Temperature, mosquito feeding status and mosquito density influence the measured bio-efficacy of insecticide-treated nets in cone assays.

BACKGROUND: The WHO cone bioassay is routinely used to evaluate the bioefficacy of insecticide-treated nets (ITNs) for product pre-qualification and confirmation of continued ITN performance during operational monitoring. Despite its standardized nature, variability is often observed between tests. We investigated the influence of temperature in the testing environment, mosquito feeding status and mosquito density on cone bioassay results. METHODS: Cone bioassays were conducted on MAGNet (alphacypermethrin) and Veeralin (alphacypermethrin and piperonyl butoxide (PBO)) ITNs, using laboratory-reared pyrethroid-resistant Anopheles funestus sensu stricto (FUMOZ strain) mosquitoes. Three experiments were conducted using standard cone bioassays following WHO-recommended test parameters, with one variable changed in each bioassay: (i) environmental temperature during exposure: 22-23 °C, 26-27 °C, 29-30 °C and 32-33 °C; (ii) feeding regimen before exposure: sugar starved for 6 h, blood-fed or sugar-fed; and (iii) mosquito density per cone: 5, 10, 15 and 20 mosquitoes. For each test, 15 net samples per treatment arm were tested with four cones per sample (N = 60). Mortality after 24, 48 and 72 h post-exposure to ITNs was recorded. RESULTS: There was a notable influence of temperature, feeding status and mosquito density on An. funestus mortality for both types of ITNs. Mortality at 24 h post-exposure was significantly higher at 32-33 °C than at 26-27 °C for both the MAGNet [19.33% vs 7%; odds ratio (OR): 3.96, 95% confidence interval (CI): 1.99-7.87, P < 0.001] and Veeralin (91% vs 47.33%; OR: 22.20, 95% CI: 11.45-43.05, P < 0.001) ITNs. Mosquito feeding status influenced the observed mortality. Relative to sugar-fed mosquitoes, The MAGNet ITNs induced higher mortality among blood-fed mosquitoes (7% vs 3%; OR: 2.23, 95% CI: 0.94-5.27, P = 0.068) and significantly higher mortality among starved mosquitoes (8% vs 3%, OR: 2.88, 95% CI: 1.25-6.63, P = 0.013); in comparison, the Veeralin ITNs showed significantly lower mortality among blood-fed mosquitoes (43% vs 57%; OR: 0.56, 95% CI: 0.38-0.81, P = 0.002) and no difference for starved mosquitoes (58% vs 57%; OR: 1.05, 95% CI: 0.72-1.51, P = 0.816). Mortality significantly increased with increasing mosquito density for both the MAGNet (e.g. 5 vs 10 mosquitoes: 7% vs 12%; OR: 1.81, 95% CI: 1.03-3.20, P = 0.040) and Veeralin (e.g. 5 vs 10 mosquitoes: 58% vs 71%; OR 2.06, 95% CI: 1.24-3.42, P = 0.005) ITNs. CONCLUSIONS: The results of this study highlight that the testing parameters temperature, feeding status and mosquito density significantly influence the mortality measured in cone bioassays. Careful adherence to testing parameters outlined in WHO ITN testing guidelines will likely improve the repeatability of studies within and between product testing facilities.

Time of exposure and assessment influence the mortality induced by insecticides against metabolic resistant mosquitoes.

BACKGROUND: Increasing metabolic resistance in malaria vector mosquitoes resulted in the development of insecticide-treated nets (ITNs) with active ingredients (AI) that target them. Bioassays that accurately measure the mortality induced by these AIs on ITNs are needed. Mosquito metabolic enzyme expression follows a circadian rhythm. Thus, this study assessed (i) influence of the time of day of mosquito exposure and (ii) timing of assessment of mortality post exposure (24 and 72 h) to ITNs against vectors that are susceptible to pyrethroids and those with metabolic and knockdown resistance mechanisms. METHODS: Two cone bioassay experiments were conducted following World Health Organization (WHO) guidelines. Firstly, on ITNs incorporated with 2 g AI/kg of deltamethrin (DM) alone, or combined with 8 g AI/kg piperonyl butoxide (PBO) synergist, during the day (9:00-14:00 h) and repeated in the evening (18:00-20:00 h). This was followed by a confirmatory experiment during the afternoon (12:00-14:00 h) and repeated in the night (22:00-24:00 h) using mosquitoes unexposed or pre-exposed to PBO for 1 h before exposure to DM ITNs. Each net piece was tested with a minimum of eight cones per time (N = 24). The outcome was mortality after 24 h (M24) or 72 h (M72) of holding. RESULTS: The cone bioassays performed using metabolic resistant mosquitoes during the evening showed significantly lower M24 than those performed in the day for DM: odds ratio (OR) 0.14 [95% confidence interval (CI) 0.06-0.30, p < 0.0001] and DM PBO [OR 0.29 (95% CI 0.18-0.49, p < 0.0001). M72 was higher than M24 for metabolic resistant mosquitoes exposed to DM [OR 1.44 (95% CI 1.09-1.88), p = 0.009] and DM PBO [OR 1.82 (95% CI 1.42-2.34), p < 0.0001]. An influence of hour of experiment and time of assessment was not observed for mosquitoes that had knockdown resistance or that were pyrethroid-susceptible. CONCLUSIONS: Time of day of experiment and hour of assessment of delayed mortality after exposure of mosquitoes are important considerations in evaluating insecticides that interact with mosquito metabolism to counter metabolic resistant mosquitoes. This is important when evaluating field-aged ITNs that may have lower concentrations of AI.

