All methods were performed in accordance with the relevant guidelines and regulations.
Study site
The laboratory experiments on regeneration and wash resistance were conducted at the KCMUCo-PAMVERC Insecticide Testing Facility; while experimental hut study was carried out at Harusini, the facility’s field site located at Mabogini village (S03˚22.764’ E03˚720.793’), adjacent to Lower Moshi rice irrigation scheme in north-eastern Tanzania. The dominant vector at this site is An. arabiensis with moderate level of resistance to pyrethroids conferred by both oxidase and esterase activities32. In this study, pyrethroid-resistant laboratory reared An. gambiae Muleba-Kis mosquitoes were released into the huts for the release-recapture experiment.
Test systems
Non-blood fed, 2–5 day old females of susceptible An. gambiae s.s. Kisumu strain and pyrethroid resistant An. gambiae s.s Muleba-Kis strain were used for the evaluation of efficacy in the laboratory (phase I). The Muleba-Kis strain has been colonized for more than 8 years and it is resistant to permethrin with fixed L1014S kdr frequency and metabolic resistance through increased oxidase activity has also been reported21. Only An. gambiae s.s Muleba-Kis were used in release-recapture experiments. The Kisumu strain is fully susceptible to insecticides and free of any detectable insecticide resistance mechanisms. The strain originated from Kisumu, Kenya and has been colonized for many years in laboratory. At the KCMUCo-PAMVERC Moshi insectary, the adult Kisumu strain mosquitoes are reared at a temperature of 24–27 °C, 75 ± 10% relative humidity (RH) and maintained under a dark:light regime of 12:12 h. The Muleba-Kis mosquitoes used for the release-recapture experiments were reared in the field insectary under ambient temperature and relative humidity and treated as previously explained21. The susceptibility status of these colonies is checked every three months using WHO susceptibility test33 and, CDC bottle bioassay test34. The colonies are regularly genotyped for kdr mutations using TaqMan assays35. To maintain the resistance of Muleba-Kis, larvae are frequently selected with alpha-cypermethrin.
Regeneration time
To determine the regeneration time of the insecticide-treated blankets, blankets were cut into 25 × 25 cm pieces and tested before washing and then washed and dried three times consecutively following WHO recommended procedures for LLINs36. The pieces were then re-tested after one, two, three, six and seven days post-washing using WHO cylinders against susceptible An. gambiae s.s (Kisumu).
Graphs for 24-h mortality and 60 min knock down (KD) correlating to insecticide bioavailability, as measured by 3 min exposure in cylinder bioassays, were established before and after washing blanket pieces three times consecutively in a day, and tested within a maximum of seven days post-washing. The time in days required to reach initial mortality or 60 min KD plateau is the period required for full regeneration of insecticide-treated blanket.
Wash resistance
WHO cylinder bioassays36 were used to assess the wash resistance for the blanket pieces washed 0, 5, 10, 15 and 20 times at the intervals equivalent to the regeneration time. Four pieces cut from 4 permethrin and 4 untreated blankets were used as positive and negative control respectively, against 4 pieces cut from 4 PBO–permethrin blankets.
Bioassay procedures
Five, non-blood fed, 2–5 day old An. gambiae Kisumu or An. gambiae Muleba-Kis mosquitoes were exposed for 3 min or 30 min to blanket pieces in WHO cylinder. Bioassays were carried out at 27 ± 2 °C and 75 ± 10% RH. Knock-down was scored after 60 min post-exposure and mortality after 24 h. Fifty mosquitoes (5 mosquitoes per cylinder) were used on each 25 × 25 cm piece of blanket sample. After exposure, the mosquitoes were held for 24 h with access to 10% glucose solution in the paper cups covered with a net material. Mosquitoes exposed to untreated blanket were referred as a negative control.
WHO tunnel test method
Blanket pieces which recorded ≤ 80% mortality in cylinder bioassay were tested in the tunnel assay using WHO guidelines. The tunnel was made of an acrylic square cylinder (25 cm in height, 25 cm in width, and 60 cm in length) divided into two sections using a blanket-covered frame fitted into a slot across the tunnel. During the assays a guinea pig was held in a small wooden cage (as a bait) in one of the sections and 50, non-blood fed, female An. gambiae Kisumu or An. gambiae Muleba-Kis aged 5–8 days were released in the other section at dusk and left overnight (13 h) for experimentation at 27 ± 2 °C and 75 ± 10% RH. The blanket surface was deliberately holed (nine 1-cm holes) to allow mosquitoes to contact the blanket material and penetrate to the baited chamber. Treated blankets were tested concurrently together with an untreated blanket. Scoring for the numbers of mosquitoes found alive or dead, fed or unfed, in each section were done in the morning. Mosquitoes found alive were removed and held in paper cups with labels corresponding to each tunnel sections under controlled conditions (25–27 °C and 75–85% RH) and fed on 10% glucose solution to monitor for delayed mortality post exposurely. Outcomes recorded were: mosquito penetration, blood feeding and mortality.
