Abstract

Objectives

To evaluate the cost-effectiveness of testing and treatment for Mycobacterium tuberculosis (Mtb) infection among Asian and Hispanic persons with diagnosed diabetes in the United States.

Methods

We estimated population size and Mtb infection prevalence for Asian and Hispanic persons aged ≥15 years with diagnosed, nongestational diabetes, by age and US-born status. We assumed a 1-time test for Mtb infection intervention, with positive-testing persons offered treatment. Using a deterministic, transmission-dynamic model of incident tuberculosis (TB) in the United States, we estimated costs, TB cases and deaths averted, and quality-adjusted life years gained under the intervention compared with no intervention. We estimated incremental cost-effectiveness ratios (ICERs), calculated as costs per quality-adjusted life years gained, from a TB health services perspective, including diagnosis and treatment for TB infection and disease. We also assessed health services and societal perspectives. We estimated 95% uncertainty intervals via probabilistic sensitivity analysis.

Results

TB cases averted per 100 000 persons tested ranged from 7.5 (95% uncertainty interval: 6.9-8.1) among US-born Hispanic persons to 238.9 (225.2-254.3) among non-US-born Asian persons. TB deaths averted per 100 000 persons tested ranged from 1.3 (1.2-1.4) among US-born Hispanic persons to 53.7 (51.4-56.1) among non-US-born Asian persons. ICERs for US-born Asian and Hispanic populations were $856 671 ($533 506-$1 234 032) and $1 081 646 ($673 142-$1 551 264), respectively. ICERs for non-US-born Asian and Hispanic populations were lower: $66 664 ($41 456-$93 625) and $68 749 ($43 136-$97 044), respectively. ICERs were 2% to 19% higher under a societal perspective.

Conclusions

Although the intervention produced health benefits for all populations assessed, health benefits were greater—and ICERs more favorable—for non-US-born Asian and Hispanic populations with diagnosed diabetes.