Efficacy of the spatial repellent product Mosquito Shield™ against wild pyrethroid-resistant Anopheles arabiensis in south-eastern Tanzania.

BACKGROUND: Spatial repellents that create airborne concentrations of an active ingredient (AI) within a space offer a scalable solution to further reduce transmission of malaria, by disrupting mosquito behaviours in ways that ultimately lead to reduced human-vector contact. Passive emanator spatial repellents can protect multiple people within the treated space and can last for multiple weeks without the need for daily user touchpoints, making them less intrusive interventions. They may be particularly advantageous in certain use cases where implementation of core tools may be constrained, such as in humanitarian emergencies and among mobile at-risk populations. The purpose of this study was to assess the efficacy of Mosquito Shield™ deployed in experimental huts against wild, free-flying, pyrethroid-resistant Anopheles arabiensis mosquitoes in Tanzania over 1 month. METHODS: The efficacy of Mosquito Shield™ transfluthrin spatial repellent in reducing mosquito lands and blood-feeding was evaluated using 24 huts: sixteen huts were allocated to Human Landing Catch (HLC) collections and eight huts to estimating blood-feeding. In both experiments, half of the huts received no intervention (control) while the remaining received the intervention randomly allocated to huts and remained fixed for the study duration. Outcomes measured were mosquito landings, blood-fed, resting and dead mosquitoes. Data were analysed by multilevel mixed effects regression with appropriate dispersion and link function accounting for volunteer, hut and day. RESULTS: Landing inhibition was estimated to be 70% (57-78%) [IRR 0.30 (95% CI 0.22-0.43); p < 0.0001] and blood-feeding inhibition was estimated to be 69% (56-79%) [IRR 0.31 (95% CI 0.21-0.44; p < 0.0001] There was no difference in the protective efficacy estimates of landing and blood-feeding inhibition [IRR 0.98 (95% CI 0.53-1.82; p = 0.958]. CONCLUSIONS: This study demonstrated that Mosquito Shield™ was efficacious against a wild pyrethroid-resistant strain of An. arabiensis mosquitoes in Tanzania for up to 1 month and could be used as a complementary or stand-alone tool where gaps in protection offered by core malaria vector control tools exist. HLC is a suitable technique for estimating bite reductions conferred by spatial repellents especially where direct blood-feeding measurements are not practical or are ethically limited.

The MTego trap: a potential tool for monitoring malaria and arbovirus vectors.

