Cost-effectiveness analysis of the multi-strategy WHO emergency care toolkit in regional referral hospitals in Uganda

Background: Low- and middle-income countries bear a disproportionate amount of the global burden of disease from emergency conditions. To improve the provision of emergency care in low-resource settings, a multifaceted World Health Organization (WHO) intervention introduced a toolkit including Basic Emergency Care training, resuscitation area guidelines, a trauma registry, a trauma checklist, and triage tool in two public hospital sites in Uganda. While introduction of the toolkit revealed a large reduction in the case fatality rate of patients, little is known about the cost-effectiveness and affordability. We analysed the cost-effectiveness of the toolkit and conducted a budget analysis to estimate the impact of scale up to all regional referral hospitals for the national level. Methods: A decision tree model was constructed to assess pre- and post-intervention groups from a societal perspective. Data regarding mortality were drawn from WHO quality improvement reports captured at two public hospitals in Uganda from 2016–2017. Cost data were drawn from project budgets and included direct costs of the implementation of the intervention, and direct costs of clinical care for patients with disability. Development costs were not included. Parameter uncertainty was assessed using both deterministic and probabilistic sensitivity analyses. Our model estimated the incremental cost-effectiveness of implementing the WHO emergency care toolkit measuring all costs and outcomes as disability-adjusted life-years (DALYs) over a lifetime, discounting both costs and outcomes at 3.5%. Results: Implementation of the WHO Toolkit averted 1,498 DALYs when compared to standard care over a one-year time horizon. The initial investment of $5,873 saved 34 lives (637 life years) and avoided $1,670,689 in downstream societal costs, resulted in a negative incremental cost-effectiveness ratio, dominating the comparator scenario of no intervention. This would increase to saving 884 lives and 25,236 DALYs annually with national scale up. If scaled to a national level the total intervention cost over period of five years would be $4,562,588 or a 0.09% increase of the total health budget for Uganda. The economic gains are estimated to be $29,880,949 USD, the equivalent of a 655% return on investment. The model was most sensitive to average annual cash income, discount rate and frequency survivor is a road-traffic incident survivor, but was robust for all other parameters. Conclusion: Improving emergency care using the WHO Toolkit produces a cost-savings in a low-resource setting such as Uganda. In alignment with the growing body of literature highlighting the value of systematizing emergency care, our findings suggest the toolkit could be an efficient approach to strengthening emergency care systems.