Table.4 Outcomes of included interventions

Barreaux et al. [19]

– Two huts with insecticide-treated inserts (In2Care tubes) placed inside of eaves or an untreated insert placed into the eaves

– All participants slept under a LLIN net and windows were open

– All mosquitoes that entered were collected the next day. Rotated treated and untreated inserts between both huts for 24 nights

Number of mosquitoes that enter the household and mortality differences between huts

Linear mixed model to assess the number of captured mosquitoes with the insert considered as an independent variable

– Inserts treated with insecticide resulted in 46% fewer Anopheles gambiae mosquitoes entering the hut vs those that were untreated

– Treated inserts also led to a 64% reduction in blood-feeding by mosquitoes that did make it through

– Mosquitoes that entered households protected by treated eaves died more than the other hut, but this rate was not statistically significant

– A statistical difference was found between nights of capture

– Used experimental houses. Released mosquitoes were reared and not wild mosquitoes which may introduce some confounding in behavior

– Insecticide treated eaves may wither in potency over time due to collection of dust or nature of chemical

– Mosquitoes used were only 4–5 days old, female, non-blood fed and were reared from larvae collected in the field. May have differed from other mosquitoes

Barreaux et al. [12]

Two huts, one with insecticide treated eaves (using the In2Care tool that the WHO recently finished testing in 2019) and closed windows and the other with nontreated, open windows. Monitored mosquitoes for 2 or 4 days

Number of mosquitoes that entered the hut and the mortality between the huts

Linear mixed models

– 0–0.4% of mosquitoes entered through the insecticide-treated eaves with closed windows

– Around 50–80% of mosquitoes entered the control hut (untreated eaves and open windows)

– Treatment hut had 25% more deaths per day compared to the control hut with 2–4% deaths per day

– Used experimental houses

– Released mosquitoes were reared and not wild mosquitoes which may introduce some confoundings in behavior and wither in potency due to collection of dust

Chinula et al. [17]

– WHO wire balls with insecticidal efficacy were placed on netting each having polyacrylate-binding agents (BA) or not having BA

– In one experiment, WSEB netting received 0, 1, or 2 g/sq m of micro-encapsulated PM

– In other experiments, all netting received 1 or 2 g/sq PM with BA or without BA

– Netting was wrapped around wire balls and mosquitoes were exposed to them for 3 min. Tested after 0, 5, 10, 15, 20 washes to see mosquito mortality

Insecticidal capabilities measured through mortality of mosquitoes

General linear mixed modeling treating experimental huts as random effects and mortality as a dependent variable

– WSEBs with PM and BA produced results similar to LLINs–killed mosquitoes at or greater than 94% after 20 total washes. This was the same result when the PM and IRS were used together. This treatment mechanism could cause WSEBs to last for years. Without the BA treatment, all mosquitoes died at 0 washes but after 10 washes, only 10% of mosquitoes died

– WSEBs without PM had fewer than 5% dying consistently throughout 0, 5, 10, 15, and 20 washes with or without BA

– Performed in experimental huts. Prototype design may not be practical for all household designs and should be affordable

– This prototype could not be used in real houses and only investigated polyester netting

– Did not test resistance to environmental exposure such as wind, rain, or sunlight

– Does not provide enough information as to how to provide major scaling up of this intervention

– Unable to effectively test durability

Gouissi et al. [15]

Installed PSP Olyset nets at windows and at eaves

Measured plasmodic index, gametocyte index, parasite density, fever, hemoglobin, and anemia

Statistica 6 and performed Chi-square tests

– More prevalent cases of anemia in the control zones compared to the treatment zones (not significant). Gametocyte index did not change

– Parasite density was 2 × lower in areas within the treatment zone

Risk of fever was similar between the two groups

– Only performed on children. No delineation between outcomes for specifically eaves vs windows.

