RESUMEN
BACKGROUND: Ghana has been implementing the indoor residual spraying (IRS) of insecticides since 2006, focusing operations in the north. Insecticide resistance concerns prompted a switch from pyrethroids to organophosphates, beginning gradually in 2011 and switching fully to the micro-encapsulated formulation of pirimiphosmethyl (PM CS), Actellic® 300CS, a third-generation indoor residual spraying (3GIRS) product, by 2014. Entomological surveillance studies have shown IRS to be a highly effective malaria control tool, but epidemiological evidence is needed as well. Countrywide prevalence surveys have shown that malaria parasite prevalence in children under 5 years of age in Northern, Upper East, and Upper West Regions had declined to less than 40% in each region by 2016. Similarly, malaria deaths in children under 5 years of age have also been declining nationally since 2009. Although IRS is suspected to have contributed to this decline, stronger evidence is needed to link the IRS interventions to the epidemiological impact. METHODS: To assess the epidemiological impact of Ghana's IRS programmatic activities, a retrospective, observational analysis using routine epidemiological data was conducted to compare malaria incidence rates from IRS and non-IRS districts in Northern, Upper East, and Upper West Regions. Routine epidemiological data consisted of passive malaria case surveillance data reported in the District Health Information System 2 (DHIS2); with cases representing patients with suspected malaria who had sought care in the public health system and had received a confirmatory diagnosis with a positive malaria RDT result. Final routine data were extracted in September 2018. All districts that had received IRS were included in the analysis and compared to all non-IRS districts within the same region. In the Northern Region, only PMI districts were included in the analysis, as they had similar historical data. RESULTS: District-level analysis from Northern Region from 2015 to 2017 of the aggregate malaria incidence reported from IRS districts relative to non-IRS comparator districts showed 39%, 26%, and 58% fewer confirmed malaria cases reported from IRS districts in 2015, 2016, and 2017, respectively. This translates to approximately 257,000 fewer cases than expected over the three years. In Upper East Region, the effect on reported malaria cases of withdrawing IRS from the region was striking; after spray operations were suspended in 2015, incidence increased an average of 485% per district (95% confidence interval: 330% to 640%) compared to 2014. CONCLUSIONS: The current observational analysis results are in line with the entomological studies in demonstrating the positive contribution of IRS with a 3GIRS product to malaria control programmes in northern Ghana and the value of using routine surveillance and implementation data to rapidly assess the impact of vector control interventions in operational settings, even in complex implementation environments.