Study setting

The study will be conducted on Mfangano Island in Homa Bay County, Kenya (Fig. 1). According to the most recent national census in 2019, the island has a land area of 66.2 km2 and a population of 24,123 distributed in 6085 households [23]. Luo and Suba are the major and minor ethnic groups on the island, respectively, with significant admixture between the two groups. The primary occupations of Mfangano residents include fishing in Lake Victoria and farming. Most compounds possess more than two house structures, and family dwell in each structure is referred to as a household. The main dwelling units are typically made of mud walls with metal sheet roofs, although units with walls made of iron sheets, concrete, or stones are also common [23].

Fig. 1

Map of Mfangano Island showing the locations of the 20 trial clusters

In general, the Lake Victoria region has two rainy periods per year, including the long rainy season from March to June and the short rainy season from October to November. However, irregular patterns have been observed in the last few years [24]. Malaria incidence peaks 1 to 2 months after the rainy season. The major vectors are Anopheles gambiae s.s., An. arabiensis, and An. funestus [25].

Mfangano has six public health facilities (three health centers and three dispensaries). The island is divided into 52 health units, each comprising 50 to 100 households. A CHV will monitor the health status of residents in each health unit. Under Homa Bay County government mandate, CHVs are to test suspected malaria cases by RDT and provide ACT treatments to confirmed positive cases. LLINs are distributed for free by the government every 3 years, and the most recent distribution in Homa Bay County began in March 2021.

Eligibility criteria

The inclusion criteria for the installation of the Olyset®Plus ceiling net are (1) residential structures housing at least one permanent resident aged 18 years or older in the household and (2) informed consent provided by at least one adult in the household. The exclusion criteria are (1) vacant dwelling structure (confirmed by at least two visits by CHVs) and (2) dwelling structure to be vacated or destroyed within the study period.

The inclusion criteria for cross-sectional malaria and anemia prevalence surveys in schools are (1) age between 3 and 15 years old, (2) living in the study area during the study period, and (3) informed consent being provided by the parent or legal guardian before each survey. The exclusion criterion is (1) having severe chronic illnesses.

The inclusion criteria for prospective cohorts are (1) living in the study area at the time of Olyset®Plus ceiling net installation, (2) having no plan to leave or stay outside the study area for an extended period (longer than 1 month) over the 12-month follow-up period, and (3) informed consent provided (if the participant is minor (less than 18 years old), informed consent has to be provided by the participant’s parent or guardian). The exclusion criteria are (1) severe chronic illnesses and (2) pregnancy at the time of Olyset®Plus ceiling net installation. Table 1 summarizes these criteria.

Table 1 The inclusion and exclusion criteria for ceiling net installation, cross-sectional survey, and prospective cohort surveyWho will take informed consent?

Written informed consent will be obtained by the study team members fluent in the local languages (Luo), Kiswahili, and English and who fully understand the study protocol. After eligibility is confirmed, the study team members will present to the potential participant a document containing all relevant information about the study in Luo and English. If the participant cannot read, study information will be conveyed verbally by the study team members. The potential participant will have opportunities to ask any questions. Agreement to participate is sought only after the participant indicates an appropriate understanding of the study.

Additional consent provisions for collection and use of participant data and biological specimens

The study information document for ceiling net installation contains the study overview. In addition, the documents for cross-sectional and cohort surveys contain details on collecting, storing, and using personal data and biological specimens during the study.

InterventionsExplanation for the choice of comparators

In Kenya, LLIN is the most widely used malaria preventive measure. The Division of National Malaria Programme coordinates free LLIN distribution, and the county governments deliver LLINs to residents in all endemic counties every 3 years. The primary purpose of this trial is to demonstrate the superiority in malaria prevention of adding Olyset®Plus ceiling nets to the standard LLIN. Thus, in the control arm, no Olyset®Plus ceiling nets will be installed. Free LLIN distribution and use will be allowed in the control and intervention arms as the current best practice. As for the planning stage, there is no plan for new LLIN distribution during the study period.

Intervention description

In the intervention arm, Olyset®Plus ceiling nets will be installed in all dwelling units where residents sleep, free of charge to the households. All participants will be encouraged to continue to use LLINs, distributed by the Homa Bay County government.

Ceiling net installation teams consisting of experienced installers who have participated in the previous ceiling net trial [14], CHVs, and community volunteers from each health unit will schedule the installation time for each eligible household. The head (or another adult) of the eligible household will be informed at least 24 h prior to the scheduled installation time.

