Baseline malaria burden and pyrethroid resistance in Muheza, Tanzania informing a cluster randomized trial of the 3D window screens

Insecticide resistance in malaria vectors has created an urgent need for alternative, non-insecticidal control tools. In response, a novel window screen, the Three-dimensional (3D) Screen, was developed. Before launching a cluster-randomized controlled trial to evaluate its impact, a baseline study was conducted in 20 hamlets (clusters) from 17 villages in Muheza District, northeastern Tanzania, to assess malaria burden, vector characteristics, and risk factors for infection. Structured questionnaires collected household information; malaria prevalence was measured among 778 children aged 6 months to 14 years using malaria rapid diagnostic tests (mRDTs), and haemoglobin was assessed by HemoCue. Indoor Centers for Disease Control and Prevention (CDC) light traps (LTs) were used for mosquito collections, yielding 14,263 mosquitoes morphologically identified, including 889 Anopheles gambiae sensu lato (s.l.) and 2,013 An. funestus senso lato (s.l.); 2,508 female Anopheles were further analysed by Polymerase chain reaction (PCR). World Health Organization (WHO) cylinder bioassays assessed pyrethroid resistance in An. gambiae s.l., and genotyping targeted the knockdown resistance (kdr)-East (L1014S) mutation. Malaria prevalence was 40.2%, and anaemia affected 55.2% of children. An. funestus s.l. (70.2%) and An. gambiae s.l. (29.8%) comprised the vector population. Mean daily entomological inoculation rates (EIRs) were 0.039 and 0.037 infectious bites per person per night, corresponding to annual EIRs of approximately 14.2 for An. funestus s.l., 13.5 for An. gambiae s.l., and 27.7 combined. Pyrethroid resistance was high, with 24-hour mortality of 56.5% to permethrin and 52.8% to deltamethrin; the kdr-East (L1014S) allele frequency in An. gambiae s.s. was 0.45. Multivariate analysis showed that children aged ≥ 5 years had approximately threefold higher odds of Plasmodium infection than those < 5 years (aOR = 3.03, 95% CI: 2.09–4.40), while sleeping under a bed net was protective (aOR = 0.36, 95% CI: 0.18–0.70). Cluster-level variance accounted for 19.5% of infection risk, highlighting community-level factors. Persistent transmission, high anaemia prevalence, and widespread insecticide resistance underscore the limitations of current measures and the need for integrated interventions, including novel non-insecticidal tools such as the 3D-Screen.