– Treatments were also applied to doors

Kampango et al. [20]

– Researchers covered gable ends with either a four-year-old bed net, untreated shade cloth, a deltamethrin-treated shade cloth, or nothing at all

– Covered the gable ends with the same material during the second week and covered the eaves with the same material during the third week

Rates of mosquito entry and the efficacy of the three different materials

Rates of mosquito entry were measured using light-trap collection. The material’s ability to protect inhabitants was measured using incidence rate ratio calculations through Generalized Estimating Equations

– Anopheles funestus mosquitoes did not differentially enter households that were protected using deltamethrin-treated shade cloths compared to those that were untreated completely (no shade cloth)

– Anopheles gambiae mosquitoes did differentially enter untreated houses compared to deltamethrin-treated houses. The latter had statistically lower rates of mosquito entry by 76%

– Households protected with bednetting encountered 84% Anopheles gambiae s.l. entry and 61.3% Anopheles funestus entry

– Households protected with untreated shade cloth experienced 70% fewer Anopheles funestus entry and 69% fewer Anopheles gambiae entry

– Small sample size due to being a pilot study

– Assumes that houses do not have any other openings. Resistance status of An. gambiae mosquitoes was not known

Menger et al. [21]

– Researchers used an attractant made of ammonia, lactic acid, tetradecanoic acid, and other compounds found in human skin (pull).

– They also used a repellent called delta-undecalactone and added it to cotton netting placed on eaves (push).

– Treated and untreated fabrics were used eight times with four replicates/day individually over 2 days.

– Four interventions used included a control, a treated push fabric in eaves, an untreated pull fabric in eaves, and a treated push–pull intervention in eaves

Rates of mosquito entry and mosquito landing on fabric

SPSS and a Shapiro–Wilk test for normality. CDC trap catches and t-tests were used for determining house entry and significant drops in mosquito entry. GLM with Dunnet's post-hoc test was used to find the mean trap catches of the four interventions

With the repellent, mosquito entry dropped by over 50% and CDC traps caught several mosquitoes. Modeling suggests that adding the push–pull system to current malaria interventions would help to reduce mosquito biting rate by 20 × even in cases of insecticidal resistance

– Only performed on male volunteers with small sample size. Mosquitoes were lab-reared. The mathematical model may have failed to account for regional variation. Mathematical model did

not include control groups

– Effects of washing materials were not thoroughly considered in the study

Menger et al. [22]

Note, only experiment II in this study used repellents. Experiment II: four interventions used as push factors were: (a) eaves screened with cotton fabric impregnated with dUDL repellent; (b) control; (c) eaves treated with sprayed PMD; (d) eaves screened with untreated fabric. CDC light traps were used as the pull factor (attractant)

Mosquito entry

SPSS to find differences in mosquito house entry. Scheffe's post-hoc tests for correction

– All interventions with eave screen treatments reduced Anopheles funestus house entry by a significant amount

– Eave screening that included PMD led to a 81% drop in house entry. MM-X traps combined with eaves treated with dUDL led to a house entry drop of 35%; combined with PMD-treated eaves led to a 80% drop

– PMD treated eaves had the highest drop in mosquito entry

– For Anopheles arabiensis mosquitoes, PMD-treated eaves dropped mosquito house entry by 89%. When combined with an MM-X trap, house entry dropped by 80%, which was the only significant drop compared to the other interventions

– Although PMD-eaves dropped Culex house entry by 84%, this was not significant

– Small sample size. Simultaneous comparison not possible in the study

– Minimal analysis of the effects of resistance, and net washing

Mmbando et al. [18]

– Transfluthrin eave ribbons and odour-baited traps were tested separately and together to study their protective efficacy

– Participants rested outdoors from 6:30 pm to 10:00 pm during biting activity periods and slept under a bednet from 10:00 pm to 6:30 am while light traps captured mosquitoes

Protective efficacy against biting either outdoors or indoors

Percentage of biting that took place and changes in biting patterns depending on the distance of these interventions

– The treated eave ribbon and light-traps together protected better from indoor-biting (83.4% protection) and outdoor-biting (79%) compared to solely using the traps to protect against indoor-biting (35.0%) and outdoor biting (31%).