BACKGROUND: Odour-baited traps are useful for vector surveillance and control. However, most existing traps have shown inconsistent recapture rates across different mosquito species, necessitating the need for more effective and efficient traps. The MTego trap with integrated thermal stimuli has been developed as an alternative trap. This study was undertaken to determine and compare the efficacy of the MTego trap to that of the Biogents (BG) modular BG-Pro (BGP) trap for sampling different mosquito species in a semi-field system. METHODS: Fully balanced Latin square design experiments (no-choice and dual choice) were conducted in semi-field chambers using laboratory-reared female Anopheles gambiae sensu stricto, Anopheles funestus, Anopheles arabiensis, Culex quinquefasciatus and Aedes aegypti. There were 16 replicates, and 50 mosquitoes of each species were released in each chamber per replicate. The evaluated traps were as follows: the MTego trap baited with PM6 (MT-PM6), the MTego trap baited with BG-Lure (BGL) (MT-BGL), and the BGP trap baited with BG-Lure (BGP-BGL). RESULTS: In the no-choice test, the MT-BGL and BGP-BGL traps captured a similar proportion of An. gambiae (31% vs 29%, P-value = 0.519) and An. funestus (32% vs 33%, P = 0.520). The MT-PM6 and BGP-BGL traps showed no significant difference in capturing Ae. aegypti (33% vs 31%, P = 0.324). However, the BGP-BGL caught more An. arabiensis and Cx. quinquefasciatus mosquitoes than the other traps (P < 0.0001). In the dual-choice test of MT-PM6 vs BGP-BGL, similar proportions of An. funestus (25% vs 27%, P = 0.473) and Ae. aegypti (29% vs 25%, P = 0.264) were captured in the traps, while the BGP-BGL captured more An. gambiae, An. arabiensis and Cx. quinquefasciatus mosquitoes than the MT-PM6 (P < 0.0001). CONCLUSIONS: This study demonstrated that the MTego trap has potential as a tool that can be used interchangeably with the BGP trap for sampling anthropophilic mosquitoes including African malaria vectors An. gambiae and An. funestus and the principal arbovirus vector Ae. aegypti.

CDC light traps underestimate the protective efficacy of an indoor spatial repellent against bites from wild Anopheles arabiensis mosquitoes in Tanzania.

BACKGROUND: Methods for evaluating efficacy of core malaria interventions in experimental and operational settings are well established but gaps exist for spatial repellents (SR). The objective of this study was to compare three different techniques: (1) collection of blood-fed mosquitoes (feeding), (2) human landing catch (HLC), and (3) CDC light trap (CDC-LT) collections for measuring the indoor protective efficacy (PE) of the volatile pyrethroid SR product Mosquito Shield™ METHODS: The PE of Mosquito Shield™ against a wild population of pyrethroid-resistant Anopheles arabiensis mosquitoes was determined via feeding, HLC, or CDC-LT using four simultaneous 3 by 3 Latin squares (LS) run using 12 experimental huts in Tanzania. On any given night each technique was assigned to two huts with control and two huts with treatment. The LS were run twice over 18 nights to give a sample size of 72 replicates for each technique. Data were analysed by negative binomial regression. RESULTS: The PE of Mosquito Shield™ measured as feeding inhibition was 84% (95% confidence interval (CI) 58-94% [Incidence Rate Ratio (IRR) 0.16 (0.06-0.42), p < 0.001]; landing inhibition 77% [64-86%, (IRR 0.23 (0.14-0.36) p < 0.001]; and reduction in numbers collected by CDC-LT 30% (0-56%) [IRR 0.70 (0.44-1.0) p = 0.160]. Analysis of the agreement of the PE measured by each technique relative to HLC indicated no statistical difference in PE measured by feeding inhibition and landing inhibition [IRR 0.73 (0.25-2.12) p = 0.568], but a significant difference in PE measured by CDC-LT and landing inhibition [IRR 3.13 (1.57-6.26) p = 0.001]. CONCLUSION: HLC gave a similar estimate of PE of Mosquito Shield™ against An. arabiensis mosquitoes when compared to measuring blood-feeding directly, while CDC-LT underestimated PE relative to the other techniques. The results of this study indicate that CDC-LT could not effectively estimate PE of the indoor spatial repellent in this setting. It is critical to first evaluate the use of CDC-LT (and other tools) in local settings prior to their use in entomological studies when evaluating the impact of indoor SR to ensure that they reflect the true PE of the intervention.

Human landing catches provide a useful measure of protective efficacy for the evaluation of volatile pyrethroid spatial repellents.

