Cost-effectiveness of adding indoor residual spraying to case management in Afghan refugee settlements in Northwest Pakistan during a prolonged malaria epidemic

INTRODUCTION Financing of malaria control for displaced populations is limited in scope and duration, making cost-effectiveness analyses relevant but difficult. This study analyses cost-effectiveness of adding prevention through targeted indoor residual spraying (IRS) to case management in Afghan refugee settlements in Pakistan during a prolonged malaria epidemic. METHODS/FINDINGS An intervention study design was selected, taking a societal perspective. Provider and household costs of vector control and case management were collected from provider records and community survey. Health outcomes (e.g. cases and DALYs averted) were derived and incremental cost-effectiveness ratios (ICERs) for cases prevented and DALYs averted calculated. Population, treatment cost, women's time, days of productivity lost, case fatality rate, cases prevented, and DALY assumptions were tested in sensitivity analysis. Malaria incidence peaked at 44/1,000 population in year 2, declining to 14/1,000 in year 5. In total, 370,000 malaria cases, 80% vivax, were diagnosed and treated and an estimated 67,988 vivax cases and 18,578 falciparum and mixed cases prevented. Mean annual programme cost per capita was US$0.56. The additional cost of including IRS over five years per case prevented was US$39; US$50 for vivax (US$43 in years 1-3, US$80 in years 4-5) and US$182 for falciparum (US$139 in years 1-3 and US$680 in years 4-5). Per DALY averted this was US$266 (US$220 in years 1-3 and US$486 in years 4-5) and thus 'highly cost-effective' or cost-effective using WHO and comparison thresholds. CONCLUSIONS Adding IRS was cost-effective in this moderate endemicity, low mortality setting. It was more cost-effective when transmission was highest, becoming less so as transmission reduced. Because vivax was three times more common than falciparum and the case fatality rate was low, cost-effectiveness estimations for cases prevented appear reliable and more definitive for vivax malaria.