– Traps set 15 m away from huts were better at protecting individuals in combination with the treated eave ribbons against indoor-biting

– Traps that were set 30 m away had better efficacy for outdoor-biting. Participants were found to have overall positive outlooks toward prototype windows, particularly the concertina windows, because of their ventilation and attraction

– Using both the treated eaves and the light-traps is complex

– Study samples are limited to one type of mosquito and are based on lab-reared mosquitoes. Wild mosquitoes may behave differently. Field data lacking other factors, such as airflow, in the push–pull system.

– This study could not be applied to younger populations since it only included adult volunteers

Mmbando et al. [10]

– Eave ribbons were treated with 0.02%, 0.2%, 1.5%, or 5% transfluthrin emulsion repellent and tested in two huts against 500 mosquitoes released every night.

– Volunteers stayed outside each hut to collect mosquitoes from 6 to 10:00 pm and slept indoors under bed nets from 10 to 6:30 am

Indoor-biting, outdoor-biting, and mosquito survival

Prokopack aspirators to measure indoor-biting. Double-net traps to measure outdoor-biting. Survival was measured by counting the number of living mosquitoes (n = 200) in a suspended cage next to huts 24 h after exposure

– Volunteers protected by eaves with 0.2% transfluthrin experienced 99% lower indoor-biting and outdoor biting

– Volunteers protected by eaves with 0.02% transfluthrin experienced 79% fewer bites indoors and 60% fewer bites outdoors

– Households with no treatment resulted in 27% fewer indoor-biting and 18% greater outdoor-biting at a non-significant rate

– 99.5% of mosquitoes died when exposed to treated huts

– Field experiments showed that treated eaves protect at 96% indoors and 84% outdoors when tested against Anopheles arabiensis. These treated eaves also protect at 42% indoors and 40% outdoors when tested against Anopheles funestus

– Mosquitoes released nightly were laboratory-reared and may not behave completely as wild mosquitoes do

– Only two volunteers were used throughout the study

– Colder temperatures (below room temperature) inhibit transfluthrin from vaporizing, which is essential for its repellency activities

– Eave ribbons treated with transfluthrin cost $7 for construction and installation

– Mosquitoes released nightly were laboratory-reared and may not behave completely as wild mosquitoes do

– Only two volunteers were used throughout the study

– Colder temperatures (below room temperature) inhibit transfluthrin from vaporizing, which is essential for its repellency activities. The exact amount of transfluthrin adsorbed was not known

– In the semi-field experiment, were only An. arabiensis; especially pertinent because the field experiment had different rates for different mosquitoes

– Did not take effects of washing into consideration

– Did not compare with other types of treatment for nets

Mwanga et al. [11]

– Volunteers stayed close to 5 huts protected with eave ribbons treated with 0.25 g/m2 transfluthrin at 0, 20, 40, 60, 80, and 100% coverage

– All volunteers collected biting mosquitoes between 6 and 10 pm and slept under bed nets (nontreated) from 10 pm to 6 am

– Mosquitoes were caged inside huts for 24 h. Huts were protected with either eave ribbons, UV-LED mosquito traps, or both, with the central hut having no insecticide-treated net to measure effects on households that do not have a net

Protection efficacy of transfluthrin eave ribbons both indoors and outdoors. Mosquito mortality against eave ribbons. Protective efficacy of combining UV-LED traps with transfluthrin eave ribbons

Collecting mosquitoes and counting remaining living ones. Also analyzed data using Generalized Linear Mixed Models

– Eave ribbons treated with transfluthrin ultimately protected volunteers at an efficacy rate of 83% indoors and 62% outdoors for volunteers who slept under a net

– Volunteers who did not use nets were protected by 57% indoors and 48% outdoors

– Volunteers who used nets had constant protection when eave ribbons were added and volunteers who did not use nets had increased protection when eave ribbons were added

– When 80% of households were using treated eaves, the protection to non-users was at its peak

– All caged mosquitoes placed inside huts with treated eave ribbons died

– UV-LED traps added with eave ribbons did not improve overall protection among those who used bed nets

