Potential long-acting tuberculosis treatment for people with HIV: cost-effectiveness benchmarks.

RATIONALE: Long-acting tuberculosis (TB) drugs could improve treatment outcomes in South Africa, where there is high prevalence of HIV and TB coinfection and high attrition from standard TB treatment, but potential benefits must be weighed against costs.

OBJECTIVES: To identify scenarios where a long-acting TB treatment would provide clinical benefit and be cost-effective compared with standard TB treatment for people with HIV in South Africa.

METHODS: We used a microsimulation model to compare clinical and cost outcomes of (1) standard-of-care 6-month oral TB treatment (SOC); and (2) a long-acting treatment regimen, entailing standard oral therapy for 2&#x2009;months followed by a one-time injection (LA). Base case parameters for the SOC/LA strategies included efficacy, 97.6%/97.6%; disengagement on oral therapy, 4.8% per month; TB relapse, 0.34%/0.34% per month starting 5&#x2009;months after regimen completion; and total regimen cost, US$50.40/$199.80 (cost of long-acting injectable drug alone, $175). We varied these parameters in sensitivity analysis. Outcomes included mortality, treatment completions, TB relapses, life-years, and health system costs. We defined LA to be cost-effective compared to SOC when its incremental cost-effectiveness ratio (ICER) was <$3,000 per year-of-life saved (YLS).

RESULTS: In base case simulations, the LA strategy would produce lower all-cause mortality (11.4% vs 11.6%) and less disengagement (12.9% vs 20.9%) than SOC at 6&#x2009;months. LA would be cost-effective in the base case at the lifetime horizon (ICER $520/YLS) and at 24&#x2009;months (ICER $2230/YLS). Over the lifetime horizon, LA would be cost-effective compared with SOC over all parameter ranges evaluated in one-way sensitivity analysis. At 24&#x2009;months, LA would not be cost-effective compared to SOC when LA efficacy is <96%, monthly probability of disengagement from the oral regimen component of SOC and LA is <4%, or cost of the long-acting injectable drug is >$200.

CONCLUSIONS: There are combinations of efficacy, cost, and care disengagement at which a long-acting injectable TB treatment regimen could be cost-effective compared with the conventional oral treatment. Cost-effectiveness could be achieved if the long-acting treatment regimen maintains high efficacy at reasonable cost, especially when disengagement from -standard oral treatment is high.