Study site
The study was conducted in 12 of the clusters in the ATSB Phase III trial in Western Province, Zambia [28]. The clusters were within three districts: Nkeyema, Kaoma and Luampa districts in Western Province, Zambia. Malaria transmission in Western Province is known to be seasonal, typically characterized by peak transmission from January to May, relating to the annual rainy season, which typically lasts from November to March. This study was conducted for seven months during the rainy season to the middle of the cold season (from December 2021 to June 2022).
Study design
This was a repeat cross-sectional study design incorporated in the ATSB main trial (ClinicalTrials.gov Identifier: NCT04800055). Briefly, an enumeration activity was used to identify 70 clusters of approximately 175 households each. Restricted randomization was used to assign 35 clusters to the intervention arm (ATSB plus IRS/LLIN) and 35 clusters to the control arm (IRS/LLIN alone). Clusters used a ‘fried egg’ design whereby the intervention was deployed in the cluster ‘core’ plus a buffer area extending 600 m beyond, while sampling for outcome ascertainment for the epidemiological outcomes was only conducted in the core areas. ATSB stations were installed on eligible structures found within the entire cluster, including the buffer and core zones. Each eligible structure received two ATSB stations during an installation campaign in the first two weeks of November 2021. The ATSB stations were monitored throughout the trial period and replaced by ATSB monitors if they met predefined criteria of damage, including evidence of defined levels of mold, leakage, holes/tears, dirt, or depletion of the bait.
Only ATSB stations installed during the initial November 2021 campaign in 12 of the 35 intervention clusters that were selected for this sub-study were eligible for this residual bioefficacy study. ATSB stations that were installed as replacements for damaged or missing stations throughout the study period were not eligible for this study. Selected clusters included 10 clusters previously selected for mosquito sampling, plus two additional clusters chosen for logistical feasibility. Once per month, one ATSB not meeting replacement criteria was removed from a structure in each cluster (for a total of 12 ATSB stations per month) and transferred to the laboratory at Macha Research Trust for bioefficacy testing. The bioefficacy evaluations were conducted monthly for seven months, beginning in December 2021 after ATSB stations had been deployed for one month.
ATSB stations
The product evaluated was the Sarabi version 1.2.1 prototype ATSB station (Fig. 2). All the bait stations had unique preprinted QR codes. Digital photos, taken using mobile devices, included a photograph of the location of each ATSB prior to removal for the bioefficacy study. Photos were visually reviewed by a single investigator, and characteristics of the structure on which ATSB stations were hung were extracted and recorded for each station sampled, including wall type (mud, concrete block, or plaster covered block), roof material (thatch or tin) and presence or absence of roof overhang.
ATSB installation
The installation of ATSB stations on household structures in the study areas was based on standard eligibility criteria (Fig. 3). Structures that were eligible to receive the ATSB stations were defined as those with (1) a complete roof, (2) walls at least 1 m high, and (3) at least 3 complete walls. Nonresidential buildings (e.g., shops, schools, churches, tobacco drying sheds), kitchens that are not used for sleeping, animal kraals, toilets, bathing shelters, and drying racks were not eligible for ATSB installation. Two ATSB stations were hung using bamboo sticks, nails, and wires on each eligible structure. The bait stations were hung approximately 1.6–1.8 meters above the ground (where possible) on opposite exterior walls, prioritizing protected locations under an eave or roof overhang.
Bioefficacy Assessment
Procedure
Each month, a list of eligible ATSB stations was compiled for each of the twelve selected bioefficacy assessment clusters, where eligibility was defined by having been installed in the first two weeks of November and having no record of removal during routine ATSB monitoring (i.e., monthly inspection of ATSB stations by the ATSB monitors). Since monthly inspection of ATSB station conditions and replacement of those that had incurred damage, started leaking or became dirty or moldy was conducted, some stations that had been deployed at the start of the trial had already been removed and replaced prior to each bioefficacy collection, and those stations that had been removed and their replacements were not considered eligible for selection in this study. Using a cluster map and the list of eligible ATSB stations, two study team members (GM and BC) alongside cluster ATSB monitors visited each cluster core and buffer area on a monthly basis to locate, capture data, and remove one eligible ATSB. The first ATSB from the list of eligible barcodes that was found in each cluster that did not meet the damage or replacement criteria was removed for bioefficacy testing. Removed ATSB stations were immediately replaced with a new ATSB. The ATSB team explained to the household head that the reason for removing a bait station was to check its operational effectiveness in killing mosquitoes in the communities during the hanging period.
Inclusion criteria
-
Bait stations located within the buffer zones and core areas of 12 participating clusters.
-
Bait stations originally installed in November 2021 (have not been removed or replaced since the original installation period).
-
Undamaged (does not meet the criteria for replacement): not leaking, no holes or tears, no mold spots larger than the rubber of a pencil, fewer than 8 depleted cells (i.e., containing low levels of bait), and fewer than 8 cells covered with dirt.
Exclusion criteria
-
ATSB stations with 1 or more cells completely torn open, leaking off of the black membrane, depletion (8 or more cells), or dirt (8 or more cells).
-
ATSB stations with mold spots exceeding the size of the rubber end of a pencil.
-
ATSB stations deployed after the November 2021 installation period.
