Cost-effectiveness analysis of adding tuberculosis household contact investigation on passive case-finding strategy in Southwestern Uganda.

Abstract

The standard passive case-finding strategy implemented by most developing countries is inadequate to detect new cases of Tuberculosis. A household contact investigation is an alternative approach. However, there is limited cost-effectiveness data to support planning and implementation in low and middle-income countries. The study aimed to evaluate the cost-effectiveness of adding household contact investigation (HCI) to the passive case-finding (PCF) strategy in the Tuberculosis control program in Southwestern Uganda. We conducted an economic evaluation using a retrospective study approach and bottom-up costing (ingredients) techniques. It was a synthesis-based evaluation of existing data extracted from the District Health Information System (DHIS 2), TB registers, and a primary cost survey. The study compared two methods of Tuberculosis (TB) case finding (PCF and HCI) strategies. Regarding PCF, patients either self-reported their signs and symptoms or were prompted by healthcare workers. At the same time, HCI was done by home visiting and screening contacts of TB patients. Patients and household contacts presumed to have Tuberculosis were requested to produce samples for analysis. We applied a static decision-analytic modeling framework to examine both strategies' costs and effectiveness. The study relied on cost and probability estimates from National Tuberculosis (TB) program data, activity costs, and published literature. It was performed from the societal and provider perspectives over 1.5 years across 12 facilities in Ntungamo, Sheema, and Rwampara Districts. The primary effectiveness measure was the number of TB cases detected (yield) and the number needed to screen (NNS). The TB yield was calculated from the number of patients screened during the period under study. The incremental cost-effectiveness ratio (ICER) was expressed as cost in 2021 US$ per additional TB case detected. We did not apply a discount rate because of the short analytic time horizon. The unit costs of detecting a Tuberculosis case were US$ (United States dollar) 204.22 for PCF and US$ 315.07 for HCI. Patient and caregiver costs are five times more in PCF than in HCI [US$26.37 Vs. US$ 5.42]. The ICER was US$ 3,596.94 per additional TB case detected. The TB screening yields were 0.52% (1496/289140) for passive case finding and 5.8% (197/3414) for household contact investigation. Household contact investigation yield among children 0-14 Vs. 15+ years [6.2% Vs.5.4%] P = 0.04. The Yield among People living with HIV (PLHIV) Vs. HIV-negative [15.8% Vs.5.3%] P = 0.03 in HHCI. The PCF yield in men Vs. Women [1.12% Vs.0.28%] P<0.01. The NNS in PCF was 193 [95% CI: 186-294] and 17 [95% CI: 14-22] in HCI. Our baseline assumptions and the specific implementations of adding HCI to existing PCF programs in the context of rural African settings prove to be not cost-effective, rather than HCI as a strategy. HCI effectively identifies children and PLHIV with TB and should be prioritized. Meanwhile, the Passive case-finding strategy effectively finds men with TB and costs lower than household contact investigation.