The ceiling net material is a rectangular sheet of Olyset®Plus net measuring 6 m × 8 m or 4 m × 6 m, depending on the size of the house [8]. The ceiling net has loops along the diagonals, allowing the net to be roped to the roof support beams. After tying the central loop to the center of the roof, all four sides of the net will be pulled taut, and the remaining loops will be tied to the roof support beams. The edges of the ceiling net are then pinned to the walls, screening the opening between the roof and the walls (Fig. 2).

Fig. 2

Scheme of the ceiling net and the process for the installation. A The rectangular LLIN sheet is reinforced along the edges and diagonals. Four loops sewn on each diagonal and one central loop sewn at the intersection of the diagonals enable the net to be tied with strings to the roof support beams and the center of the roof, respectively. B The central loop is tied to the center of the roof. C The net is pulled tight to avoid slack, and the rest of the loops are tied to the support beans. D The edges of the net are pinned to the top of the walls, and the gap between the walls and the roof is screened

Criteria for discontinuing or modifying allocated interventions

Since the ceiling net is semi-permanently installed, the intervention will be discontinued only when the participant specifically requests the removal of the ceiling net by the study team. We do not allow any crossover from the control arm to the intervention arm during the follow-up period. Those who migrate between the arms or emigrate from the study areas will be dropped from the study follow-up.

Strategies to improve adherence to interventions

Adherence to the intervention cohort in this study is defined as sleeping in houses with Olyset®Plus ceiling nets. Adherence is monitored indirectly by assessing the number of nights each participant spends outside their house during the bi-weekly interview. During each house visit, CHVs will visually inspect the condition of the ceiling nets. Any visible tear and damage to the ceiling net will be reported to the research team by the size and location of the damage.

Relevant concomitant care permitted or prohibited during the trial

There is no specific concomitant care prohibited during the trial. All participants in both arms will continue to receive and use free LLIN and have access to standard medical care, including malaria testing by RDT and treatment with ACT.

Provisions for post-trial care

All participants will be under the normal healthcare system in the study setting. No perceived health risks for the intended population are expected with the intervention. Our plan of continuous cross-sectional malaria surveillance after the study period allows us to monitor further parasite transmission in the population.

Outcomes

The primary outcome of the study is malaria prevalence by microscopy in schoolchildren (3 to 15 years old) 12 months after Olyset®Plus ceiling net installation. The secondary outcomes are (1) malaria prevalence by PCR in schoolchildren at 6 and 12 months post-installation; (2) cumulative malaria incidence in the prospective cohort for 12 months post-installation; (3) anemia prevalence in schoolchildren at 6 and 12 months post-installation; (4) prevalence and level of antibodies (IgG) against Anopheles salivary proteins gSG6 and cE5 in the prospective cohort at 6 and 12 months post-installation, as a measure of exposure to mosquito bites; (5) anopheline mosquito density, species composition, and sporozoite infection prevalence at 6 and 12 months post-installation; (6) the percentage of cohort participants who sleep under the bed net at 6 and 12 months post-installation; and (7) perception of the Olyset®Plus ceiling nets among community members.

Participant timeline

The study flowchart and sampling timeline are presented in Table 2.

Table 2 The study flowchartSample size

Parasite prevalence by microscopy among schoolchildren on Mfangano Island is approximately 20%, based on our previous study [26]. To detect the expected effect of 60% parasite prevalence reduction (to a prevalence of 8%) relative to control, with a power of 0.8, two-sided type 1 error of 5%, an intraclass correlation coefficient (ICC) for clusters of 0.063 (estimated based on our previous school study by assuming each school as a cluster), and average cluster size 150 (number of schoolchildren), the required number of clusters is 10 (or 1500 children) per arm using Stata/MP16.1. Based on the 2019 national census, children 3 to 15 years old account for 40.3% of the total population in Homa Bay County. The study requires a total population (all ages) of 7444 people or 1880 households (mean of 3.96 persons per household on Mfangano). By assuming a 3% dropout per month, we plan to enroll 1315 households for the intervention arm.

For the longitudinal cohort, we estimate the infection incidence on Mfangano Island as approximately 2.0 per person-year based on the data on Kenya Health Information System (KHIS). To detect the expected effect of a 40% reduction (1.2 per person-year), with a power of 0.8, two-side type 1 error of 5%, and an average cluster size of 25 individuals, the required number of clusters is 7 (or 350 individuals) per arm based on Eq. 7.3 on reference [27]. Assuming 3% of cohort participants will drop out per month, we plan to enroll 490 individuals in the cohort.