BACKGROUND: The human landing catch (HLC) method, in which human volunteers collect mosquitoes that land on them before they can bite, is used to quantify human exposure to mosquito vectors of disease. Comparing HLCs in the presence and absence of interventions such as repellents is often used to measure protective efficacy (PE). Some repellents have multiple actions, including feeding inhibition, whereby mosquitoes may be unable to bite even if they land on a host. A comparison was made between the PE of the volatile pyrethroid spatial repellent (VPSR) transfluthrin determined using a landing method (HLC) and a biting method (allowing the mosquitoes that landed to blood-feed) to evaluate whether HLC is a suitable method for the estimation of the personal PE of a VPSR. METHODS: A fully balanced, two-arm crossover design study was conducted using a 6 × 6 × 2-m netted cage within a semi-field system. Hessian strips (4 m × 0.1 m) treated with a 5-, 10-, 15-, or 20-g dose of transfluthrin were evaluated against a paired negative control for three strains of laboratory-reared Anopheles and Aedes aegypti mosquitoes. Six replicates were performed per dose using either the landing or the biting method. The number of recaptured mosquitoes was analysed by negative binomial regression, and the PEs calculated using the two methods were compared by Bland-Altman plots. RESULTS: For Anopheles, fewer mosquitoes blood-fed in the biting arm than landed in the landing arm (incidence rate ratio = 0.87, 95% confidence interval 0.81-0.93, P < 0.001). For Ae. aegypti, biting was overestimated by around 37% with the landing method (incidence rate ratio = 0.63, 95% confidence interval 0.57-0.70, P = 0.001). However, the PEs calculated for each method were in close agreement when tested by the Bland Altman plot. CONCLUSIONS: The HLC method led to underestimation of mosquito feeding inhibition as a mode of action of transfluthrin, and there were species- and dose-dependent differences in the relationship between landing and biting. However, the estimated PEs were similar between the two methods. The results of this study indicate that HLC can be used as a proxy for personal PE for the evaluation of a VPSR, especially when the difficulties associated with enumerating blood-fed mosquitoes in a field setting are taken into consideration.

Insecticide-treated eave nets and window screens for malaria control in Chalinze district, Tanzania: a study protocol for a household randomised control trial.

BACKGROUND: Long-lasting insecticidal nets (LLINs) have contributed to the reduction of malaria in sub-Saharan Africa, including Tanzania. However, they rely on daily user behaviour and high coverage which is difficult to maintain. Also, insecticide resistance among malaria vector mosquitoes is contributing to reduced efficacy of control tools. To overcome these problems, we propose to evaluate a new tool for house modification, the insecticide-treated eave nets (ITENs) in combination with insecticide-treated window screens (ITWS) incorporated with dual active ingredient (dual AI) for the control of malaria. METHODS: Four hundred and fifty (450) households with intact walls, open eaves without screens or nets on the windows in Chalinze district will be eligible and recruited upon written informed consent. The households will be randomly allocated into two arms: one with ITENs and ITWS installed and the other without. Malaria parasite detection using a quantitative polymerase chain reaction (qPCR) will be conducted shortly after the long rain (June/July, 2022) as the primary outcome and shortly after the short rain (January/February, 2022) as the secondary outcome. Other secondary outcomes include clinical malaria cases, and density of malaria vectors and nuisance after the short rain and long rain. In addition, surveys will be conducted in households with ITENs and ITWS to estimate the intervention's cost during installation, adverse effects one month after installation, and presence, fabric integrity and user acceptance six and twelve months after installation. Bioefficacy and chemical content will be evaluated twelve months after installation. DISCUSSION: ITENs and ITWS have been shown in Kenya to reduce indoor mosquito density. However, it is not known if indoor mosquito density reduction translates into reduction of malaria cases. Data from the study will measure the potential public health value of an additional intervention for malaria control at the household level in areas of mosquito insecticide resistance that does not require daily adherence. TRIAL REGISTRATION: The study is registered on ClinicalTrials.gov .

Influence of testing modality on bioefficacy for the evaluation of Interceptor® G2 mosquito nets to combat malaria mosquitoes in Tanzania.

