Impact of malaria vector control interventions implemented in Luangwa District southeastern of Zambia: A 13-year observational time series analysis of malaria trends.

INTRODUCTION: Luangwa District has one of the longest running legacy datasets in Zambia regarding reliable monitoring of confirmed malaria cases through the national Health Management Information System (HMIS). It was also one of the first districts to achieve sustained coverage with long-lasting insecticidal nets (LLINs) and indoor residual spraying (IRS) of insecticides. METHODS: HMIS data from2009 to 2021 were analysed using generalized linear mixed models, to assess the effects of LLINs and IRS on rates of inpatient admissions with severe malaria and total confirmed malaria cases. IRS treatments included the pyrethroids deltamethrin and lambda-cyhalothrin, the organophosphate pirimiphos-methyl as emulsifiable concentrate and micro-encapsulated formulations (PM-EC and PM-CS, respectively) and a deltamethrin coformulation with the neonicotinoid clothianidin (DC). RESULTS: IRS with PM-CS reduced both inpatient admissions (Relative rates (RR) and 95% confidence intervals (CI) for ≤3 months, 4-6 months and 7-12 months post spray = 0.28 [0.19, 0.98] (P = 0.0019), 0.46 [0.31, 0.96] (P = 0.0346) and 0.41 [0.35, 0.94] (P = 0.0174), respectively), and total cases (RR [95%CI] = 0.25 [0.01, 0.67] (P = 0.0017), 0.66 [0.11, 0.88] (P = 0.0087) and 0.48 [0.28, 0.96] (P = 0.0018) for the same post-spray intervals, respectively) for a full year. Furthermore, while reductions of inpatient admissions with severe malaria could only be attributed to DC for the first three months after spraying (RR [95% CI] = 0.27 [0.10, 0.68], P = 0.0379), impacts upon total malaria cases were also apparent for a full year (RR [95% CI] for 1-3 months, 4-6 months and 7-12 months post spray = 0.15 [0.10, 0.68] (P = 0.0017), 0.23 [0.05, 0.56] (P = 0.0013) and 0.43 [0.25, 0.86], P = 0.0029, respectively). Overall, there were >90% fewer inpatient admissions and >80% fewer cases by the end of the study, much of which could be attributed to the immediate effects of scaling up IRS with PM-CS or DC in late 2014 (RR [95% CI = 0.36[0.26,0.51] per year, P = 0.0001) and simultaneously almost doubling the number of health facilities across the district in mid-2016 (RR [95% CI = 0.85 [0.76,0.96] per year, P = 0.0088). CONCLUSIONS: IRS with durable non-pyrethroid insecticide formulations and improved access to diagnosis and treatment were both clearly associated with substantial incremental reductions of malaria incidence. While no epidemiological effect could be attributed to LLINs, this presumably occurred because coverage was already high at the outset and remained so throughout the study.