Safety measures on handling the ATSB bait stations
The personnel who were handling ATSB stations from the field site were provided with gloves for proper handling and removal of ATSB stations. Personal protective equipment, including laboratory coats and masks, was used during bioassays to avoid the inhalation of mold spores and to mitigate potential hazards associated with handling bait stations.
Data capture during bait station removal
Data capture was completed in Commcare (Diamgi Inc., Cambridge MA, USA) using Android smartphones at the time of removal of ATSB stations selected for bioefficacy. Data were recorded by a study team member familiar with the overall ATSB trial’s standardized ATSB damage criteria. This included the collection of the GPS coordinates of the identified structure at the entrance door and ATSB barcode. Data forms also included the capturing of the following characteristics: close-up photograph of ATSB, whether the ATSB had any holes and if any cells were completely torn open, if the ATSB was leaking (liquid dripping from the black surface onto white border or other surfaces), if 0 or 1–7 or 8 + cells were depleted, if 0 or 1–7 or 8 + cells were dirty, presence of any mold and if any mold spots were greater than the size of the rubber end of a pencil, a photo of ATSB’s position on the structure, and the cardinal direction the ATSB was facing while installed (e.g., north, south, etc.
Packaging, storage and transportation of ATSB stations
The removed ATSB stations were labeled with permanent markers as follows: the cluster number, month (i.e., M1 for month one), and date of removal. The bait stations were placed in a predesigned slot in a lockable wooden box for transport to the Macha Research Trust (MRT) for bioefficacy assessment. The wooden box had 12 removable wooden slots to which the ATSB stations were attached, so they were held vertically when transported. Bait stations were transported to MRT within a 4-day +/-1-day span (2–3 days collection and 1 day transit from Kaoma, Western Province to MRT) in the second week of every calendar month from December 2021 to June 2022. The ATSB stations were stored in a transport box awaiting transportation and prior to the assays at MRT.
Mortality Assessment of the bait stations
Prior to the assessment, ATSB stations were not cleaned or wiped to avoid altering the condition they were in when recovered from the field. The ATSB stations were tested using a laboratory colony of insecticide-susceptible An. gambiae s.s. (Kisumu strain) mosquitoes. Prior to the assessment, female and male mosquitoes of known age (3–5 days old) were reared under controlled conditions (hum; 80 +/-10%, temp; 27 +/-2 ⁰C)) in an insectary at Macha Research Trust. The conditions for the second insectary where the experiments were carried out were maintained within the same range.
Female and male mosquitoes were selected, placed in separate cages, and starved (no sugar) for 24 and 12 hours, respectively. Water-soaked cotton wool was provided in every cage. After the starvation period, mosquitoes were placed into release cups (50 in each cup or 25 in each cup if using small release cups). The mosquitoes were allowed to stay in cups for a minimum of one hour before releasing them into cages to allow them to stabilize from any stress that may have been incurred during the selection process.
A total of 84 field-deployed bait stations were collected, 12 every month for seven months. Out of the total collected, 71 were evaluated. Each bait station was mounted inside one of the walls of a separate 30 cm x 30 cm cage in portrait orientation (Fig. 4). Control cages with water only and with both water and 77% sugar solutions were also constructed to assess control water only and sugar fed mortality.
Cohorts of fifty starved male and fifty starved female mosquitoes were released into each testing cage and each of two control cages. To ensure that all released mosquitoes were in good condition, mosquitoes were allowed to fly out of the release cups by gently tapping on the side of the cups and removing the cups with any mosquitoes that did not leave the release cup. All cages were provided with cotton wool soaked in water. The cages were all covered with wet towels on the top surface to increase the humidity levels. The temperature/humidity was noted at the start and end of the experiments.
Data collection, management and analysis
In each of the eight assay rounds (seven from field-collected ATSB stations and one from the new non-deployed stations), mosquitoes were exposed for 48 hours to the two negative controls and to recently removed ATSB stations that met the inclusion criteria (up to 12 ATSB stations/month). In addition, 12 new, non-deployed ATSB stations were bioassayed in month 0 as a positive control. The dead mosquitoes were counted and removed at 24 hours for each treatment, including the control cages, and recorded in a standardized Excel worksheet. The remaining mosquitoes in each of the cages were left for an additional 24 hours. At the end of 48 hours, the numbers of dead and living mosquitoes in all treatment and control cages were recorded in a standardized Excel worksheet.
All data analyses and manipulation were conducted with R (v. 4.1.3) and R Studio (v. 2022.07.1, build 554) [29, 30].
Bioassay mortality was calculated by summing the number of mosquitoes that had died within 48 hours and dividing by the total number of mosquitoes released to feed on the ATSB in each assay. This proportion was then corrected for starvation and natural mortality rates from active sugar or water controls for each round of assays using Abbott’s formula:
$$\frac{\left(\% observed mortality-\% control mortality\right)}{(1-\% control mortality)}$$
The corrected natural mortality proportions were used for subsequent analyses. Bivariate and multivariable linear regression analyses were conducted using the corrected natural mortality (sugar control) proportion as the outcome. Predictor variables examined were condition (new/field deployed), holes, leaks, mold, depletion, dirtiness, wall material and roof material. Predictor variables were used to determine the potential factors that may influence the bioefficacy of field-deployed ATSB. Unadjusted estimates are presented as well as adjusted estimates that account for trial round as both a continuous and categorical (independent) variable.