BACKGROUND: Insecticide-treated net (ITN) durability is evaluated using longitudinal bioefficacy and fabric integrity sampling post-distribution. Interceptor® G2 was developed for resistance management and contains two adulticides: alpha-cypermethrin and chlorfenapyr; it is a pro-insecticide that is metabolized into its active form by mosquito-detoxifying enzymes and may be enhanced when the mosquito is physiologically active. To elucidate the impact of bioassay modality, mosquito exposures of the alphacypermethrin ITN Interceptor® and dual adulticide Interceptor® G2 were investigated. METHODS: This study evaluated the performance of Interceptor® G2 compared to Interceptor® against local strains of mosquitoes in Tanzania. Unwashed and 20× times washed nets were tested. Efficacy of ITNs was measured by four bioassay types: (1) World Health Organisation (WHO) cone test (cone), (2) WHO tunnel test (tunnel), (3) Ifakara ambient chamber test (I-ACT) and (4) the WHO gold standard experimental hut test (hut). Hut tests were conducted against free-flying wild pyrethroid metabolically resistant Anopheles arabiensis and Culex quinquefasciatus. Cone, tunnel and I-ACT bioassays used laboratory-reared metabolically resistant An. arabiensis and Cx. quinquefasciatus and pyrethroid susceptible Anopheles gambiae sensu stricto and Aedes aegypti. RESULTS: Against resistant strains, superiority of Interceptor® G2 over Interceptor® was observed in all "free-flying bioassays". In cone tests (which restrict mosquito flight), superiority of Interceptor® over Interceptor® G2 was recorded. Mortality of unwashed Interceptor® G2 among An. arabiensis was lowest in hut tests at 42.9% (95% CI: 37.3-48.5), although this increased to 66.7% (95% CI: 47.1-86.3) by blocking hut exit traps so mosquitoes presumably increased frequencies of contact with ITNs. Higher odds of mortality were consistently observed in Interceptor® G2 compared to Interceptor® in "free-flying" bioassays using An. arabiensis: tunnel (OR = 1.42 [95% CI:1.19-1.70], p < 0.001), I-ACT (OR = 1.61 [95% CI: 1.05-2.49], p = 0.031) and hut (OR = 2.53 [95% CI: 1.96-3.26], p < 0.001). Interceptor® and Interceptor® G2 showed high blood-feeding inhibition against all strains. CONCLUSION: Both free-flying laboratory bioassays (WHO Tunnel and I-ACT) consistently measured similarly, and both predicted the results of the experimental hut test. For bioefficacy monitoring and upstream product evaluation of ITNs in situ, the I-ACT may provide an alternative bioassay modality with improved statistical power. Interceptor G2® outperformed Interceptor ® against pyrethroid-resistant strains, demonstrating the usefulness of chlorfenapyr in mitigation of malaria.

A crossover study to evaluate the diversion of malaria vectors in a community with incomplete coverage of spatial repellents in the Kilombero Valley, Tanzania.

BACKGROUND: Malaria elimination is unlikely to occur if vector control efforts focus entirely on transmission occurring indoors without addressing vectors that bite outdoors and outside sleeping hours. Additional control tools such as spatial repellents may provide the personal protection required to fill this gap. However, since repellents do not kill mosquitoes it is unclear if vectors will be diverted from households that use spatial repellents to those that do not. METHODS: A crossover study was performed over 24 weeks in Kilombero, Tanzania. The density of resting and blood-engorged mosquitoes and human blood index (HBI) of malaria vector species per household was measured among 90 households using or not using 0.03 % transfluthrin coils burned outdoors under three coverage scenarios: (i) no coverage (blank coils); (ii) complete coverage of repellent coils; and (iii) incomplete coverage of repellent and blank coils. Mosquitoes were collected three days a week for 24 weeks from the inside and outside of all participating households using mosquito aspirators. Paired indoor and outdoor human landing collections were performed in three random households for six consecutive nights to confirm repellent efficacy of the coils and local vector biting times. RESULTS: The main vectors were Anopheles arabiensis and Anopheles funestus (sensu stricto), which fed outdoors, outside sleeping hours, on humans as well as animals. Anopheles arabiensis landings were reduced by 80 % by the spatial repellent although household densities were not reduced. The HBI for An. arabiensis was significantly higher among households without repellents in the incomplete coverage scenario compared to houses in the no coverage scenario (Odds ratio 1.71; 95 % CI: 1.04-2.83; P = 0.03). This indicated that An. arabiensis mosquitoes seeking a human blood meal were diverted from repellent users to non-users. The repellent coils did not affect An. funestus densities or HBI. CONCLUSIONS: Substantial malaria vector activity is occurring outside sleeping hours in the Kilombero valley. Repellent coils provided some protection against local An. arabiensis but did not protect against local (and potentially pyrethroid-resistant) An. funestus. Pyrethroid-based spatial repellents may offer a degree of personal protection, however the overall public health benefit is doubtful and potentially iniquitous as their use may divert malaria vectors to those who do not use them.

Evaluating preservation methods for identifying Anopheles gambiae s.s. and Anopheles arabiensis complex mosquitoes species using near infra-red spectroscopy.

