An observational analysis of the impact of indoor residual spraying with non-pyrethroid insecticides on the incidence of malaria in Ségou Region, Mali: 2012–2015

BACKGROUND: Ségou Region in Central Mali is an area of high malaria burden with seasonal transmission, high access to and use of long-lasting insecticidal nets (LLINs), and resistance to pyrethroids and DDT well documented in Anopheles gambiae s.l. (the principal vector of malaria in Mali). Ségou has recently received indoor residual spraying (IRS) supported by Mali’s collaboration with the US President’s Malaria Initiative/Africa Indoor Residual Spraying programme. From 2012 to 2015, two different non-pyrethroid insecticides: bendiocarb in 2012 and 2013 and pirimiphos-methyl in 2014 and 2015, were used for IRS in two districts. This report summarizes the results of observational analyses carried out to assess the impact of these IRS campaigns on malaria incidence rates reported through local and district health systems before and after spraying. METHODS: A series of retrospective time series analyses were performed on 1,382,202 rapid diagnostic test-confirmed cases of malaria reported by district routine health systems in Ségou Region from January 2012 to January 2016. Malaria testing, treatment, surveillance and reporting activities remained consistent across districts and years during the study period, as did LLIN access and use estimates as well as An. gambiae s.l. insecticide resistance patterns. Districts were stratified by IRS implementation status and all-age monthly incidence rates were calculated and compared across strata from 2012 to 2014. In 2015 a regional but variable scale-up of seasonal malaria chemoprevention complicated the region-wide analysis; however IRS operations were suspended in Bla District that year so a difference in differences approach was used to compare 2014 to 2015 changes in malaria incidence at the health facility level in children under 5-years-old from Bla relative to changes observed in Barouéli, where IRS operations were consistent. RESULTS: During 2012–2014, rapid reductions in malaria incidence were observed during the 6 months following each IRS campaign, though most of the reduction in cases (70% of the total) was concentrated in the first 2 months after each campaign was completed. Compared to non-IRS districts, in which normal seasonal patterns of malaria incidence were observed, an estimated 286,745 total fewer cases of all-age malaria were observed in IRS districts. The total cost of IRS in Ségou was around 9.68 million USD, or roughly 33.75 USD per case averted. Further analysis suggests that the timing of the 2012–2014 IRS campaigns (spraying in July and August) was well positioned to maximize public health impact. Suspension of IRS in Bla District after the 2014 campaign resulted in a 70% increase in under-5-years-old malaria incidence rates from 2014 to 2015, significantly greater (p = 0.0003) than the change reported from Barouéli District, where incidence rates remained the same. CONCLUSIONS: From 2012 to 2015, the annual IRS campaigns in Ségou are associated with several hundred thousand fewer cases of malaria. This work supports the growing evidence that shows that IRS with non-pyrethroid insecticides is a wise public health investment in areas with documented pyrethroid resistance, high rates of LLIN coverage, and where house structures and population densities are appropriate. Additionally, this work highlights the utility of quality-assured and validated routine surveillance and well defined observational analyses to rapidly assess the impact of malaria control interventions in operational settings, helping to empower evidence-based decision making and to further grow the evidence base needed to better understand when and where to utilize new vector control tools as they become available. ELECTRONIC SUPPLEMENTARY MATERIAL: The online version of this article (10.1186/s12936-017-2168-2) contains supplementary material, which is available to authorized users.