Washing of blankets and whole nets for hut trial
Blankets and whole nets were separately washed following WHOPES guidelines. In brief, each blanket/net was washed in Savon de Marseilles soap solution (2 g/L) for 10 min: 3 min stirring, 4 min soaking, then another 3 min stirring. This was followed by 2 rinse cycles of the same duration with water only. The water pH was 6 for all washes. The mean water hardness was within the WHOPES limit of ≤ 89 ppm. All nets used in the experimental hut study were cut with holes (4 cm × 4 cm) to simulate the conditions of a torn net. While nets were washed 20 times as per guidelines, blankets were only washed 10 times. To simulate a situation in emergence situations where washing is less frequent due to water scarcity30,31.
Experimental hut trial:experimental hut design
Experimental hut study was done in Lower Moshi using typical East African experimental huts design as described in the WHOPES35. Huts were constructed with brick walls and featured with cement plaster on the inside and a ceiling board, a metal iron sheet roof, open eaves with window and veranda traps on each side and window traps. Slight modifications from the original structure were made by installing metal eave baffles on two sides. The baffles allow mosquito entry but prevent exits. The window traps were used to collect mosquitoes that tend to exit the huts.
Test item labelling, washing and perforating
Both blankets and LLINs for the trial were distinctively labelled with fabric labels that withstand washes. For wash resistance, the blankets and nets were separately washed according to a protocol adapted from the standard WHO washing procedure36 at the interval equivalent to the regeneration time established in the laboratory for blanket and LLIN respectively. Before testing in the experimental huts, all nets were deliberately holed i.e. 30 holes measuring 4 × 4 cm were made in each net, 9 holes in each of the long side panels, and 6 holes at each short side (head- and foot-side panels) to enhance blood-feeding on the control arm.
Test items packaging
Each blanket and net were sealed in a plastic bag and then packed in the large plastic container. Each container was labelled for a single treatment to avoid cross contamination between test items.
Experimental hut decontamination
A cone assay with 10 susceptible mosquitoes was performed on one wall per hut to rule out any contamination of the wall surface. Only huts with 24 h mortality of susceptible mosquitoes < 10% were used.
Predator check
Experimental hut rooms, verandas and exit traps were searched for ants. Where found, clean petri dishes containing boric acid-sugar mixture (1:1) was used to put down ants. Subsequently, the ant numbers were monitored by using dead mosquitoes as bait. The trials started only when the mosquito baits were still in the huts for 24 h. Rooms and verandas of all huts were checked for spiders and spider webs and the baffles were checked for the presence of lizards.
Leakage check
All veranda walls and screens were checked for cracks and holes respectively and repaired as appropriate. The baffles in the ‘entry eave gap’ sides and the eave gap in the ‘exit eave gap’ sides were checked for uniformity targeting the gap level of 5 ± 2 cm when measured from the inside of the hut. Water moats around the huts were checked for leakage and repaired when found leaky. Exit traps were checked for holes and repaired, cleaned and hung back up. Any gap around the window trap was sealed with clean cotton wool.
Volunteer recruitment, medical care and assessment of adverse events
Six adult male volunteer sleepers were consented, recruited and trained for the study; the volunteers slept individually in the huts every test night from 19:00 to 06:30 h. Before the study, all sleepers underwent a medical checkup for malaria and symptoms for COVID-19. Every Monday and Friday morning before mosquito collection, sleepers were asked if they experience any apparent side effect due to skin contact with treated blankets.
Release-recapture experiment
The efficacy of treated blanket was evaluated against pyrethroid resistant An. gambiae Muleba-Kis which were released in each experimental hut early in the evening and re-captured the next morning. During the 36 nights of collections, a total of 4320 An. gambiae s.l. Muleba-Kis were used for the release-recapture study with a recapture success rate of approximatively 90%. Six trial arms: untreated blankets (washed 10 times), Olyset LLIN (washed 20 times), Treated Blanket & Olyset LLIN roof-treated IG2 (10 times and 20 times washed respectively), Olyset Plus LLIN (washed 20 times), treated blanket (10 times washed) and unwashed treated blanket were tested in separate huts for 36 collection nights from July to August 2021.
Rotation of treatments and sleepers
A full 6 × 6 Latin square rotation schedule was automatically generated37. Treatments were rotated between experimental huts every six days, and sleepers every day to limit any location bias resulting from differences between sleepers, huts and treatments. Three replicates per treatment arm were used and exchanged daily to capture any variation within treatments. The huts were cleaned and left for ventilation every 7th day before the next rotation.
