Eave tubes for malaria control in Africa: prototyping and evaluation against Anopheles gambiae s.s. and Anopheles arabiensis under semi-field conditions in western Kenya

Abstract

Background: Whilst significant progress has been made in the fight against malaria, vector control continues to rely on just two insecticidal methods, i.e., indoor residual spraying and insecticidal bed nets. House improvement shows great potential to complement these methods and may further reduce indoor mosquito biting and disease transmission. Open eaves serve as important mosquito house entry points and provide a suitable location for intercepting host-seeking anophelines. This study describes semi-field experiments in western Kenya with eave tubes, a household protection product that leverages the natural behaviour of host-seeking malaria mosquitoes.

Methods: Semi-field experiments were conducted in two screen-houses. In both of these a typical western Kenyan house, with mud walls and corrugated iron sheet roofing, was built. Eave tubes with bendiocarb- or deltamethrin-treated eave tube inserts were installed in the houses, and the impact on house entry of local strains of Anopheles gambiae and Anopheles arabiensis was determined. Experiments with open eave tubes (no netting) were conducted as a control and to determine house entry through eave tubes. Insecticidal activity of the inserts treated with insecticide was examined using standard 3-min exposure bioassays.

Results: Experiments with open eave tubes showed that a high percentage of released mosquitoes entered the house through tubes during experimental nights. When tubes were fitted with bendiocarb- or deltamethrin-treated inserts, on average 21% [95% CI 18-25%] and 39% [CI 26-51%] of An. gambiae s.s. were recaptured the following morning, respectively. This contrasts with 71% [CI 60-81%] in the treatment with open eaves and 54% [CI 47-61%] in the treatment where inserts were treated with fluorescent dye powder. For An. arabiensis recapture was 21% [CI 14-27%] and 22% [CI 18-25%], respectively, compared to 46% [CI 40-52%] and 25% [CI 15-35%] in the treatments with open tubes and fluorescent dye.

Conclusions: Insecticide-treated eave tubes resulted in significant reductions in recapture rates for both malaria vector species, representing the first and promising results with this novel control tool against Kenyan malaria vectors. Further field evaluation of eave tubes under more realistic field conditions, as well as their comparison with existing approaches in terms of cost-effectiveness and community acceptance, is called for.

Keywords: Anopheles arabiensis; Anopheles gambiae; Eave tubes; House improvement; Kenya; Semi-field system.

Figures

Fig. 1

a Experimental house (3 × 3 m) inside screen-house with mud wall, corrugated iron roofing, screened window and door. b Inside the house with untreated bed net, sealed eaves and eave tubes. c Eave tube with treated insert as seen from the outside

Fig. 2

Eave tube prototypes. Originally, electrostatic netting was fitted over the PVC pipe using a rubber or PVC ring (a–c). Subsequently a second generation of tubes with special inserts (d–f and g–i) was developed. Unfortunately these inserts were too close to the outside of the house, which resulted in the development of an eave tube insert that can slide inside the PVC pipe (j), can easily be stacked (k), and is slightly conical to fit in different diameter tubes (l). 250 of these eave tube inserts (l) can be packed in a box of 60 × 40 × 40 cm

Fig. 3

Mortality of An. gambiae s.s. 24 h after a 1-min exposure to control or insecticide-treated netting at 18 or 27 °C. While deltamethrin killed all mosquitoes regardless of temperature, bendiocarb was significantly less lethal at 18 than at 27 °C (p < 0.001)

Fig. 4

House entry by mosquitoes through open eave tubes. For An. gambiae s.s., 92% of the released mosquitoes were retrieved indoors (light blue backpack aspirator, dark blue CDC light trap, the rest outdoors (white backpack aspirator). For An. arabiensis, indoor captures totalled 76%

Fig. 5

Temperature and relative humidity inside (squares) and outside (diamonds) the experimental house during the experimental period. Measurements of temperature and humidity were taken at 30 min intervals. Environmental data was collected between 19:00 and 7:00 h. Horizontal stripes represent the average per experimental night