Recruitment Community sensitization

We first sought approval from the Homa Bay County Government Ministry of Health and based on their advice, hosted meetings with CHVs, village chiefs, and public health officers on Mfangano Island, and other key stakeholders from the county to explain the purpose, scope, objectives, methods, timeline, and potential significance of our study. CHVs and village leaders were asked to disseminate study information to and answer questions from community members. Feedback from CHVs and village leaders was sought, and regular meetings were held among CHVs, village leaders, and the study team to devise and refine approaches to engage communities. Finally, broad-level community consent to participate in the ceiling net trial was sought through CHVs and village chiefs.

Community census

After community sensitization and affirmation of agreement to participate provided by village chiefs, a census will be conducted by CHVs and experienced local research assistants to enumerate and record demographic information from all households in the health units that are receptive to our study. The following information will be collected from each household: (1) the number of residents; (2) the age, gender, and occupation of each resident; (3) the name of the school attended by each child; (4) the number, type, size, and functions of house structures; (5) the current LLIN ownership and usage; and (6) the GPS coordinates of the household. Written informed consent to participate in the ceiling net trial will be sought from the head of each household during census house visits.

Assignment of interventions: allocationSequence generation

Random numbers are generated using Microsoft Excel 2016.

Concealment mechanism

Each cluster is assigned a computer-generated random number. The random numbers are then arranged in ascending order. Clusters with random numbers in the smaller half are assigned to the intervention arm, while those in the larger half are in the control arm.

Implementation

The allocation sequence and assignment are generated by a volunteer who has no knowledge about the study. Local study assistants will perform participant enrollment.

Assignment of interventions: blindingWho will be blinded

Due to the visibility of the Olyset®Plus ceiling net, neither the trial participants nor the members of the study team who take part in field activities can be blinded. However, laboratory- and office-based personnel (e.g., microscopists, laboratory technicians, and data analysts) will be blinded to the identity and intervention status of the trial participants since all biological specimens will be identified by a unique numeric study identifier, and personal information will be removed before analyses.

Procedure for unblinding if needed

This is an open-label trial, and only the data measurers are blinded. Therefore, there is no circumstance that they need to be unblinded.

Data collection and managementPlans for assessment and collection of outcomes Cross-sectional school surveys

Malaria prevalence in children will be estimated using cross-sectional malariometric surveys in schools. These surveys will be conducted at baseline (before ceiling net installation), 6 months, and 12 months post-installation. In addition, class rosters will be cross-referenced with the community census to determine the allocation of each eligible student in the trial clusters.

Malaria status will be determined using three methods: RDT, microscopy, and PCR. First, a finger prick blood sample will be collected for on-site diagnosis by Paracheck-Pf® RDT (Orchid Biomedical Systems, India). Survey participants with positive test results will receive a treatment course of artemether-lumefantrine with dosing instructions as per the guidelines from the Ministry of Health in Kenya. Blood smears will be prepared on-site and transported to the main laboratory in Homa Bay, where thin smears are fixed with methanol. All smears are stained with 3% Giemsa solution for 30 min, then examined by experienced microscopists. Two blood samples (70 µl each) will be collected with a 75-mm EDTA-coated micro-hematocrit capillary tube (Marienfeld, Lauda-Königshofen, Germany) and spotted on Whatman ET31 Chr filter paper (Whatman International. Maidstone, UK). The blood spots will be allowed to dry at ambient temperature and stored in individual zipped plastic bags at − 20 °C. The dried blood spots (DBS) will be used for DNA extraction and determination of malaria status by PCR [28].

Hemoglobin (Hb) levels will be measured using the HemoCue Hb 801 Analyzer (HemoCue, Ängelholm, Sweden). Survey participants with severe anemia based on the WHO criteria [29] will be referred to local health facilities for further consultation.

Cohort survey

Malaria incidence, exposure to Anopheles bites, will be assessed during and after 12 months of Olyset®Plus ceiling net installation. Every 2 weeks, CHVs will visit the homes of cohort participants. A structured questionnaire created using the Research Electronic Data Capture (REDCap) application and loaded on an Android-based tablet computer will be used to collect any history of fever, malaria episode, visit to the local health facilities, and travel in the previous 2 weeks [30].

