High net coverage, incomplete protection: factors associated with malaria resurgence in three elimination districts of Zimbabwe1. Rameck Makokove, Plan International Zimbabwe, Malaria Prevention Control and Elimination Programme

Abstract Background In Zimbabwe, insecticide-treated nets (ITNs) are the core intervention in low-transmission districts pursuing sub-national malaria elimination. Masvingo, Chegutu and Gokwe South entered the elimination phase in 2017, yet malaria incidence rose despite ITN coverage exceeding 95%. The drivers of this resurgence under high net coverage remain uncharacterised. This study assessed the behavioural, socio-economic, entomological and environmental factors associated with malaria resurgence and ITN utilisation across the three districts. Methods A mixed-methods design was used, comprising a cross-sectional analytical survey complemented by a qualitative strand, an entomological survey and environmental remote sensing. A stratified multistage sample of 314 household heads was surveyed on knowledge, attitudes and practices, net utilisation and outdoor activities. Eleven environmental health technicians and 14 village health workers were interviewed as key informants. Host-seeking mosquitoes were sampled indoors and outdoors using CDC light traps, and climatic and vegetation trends (2016–2020) were analysed using NASA POWER and Landsat data. Associations of socio-demographic factors with ITN utilisation were assessed using a population-averaged logistic regression model, accounting for clustering of households within wards, while qualitative and climate data were analysed thematically and with the Mann–Kendall test, respectively. Results Knowledge of malaria was high, with 81.8% identifying mosquitoes as the mode of transmission, but only 11.8% knew the correct three-year net replacement interval. Although 83% owned ITNs, only 49% reported sleeping under a net the previous night; use was seasonal, with 54.1% using nets only during the rainy season, and 68% of those needing replacements failed to obtain them. Higher education independently predicted positive attitudes towards nets. Receiving malaria information (adjusted odds ratio [aOR] 1.75, 95% CI 1.18–2.59) and older age (aOR 1.63, 95% CI 1.12–2.38) were independently associated with net use, whereas women were less likely than men to have used a net (aOR 0.73, 95% CI 0.55–0.96). Outdoor livelihood, water-collection and socio-cultural activities coincided with peak vector-biting hours. Anopheles gambiae sensu lato and the Anopheles funestus group were active both indoors and outdoors, with An. gambiae showing bimodal biting. No statistically significant long-term trends in temperature, rainfall or humidity were detected. Village health workers were the main and most preferred source of malaria information. Conclusion Malaria resurgence in these elimination districts is multifactorial rather than the failure of a single intervention. High awareness (81.8%) and net ownership (83%) coexisted with seasonal and inconsistent use, barriers to replacement, and outdoor activity overlapping with vector biting. Sustaining elimination gains requires accessible net replacement, practical and segmented social and behaviour change communication, strengthened entomological surveillance, and contextually tailored interventions for residual outdoor transmission.