Cost-consequence analysis of ambulatory clinic- and home-based multidrug-resistant tuberculosis management models in Eswatini

Background: We compared the cost-consequence of a home-based multidrug-resistant tuberculosis (MDR-TB) model of care, based on task-shifting of directly observed therapy (DOT) and MDR-TB injection administration to lay health workers, to a routine clinic-based strategy within an established national TB programme in Eswatini. Methods: Data on costs and effects of the two ambulatory models of MDR-TB care was collected using documentary data and interviews in the Lubombo and Shiselweni regions of Eswatini. Health system, patient and caregiver costs were assessed in 2014 in US$ using standard methods. Cost-consequence was calculated as the cost per patient successfully treated. Results: In the clinic-based and home-based models of care, respectively, a total of 96 and 106 MDR-TB patients were enrolled in 2014, with treatment success rates of 67.8% and 82.1%. Health system costs per patient treated were slightly lower in the home-based strategy (US$19 598) compared to the clinic-based model (US$20 007). The largest costs in both models were for inpatient care, administration of DOT and injectable treatment, and drugs. Costs incurred by patients and caregivers were considerably higher in the clinic-based model of care due to the higher direct travel costs to the nearest clinic to receive DOT and injections daily. In total, MDR patients in the clinic-based strategy incurred average costs of US$670 compared to US$275 for MDR-TB patients in the home-based model. MDR-TB patients in the home-based programme, where DOT and injections was provided in their homes, only incurred out-of-pocket travel expenses for monthly outpatient treatment monitoring visits averaging US$100. The cost per successfully treated patient was US$31 106 and US$24 157 in the clinic-based and home-based models of care, respectively. The analysis showed that, in addition to the health benefits, direct and indirect costs for patients and their caregivers were lower in the home-based care model. Conclusion: The home‐based strategy used less resources and generated substantial health and economic benefits, particularly for patients and their caregivers, and decision makers can consider this approach as an alternative to expand and optimise MDR-TB control in resource-limited settings. Further research to understand the appropriate mix of treatment support components that are most important for optimal clinical and public health outcomes in the ambulatory home-based model of MDR-TB care is necessary.