Every month, certified medical laboratory staff will accompany CHVs on home visits to collect blood samples from cohort participants. Blood sampling by finger prick and venipuncture will alternate between monthly visits, starting with a finger prick at baseline. Blood samples will be used to determine malaria infection status by RDT, microscopy, PCR, and Hb measurement as described for the cross-sectional surveys. In addition, whole blood (500 µl for capillary blood and 3 ml for venous blood) will be collected in EDTA tubes. These samples will be transported to the main laboratory in Homa Bay County Teaching and Referral Hospital and stored at − 80 °C.

Training sessions will be held for CHVs to familiarize themselves with the questionnaire’s content and the REDCap application’s use to record the responses. Built-in validation and completion checks will ensure data quality and completeness, respectively. To avoid duplication, all microscope slides, filter papers, and sample tubes will be pre-labeled with auto-generated serial numbers. CHVs and certified medical laboratory staff will be prompted by the REDCap application before blood sampling to confirm the identity and serial number of the cohort participant. The completeness of blood sampling will be confirmed twice after the sampling step in the field and at the sample storage step in the laboratory.

Mosquito density, species composition, and infection rates

A cross-sectional entomological survey will be conducted to obtain baseline data before the intervention. Indoor-resting female anopheline mosquitoes will be sampled from five houses in each cluster using pyrethroid spray catches and light traps. All selected houses are made of mud walls and consist of one room. Using the baseline data, the cluster size (the number of houses) will be estimated for a post-intervention survey with a 50% density reduction and power of 0.8. The post-intervention survey will be conducted at the end of the rainy season during the study period. Sampled anophelines will be identified at the species level with PCR. The occurrence of Plasmodium parasites in the salivary glands of individual mosquitoes will also be examined using PCR.

Acceptability of Olyset®Plus ceiling nets

Focus group discussions (FGDs), a structured questionnaire, and in-depth interviews will be used to assess the acceptance of the Olyset®Plus ceiling nets. At the end of the 12-month follow-up, all cohort members will be subjected to the structured questionnaire, and a part of them will be invited to in-depth interviews based on their responses to the questionnaires.

Plans to promote participant retention and complete follow-up

To promote retention among participants in the control arm, we will install ceiling nets in their houses after the 12-month follow-up period, irrespective of the study outcome. This arrangement was agreed upon by CHVs and village leaders as an acceptable study incentive, conveyed to potential study participants during community sensitization, and reiterated to participants during the census survey and informed consent process.

In the cohort survey, CHVs will make an appointment with participants and confirm their available date and time before each bi-weekly home visit. Participants will receive a small remuneration (i.e., rice, beans, cooking oil, and soap) when they provide a venous blood sample. CHVs will be instructed to relay to the research team any issue raised by cohort participants, and discussions will be held to resolve issues that cannot be immediately addressed. The research team will periodically accompany the CHVs in their home visits to reinforce to cohort participants the importance of the study.

Data management

Study data will be collected on Android-based tablet computers using the REDCap application to promote data quality and security. Data validation, such as range checks and completeness checks, will be enabled in all survey instruments. For cross-sectional surveys, data will be uploaded to the REDCap server at Mount Kenya University at the conclusion of each survey day. After the data manager confirms the data quality on the server, data stored locally on the tablet computers will be deleted before the next survey to avoid potential overwriting of existing data. Survey data will be uploaded to the REDCap server at least once a week for the longitudinal cohorts. Each cohort participant is given a unique identifier, and each visit is preprogrammed as a defined event in the REDCap data collection instrument to facilitate data entry. Cohort surveys will be conducted by CHVs familiar with cohort participants and will be prompted to confirm the identity of the cohort participants before data entry. In addition, the data manager will confirm the data quality on the server once a week.

Access to survey data will be limited to data analysts and the data manager in the research team. In addition, personally identifiable information will be removed before data analyses.

Confidentiality

To maintain confidentiality, each participant in cross-sectional surveys and the longitudinal cohort is assigned a unique identifier. The data collected will be labeled using the unique identifier and stored separately from the key linking personal information (name, date of birth, GPS, and phone number). The data will be kept on a secure server that is only accessible to the research staff. Publications will contain only aggregated data, and no personal information will be included.

Plans for collection, laboratory evaluation, and storage of biological specimens for genetic or molecular analysis in this trial/future use

Blood samples will be collected to examine malaria infections by multiple methods, hemoglobin levels, immunity against malaria parasites and mosquito saliva, and malaria parasite genomics. No human genetic studies are planned in this study. However, any biological specimens remaining after analyses described in this study will be stored indefinitely for future studies unless the participants opt out during the informed consent process. Contact information of the study team is provided in the consent form to study participants, who can remove themselves from this study and any future studies that may use their blood samples at any time without penalty or prejudice.