– Took place in semi-field settings and used lab-reared mosquitoes; may, therefore, not be representative of actual conditions

– Volunteers who slept under a net slept under untreated nets, which is unrealistic since most distributed nets are treated with some form of insecticide

Oumbouke et al. [16]

Eleven powder formulations from six insecticide groups (pyrethroid, carbamate, organophosphate, neonicotinoid, entomopathogenic fungus, and boric acid) were tested against pyrethroid resistant Anopheles gambiae s.l. mosquitoes for four weeks

The insecticide of all 11 formulations that had the best residual efficacy, mortality data over its various transient exposures to mosquitoes, and mosquito mortality from that insecticide

Measured contact with eaves using fluorescent powder

– After four weeks, mosquito mortality dropped by 25% for the majority of all 11 insecticides even though they killed mosquitoes at 45–100% during the initial 2 weeks

– Beta-cyfluthrin remained potent (100% mortality) for one month

– Only < 5% of mosquitoes that came into contact with untreated eaves died. About 55% of mosquitoes that came into contact with beta-cyfluthrin treated eaves died the next morning and 64% were dead after 24 h of exposure

– An average of 44% of mosquitoes came into contact with the treated eaves and 75% of mosquitoes passed through open, untreated eave tubes to enter households

– Limited in the number of insecticides this study tested and took place in experimental huts that may not represent real conditions

– Mosquitoes were lab-reared

– Only two volunteers were used in the experimental huts, which is a considerably small sample size.

– Only Anopheles gambiae mosquitoes used; may have different results with different mosquito types

– Did not evaluate long-term properties of all insecticides, and was not able to entirely identify what made beta-cyfluthrin the most successful

Snetselaar et al. [13]

– In two houses, eave tubes treated with bendiocarb or deltamethrin were implanted to test against Anopheles gambiae and Anopheles arabiensis mosquitoes through 3 min bioassays

– Same experiment took place with untreated eave tubes (open)

The number of mosquitoes that enter each household (recapture)

Collected using record sheets and calculated by counting the number of mosquitoes retrieved. Normality was measured using the Shapiro–Wilk test and Levene's test with Mann–Whitney U Test and Bonferroni correction

– Households protected with bendiocarb-treated eaves had a recapture percentage of 21% on average for both Anopheles gambiae s.s. and Anopheles arabiensis mosquitoes

– Those treated with deltamethrin had a recapture percentage of 39% on average for Anopheles gambiae mosquitoes and 22% for Anopheles arabiensis mosquitoes

– Untreated eaves had 71% and 54% recapture rates for households only protected with fluorescent dye powder for Anopheles gambiae mosquitoes and 46% and 25% for Anopheles arabiensis mosquitoes

Was conducted in semi-field conditions with lab-reared mosquitoes. Assessed effects over several weeks, but longer term effects not evaluated in the study

Sternberg et al. [14]

– Eave tubes of various heights, diameters, angles were treated with different ingredients (bendiocarb, LLIN fabric, and fungus) applied on netting

– These were tested on compartments made up of a hut and one sleeper, against open eaves, and again in a larger compartment that consisted of a mosquito population, vegetation, 2+ houses, and cattle sheds meant to represent a village. There, they brought in LLINs and then brought in eave tubes

Mortality rates between applied active ingredients. Recapture rates from treated netting. Larval densities based on LLIN introduction and larval densities based on installation of eave tubes and screening

Generalized linear modeling using quasi binomial error distributions

– Bendiocarb and LLIN netting had equal effects on mosquito mortality rates. Deltamethrin and bendiocarb applied to eaves led to 99–100% mortality of all A. gambiae s.s. mosquitoes, whereas 39% of A. arabiensis mosquitoes died under bendiocarb treatment

– Treated netting resulted in 50–70% lower recapture rates compared to controls. Closed eaves (specifically ones treated with bendiocarb netting) performed similar to LLIN and open eaves together

– Within the constructed village, LLINs produced a 60% drop in larval densities and 85% drop in indoor mosquitoes

– Only performed using adult human volunteers