House modifications using insecticide-treated screening of eave and window as a vector control tool: evidence from a semi-field system in Tanzania and simulated epidemiological impact

BACKGROUND: Gaps in unimproved house structures, especially in eaves and windows, allow mosquito entry, increasing indoor vector-borne disease transmission. Simple modifications to such houses may reduce human exposure, and insecticide treatment may kill mosquitoes, benefiting all community members. This study evaluated insecticide-treated screening (ITS) for eaves and windows, incorporated with deltamethrin and piperonyl butoxide (PBO), compared to a permethrin and PBO-treated bednet in Tanzania. METHOD: A randomised Latin-square design (4 × 4) was used in four experimental huts within a large netting cage to allow mosquito recapture inside and outside of huts. Four treatments were evaluated: (1) new (12-month stored) ITS; (2) 12-month naturally-aged ITS; (3) 12-month field-used pyrethroid-PBO bednet (standard of care in Tanzania), and (4) no treatment. The study was performed for 32 nights using 30 mosquitoes per strain, per hut per night. Four laboratory-reared strains were used: malaria vectors (Anopheles arabiensis and An. funestus), dengue vector (Aedes aegypti), and nuisance biting (Culex quinquefasciatus). Recaptured mosquitoes were assessed for mortality at 72 h, blood-feeding, and hut entry. A simulation with a modified mechanistic model tracking Plasmodium falciparum malaria was used to illustrate potential epidemiological impact from these products. RESULTS: Against all mosquito species compared to 12-month aged pyrethroid-PBO-treated bednet, new ITS induced higher mortality [Odds Ratio:2.25(95%Confidence Interval:1.65-3.06),p < 0.0001], and aged-ITS was similar [OR:0.80(95%CI:0.59-1.08),p = 0.141]. Both new and aged ITS significantly (p < 0.0001) reduced mosquito blood-feeding [new OR:0.02(95% CI:0.01-0.03); aged OR:0.09(95%CI:0.05-0.14)] and hut entry [new IRR:0.10(95%CI:0.08-0.13); aged IRR:0.25(95%CI:0.21-0.31)]. Transmission model estimates indicate epidemiological impacts of ITS may supersede pyrethroid-PBO-treated bednets at the population level. The model results indicate impact potency depends on assumed intervention percentage cover, durability, and mosquito bionomics. CONCLUSIONS: This study introduces a standardised semi-field bioassay to evaluate ITS, generating entomological data that, for the first time, enabled modelling of its potential impact on malaria transmission. ITS is an efficacious tool for public health as it kills substantial proportions of malaria and dengue vectors, and reduces nuisance biting. Given its simplicity, it should be considered an additional or stand-alone tool for screening unimproved houses.