, treatment, and prevention) across all facility levels of the country to model three scenarios of HIV and tuberculosis programme scale-up from Jan 1, 2023, to Dec 31, 2033: a baseline scenario, when coverage remains static using existing consumable constraints; a constrained scenario, in which prioritised interventions are scaled up with fixed consumable constraints; and an unconstrained scenario, in which prioritised interventions are scaled up with maximum availability of all consumables related to HIV and tuberculosis care. Findings: With uninterrupted medical supplies, in Malawi, we projected HIV and tuberculosis incidence to decrease to 26 (95% uncertainty interval [UI] 19-35) cases and 55 (23-74) cases per 100 000 person-years by 2033 (from 152 [98-195] cases and 123 [99-160] cases per 100 000 person-years in 2023), respectively, with programme scale-up, averting a total of 12·21 million (95% UI 11·39-14·16) disability-adjusted life-years. However, the effect was compromised by restricted access to key medicines, resulting in approximately 58 700 additional deaths (33 400 [95% UI 22 000-41 000] due to AIDS and 25 300 [19 300-30 400] due to tuberculosis) compared with the unconstrained scenario. Between 2023 and 2033, eliminating HIV treatment stockouts could avert an estimated 12 100 deaths compared with the baseline scenario, and improved access to tuberculosis prevention medications could prevent 5600 deaths in addition to those achieved through programme scale-up alone. With programme scale-up under the constrained scenario, consumable stockouts are projected to require an estimated 14·3 million extra patient-facing hours between 2023 and 2033, mostly from clinical or nursing staff, compared with the unconstrained scenario. In 2033, with enhanced screening, 188 000 (81%) of 232 900 individuals projected to present with active tuberculosis could start tuberculosis treatment within 2 weeks of initial presentation if all required consumables were available, but only 8600 (57%) of 15 100 presenting under the baseline scenario. Interpretation: Ignoring frailties in the health-care system, in particular the potential non-availability of consumables, in projections of HIV and tuberculosis programme scale-up might risk overestimating potential health impacts and underestimating required health system resources. Simultaneous health system strengthening alongside programme scale-up is crucial, and should yield greater benefits to population health while mitigating the strain on a heavily constrained health-care system. Funding: Wellcome and UK Research and Innovation as part of the Global Challenges Research Fund.