Cost-effectiveness of in-kind nutritional support for impoverished persons with tuberculosis to reduce mortality and disengagement from care in India: a modelling study

BACKGROUND: Undernutrition is a leading risk factor for tuberculosis (TB) mortality and poor treatment outcomes. We evaluated the cost-effectiveness of providing in-kind nutritional support to impoverished persons with TB (PWTB) in India. METHODS: We developed a Markov model comparing standard care with a household-level nutritional food-basket intervention. Model parameters were informed from systematic reviews and a recent cluster-randomised study. We estimated TB mortality, disability-adjusted life years (DALYs) and costs from the healthcare perspective. We calculated incremental cost-effectiveness ratios (ICERs), and parameter uncertainty was quantified through deterministic and probabilistic sensitivity analyses. We estimated uncertainty intervals (UIs) with 10 000 Monte Carlo iterations. FINDINGS: Nutritional support was projected to avert 10 470 DALYs per 10 000 PWTB (95% UI 1 775 to 33 255). When scaled to India's 2.8 million annual TB cases, this corresponds to approximately 120 120 TB deaths averted nationwide (95% UI 60 760 to 162 960) under full coverage. The ICER was US$141 per DALY averted (95% UI US$44 to 836). The ICER was most sensitive to the mortality reduction achieved through nutritional support but remained below India's willingness-to-pay threshold (US$550 per DALY averted) as long as the intervention achieved at least a 13% relative reduction in mortality. Nutritional support was cost-effective in 94% of simulations. INTERPRETATION: In-kind nutritional support for PWTB is highly cost-effective and could substantially reduce TB mortality and disengagement from care. Scaling up such interventions could meaningfully improve TB outcomes nationwide and accelerate progress towards the end TB strategy targets.