In each hut one volunteer slept every trial night from 19:00 h to 06:30 h the following morning and were instructed to sleep under the net (Olyset or Olyset Plus or untreated) and/or to cover themselves with a blanket (untreated or treated). During each trial night supervisors performed two inspections and scored the level of blanket coverage on sleepers, following the scoring chart adopted from Kitau et al.8.
Mosquito release and collection procedures
Blankets and LLINs were evaluated in the Mabogini experimental huts from June to July 2021, corresponding to 36 nights of collections from 6 huts. In each hut, 20 blood-unfed, 2–5 days old female An. gambiae Muleba-Kis were released inside the hut and collected in the morning. Collection of the mosquitoes started at the negative control hut and proceeded to the single insecticide treatment arms, then to the permethrin–PBO positive control, then lastly PBO–permethrin blanket. The mosquitoes were first collected from the floor, inside the nets, walls, ceiling and exit traps. The time spent on collecting mosquitoes was strictly followed with an acceptable variation of time to prevent collector bias between huts; exit traps (3 min), room floor (5 ± 1 min), room walls and ceiling (5 ± 1 min) and in the net (6 ± 1 min).
Mosquito scoring
Two technicians were involved in scoring, mosquitoes identification and filling the data forms. Mosquitoes were systematically scored by species (Anopheles then Culex), by location (floor, net, wall, ceiling, window trap), gonotrophic and alive/dead status. Alive mosquitoes were placed in small paper cups and provided access to 10% glucose solution. Temperature and RH were also recorded during scoring. Scoring was repeated after 24 h for delayed mortality.
Outcome measures
Each net treatment, blanket treatment or blanket in combination with net was evaluated for blood feeding inhibition and induced mortality as primary outcomes. Repellence was also scored as a secondary outcome. Immediate and delayed mortality were determined as the proportions of total mosquitoes released into the hut and collected dead in the morning (immediate mortality) or when caught alive and die at 24 h. Blood-feeding inhibition was determined as reduction in blood-feeding in treatments relative to the negative control hut. Repellence (induced exophily) was determined as a proportion of mosquitoes collected from the exit traps from treated huts relative to proportion caught in exit traps of negative control hut.
Statistical methods
Statistical analyses were performed using Stata 16 (Stata Corp LP, College Station, TX, USA). Mortality from the cylinder bioassay and tunnel tests were expressed as percentages and 95% confidence intervals was calculated. For the experimental hut trial, data were entered in an excel database and transferred to Stata for processing and analysis. In brief, proportion of blood feeding inhibition, exophily and overall killing effect were calculated. Data were analysed with a linear mixed model that included the adjacent hut treatment as independent variable and negative binomial regression for count data. The night of capture, the hut and the sleeper were considered as fixed effects. The primary criteria in the evaluation were blood-feeding inhibition and mortality, with exophily as a secondary outcome.
Percentage mortality was estimated using the formula
$$left[ {left( {{text{N}}_{{text{d}}} } right){text{/N}}_{{text{t}}} } right], times ,100%$$
where Nd = the number of mosquitoes found dead in hut, Nt = total number of mosquitoes released in hut.
Percentage blood feeding inhibition (BFI) was estimated using the formula:
$$left[ {left( {F_{{text{c}}} – F_{{text{t}}} } right)/F_{{text{c}}} } right] times 100%$$
where Fc = number of mosquitoes found fed in untreated control hut, Ft = number of mosquitoes found fed in treated hut.
Pertentage induced exophily was estimated using the formula
$$left( {{text{Nv}}/{text{Nt}}} right) times {1}00$$
where Nv = number of mosquitoes found in verandah, Nt = the total number of mosquitoes found inside the hut and verandah.
The overall protective efficacy (%) of the treatment considered that a significant number of mosquitoes were inhibited from penetration and not killed by the treatment and was estimated using the formula
$$left[ {left( {D_{t} {-}D_{c} } right)/P_{c} } right] times 100$$
where Dt is the total number of mosquitoes found dead in the treatment tunnel, Dc is the total number of mosquitoes found dead in the control tunnel, Pc is the total number of mosquitoes penetrating in the control tunnel.
Institutional review board statement
The required data were collected after obtaining ethical clearance from Kilimanjaro Christian Medical University College (Research Ethical Clearance Certificate No. 2512), and it is part of the large ongoing research program that was reviewed and approved by the Tanzania National Institute for Medical Research (NIMR/HQ/R.8c/Vol.1/554). Written informed consent in local language (Swahili) was obtained from the participants before participation.
Informed consent
Informed consent was obtained from all subjects involved in the study.
Source: Ecology - nature.com
