A closer look at the WHO cone bioassay: video analysis of the hidden effects of a human host on mosquito behaviour and insecticide contact

Abstract Background The WHO cone test is one of three tests currently used to evaluate the efficacy of insecticide-treated bed nets (ITNs). It generates two test outputs, knockdown and 24-h mortality, both indicative of immediate toxicity but that reveal little about the nature of mosquito and ITN interaction or how results translate to real-world settings. Methods A human arm held 5 mm behind the net surface acted as a host attractant during cone tests and a smartphone was used to capture mosquito behaviour in the cone. Post-exposure blood feeding and survival for nine days were recorded; ingested blood meal size was determined by measuring excreted haematin. Four strains of Anopheles gambiae (insecticide susceptible: Kisumu and N’gousso; insecticide resistant: Banfora and VK7) were tested with and without the host attractant using untreated, Permanet 2.0 and Olyset nets. Video recordings were scan sampled every five seconds to record mosquito positions on either the net, in flight or in contact with the cone. Generalized estimating equations were used to analyse all data except survival within nine days which was analysed using Weighted Cox Regression. Results Net contact was the most frequently recorded behaviour in all Anopheles spp. strains on all nets. Adding the human host as attractant triggered excitatory behaviours: in all strains, the magnitude of net contact was significantly decreased compared to tests without a host. ITN exposure altered the observed behaviour of the two susceptible strains, which exhibited a decreased response to the host during ITN tests. The resistant strains did not alter their behaviour during ITN tests. Significantly less net contact was observed during Olyset Net tests compared to Permanet 2.0. The host presence affected survival after exposure: Banfora and VK7 mosquitoes exposed to Permanet 2.0 with a host lived longer compared to tests performed without a host. However, mosquitoes that blood-fed and survived long enough to digest the blood meal did not exhibit significantly reduced longevity regardless of the presence of the host attractant. Conclusions Simple modifications to the WHO cone test and extension of post-test monitoring beyond the current 24 h enable detailed behavioural characterizations of individual ITNs to be compiled. The effects observed from testing with a host and including blood feeding suggest that more representative estimates of true of ITN efficacy are gained with these modifications than when using the current testing protocol.