BACKGROUND: Near-infrared spectroscopy (NIRS) has been successfully used on fresh and RNAlater-preserved members of the Anopheles gambiae complex to identify sibling species and age. No preservation methods other than using RNAlater have been tested to preserve mosquitoes for species identification using NIRS. However, RNAlater is not the most practical preservative for field settings because it is expensive, requires basic laboratory conditions for storage and is not widely available in sub-Saharan Africa. The aim of this study was to test several cheaper and more field-friendly preservation methods for identifying sibling species of the An. gambiae complex using NIRS. METHODS: In this study we describe the use of NIRS to identify sibling species of preserved An. gambiae s. s. and An. arabiensis. Mosquitoes of each species were placed in sample tubes and preserved using one of the following preservation methods: (i) refrigeration at 4°C, (ii) freezing at -20°C, (iii) drying over a silica-gel desiccant, (iv) submersion in RNAlater at room temperature, (v) submersion in RNAlater at 4°C, and (vi) submersion in RNAlater at -20°C. Mosquitoes were preserved for 1, 4, 10, 32 or 50 weeks before they were scanned. RESULTS: Storage at 4°C was the only preservation method that, up to 32 weeks, did not result in significantly lower predicted values than those obtained from fresh insects. After 50 weeks, however, refrigerated samples did not give meaningful results. When storing for 50 weeks, desiccating samples over silica gel was the best preservation method, with a partial least squares regression cross-validation of >80%. Predictive data values were analyzed using a generalized linear model. CONCLUSION: NIRS can be used to identify species of desiccated Anopheles gambiae s.s. and Anopheles arabiensis for up to 50 weeks of storage with more than 80% accuracy.

Experimental hut evaluation of linalool spatial repellent agar gel against Anopheles gambiae sensu stricto mosquitoes in a semi-field system in Bagamoyo, Tanzania.

BACKGROUND: Malaria vector control is in need of new tools to face its current challenges such as the spread of pyrethroid-resistance and the increase of outdoor feeding mosquitoes. New strategies such as spatial repellents need to be evaluated as supplemental tools to existing control measures such as insecticide treated bed nets and indoor residual spraying. Linalool is a naturally occurring terpene alcohol commonly found in flowers and spices with reportedly repellent properties. METHODS: Four experimental huts fitted with exit traps and enclosed inside a large screened semi-field system were used for the evaluation. The tested spatial repellent product consisted of an agar gel emanator containing 73% linalool. Two rounds of experiments using a Latin square design were conducted to evaluate the efficacy of the linalool emanators compared to no treatment (negative control) and a transfluthrin coil (positive) against lab-reared disease free Anopheles gambiae s.s.. The emanators were hung inside experimental huts where two volunteers were sleeping unprotected. The outcome measures were repellency, % feeding inhibition, %mortality and post 24 h % mortality. RESULTS: Unlike the mosquito coil, the linalool emanators did not show any feeding inhibition, repellency or induced mortality compared to the negative control. On the other hand mosquitoes kept for 24 h post exposure were 3 times more likely to die after being exposed to two 73% linalool emanators than the negative control. CONCLUSIONS: Our results indicate that linalool agar gel emanators are not adequate as a spatial repellent against Anopheles gambiae s.s.. However adding linalool to known repellent formulations could be advantageous, not only because of its pleasant scent but also because of the delayed mortality effect it has on mosquitoes.

Do topical repellents divert mosquitoes within a community? Health equity implications of topical repellents as a mosquito bite prevention tool.

OBJECTIVES: Repellents do not kill mosquitoes--they simply reduce human-vector contact. Thus it is possible that individuals who do not use repellents but dwell close to repellent users experience more bites than otherwise. The objective of this study was to measure if diversion occurs from households that use repellents to those that do not use repellents. METHODS: The study was performed in three Tanzanian villages using 15%-DEET and placebo lotions. All households were given LLINs. Three coverage scenarios were investigated: complete coverage (all households were given 15%-DEET), incomplete coverage (80% of households were given 15%-DEET and 20% placebo) and no coverage (all households were given placebo). A crossover study design was used and coverage scenarios were rotated weekly over a period of ten weeks. The placebo lotion was randomly allocated to households in the incomplete coverage scenario. The level of compliance was reported to be close to 100%. Mosquito densities were measured through aspiration of resting mosquitoes. Data were analysed using negative binomial regression models. FINDINGS: Repellent-users had consistently fewer mosquitoes in their dwellings. In villages where everybody had been given 15%-DEET, resting mosquito densities were fewer than half that of households in the no coverage scenario (Incidence Rate Ratio [IRR]=0.39 (95% confidence interval [CI]: 0.25-0.60); p<0.001). Placebo-users living in a village where 80% of the households used 15%-DEET were likely to have over four-times more mosquitoes (IRR=4.17; 95% CI: 3.08-5.65; p<0.001) resting in their dwellings in comparison to households in a village where nobody uses repellent. CONCLUSIONS: There is evidence that high coverage of repellent use could significantly reduce man-vector contact but with incomplete coverage evidence suggests that mosquitoes are diverted from households that use repellent to those that do not. Therefore, if repellents are to be considered for vector control, strategies to maximise coverage are required.