The study was conducted on Pemba, one of the two main islands of the Zanzibar archipelago in the United Republic of Tanzania. Pemba has two regions (North and South) that are divided into four districts (i.e. Chake Chake, Micheweni, Mkoani and Wete), which are further subdivided into 129 small administrative units, known as shehias [22]. The SchistoBreak project was implemented from 2020 to 2024 in 20 shehias in the two northern districts of Pemba, Micheweni and Wete. In 2022, the 20 shehias had an estimated population of 95,000 [22]. A total of 81.5% of households in the Northern Region use improved sources of drinking water [23]. Piped water via taps and wells is available to 74.9% of household in this region [23]. For domestic activities and some recreational activities, the population of Pemba including children and adults rely mainly on the extraction of underground water from springs and wells, as tap water supply is often erratic [24, 25].

In 2020, there were 26 primary schools in the study area [26]. Furthermore, there were 239 Islamic schools (madrassas) located in the study area, which children attend in addition to primary or secondary schooling [17, 27]. The study presented here reports findings on the schistosomiasis-related KAP of children who attended the 16 main public primary schools in the study area and received BCC interventions in hotspot schools (comparing baseline results from late 2020 with follow-up results from early 2022).

The SchistoBreak project was designed as an intervention study with repeated cross-sectional surveys [20]. The cross-sectional surveys served to assess the impact of interventions annually and to target interventions to the local micro-epidemiology. Shehias with a S. haematobium prevalence of ≥ 3% in schoolchildren and/or ≥ 2% in community members were considered as hotspot IUs and received an intervention package consisting of (i) annual preventive chemotherapy with praziquantel in schools and communities; (ii) focal snail control with the molluscicide niclosamide; and (iii) BCC interventions. Shehias with a S. haematobium prevalence below the aforementioned thresholds were considered low-prevalence IUs and received surveillance-response interventions that included targeted treatment and focal snail control measures but did not include any BCC interventions.

The cross-sectional surveys for this study were conducted in the 16 main public primary schools in the study area. The baseline survey was carried out in November 2020 and the follow-up survey in March 2022. S. haematobium infection and schistosomiasis-related KAP were assessed in children attending grades 3, 4 or 5. In each grade, 25 children were selected; hence, 75 children per school [20]. For random selection, all children in a grade lined up, stratified by sex, and each third child, alternatively from the male and female rows, was systematically selected to be included in the study until 25 children per grade were reached [20].

Selected children received an information sheet and informed consent form (ICF) for their parents to sign. On the next day, when children had submitted the ICF signed by their parents or legal guardians, they were registered electronically by assigning them a unique identifier code (ID) and recoding their socio-demographic information (sex and age) using computer tablets (Samsung Galaxy Tab A; Samsung Electronics, Seoul, the Republic of Korea, 2019) and Open Data Kit (ODK, www.opendatakit.org) software.

Subsequently, each child was given a plastic container labelled with their ID and invited to provide its own urine sample. Urine samples were produced between 10 am and 2 pm and collected by the members of the study team who also interviewed the children about their schistosomiasis-related KAP. For this purpose, each child sat down in private with a study team member who asked questions in Kiswahili and recorded responses using a pre-tested ODK questionnaire (Additional file 1). The questions focused on children’s KAP regarding schistosomiasis transmission and prevention. There were questions about children’s knowledge of the lifecycle of S. haematobium, ways to prevent both getting infected with and spreading S. haematobium infection, their attitude towards schistosomiasis prevention of transmission and infection, and their access to improved water sources, their water contact behaviour and play practices.

Urine samples were transferred to the Public Health Laboratory-Ivo de Carneri (PHL-IdC) in Chake Chake on the day of collection. Urine samples were examined for S. haematobium eggs using a urine filtration method and microscopy by experienced laboratory technicians [26]. Data of the egg counts were recorded on paper sheets by laboratory technicians and subsequently double-entered into a Microsoft Excel database version 2016 (Microsoft Cooperation; Redmond Washington, USA) by two data entry clerks from PHL-IdC.

The BCC interventions in the SchistoBreak project were implemented from May to October 2021 in the five shehias that were stratified as hotspot IUs, based on the results of the first cross-sectional survey. The BCC tools and strategies used in the hotspot IUs in the SchistoBreak project area were originally developed in the ZEST project [17,18,19]. The measures were identified and designed together with the local population in Zanzibar, using a human-centred process, and included community-based and school-based interventions [28]. We adopted the following BCC measures that had worked well in the ZEST project and improved children’s knowledge and behaviour [29].

In communities, small-scale WASH measures (i.e. washing platforms) were installed in close proximity to a pump, tap or well. For this purpose, the behaviour study team organised meetings with community leaders (shehas) and community members to discuss the purpose and placement of the washing platform. Once an appropriate location was identified, the washing platforms were constructed with the labour support of the communities. The prerequisite for the location was that a working public tap or well was nearby or could be included directly into the platform. The round concrete platforms had a diameter of approximately 4 m, a sitting wall at the edge, and a drainage system with a slope where water could drain effectively into a large draining hole. Two washing platforms per hotspot IU were installed. Moreover, additional community meetings were held after construction to explain the advantages of using washing platforms for the prevention of schistosomiasis to the communities.

In schools, classroom- and school-based BCC interventions were implemented. For this purpose, teachers of all public primary schools and madrassas of the five hotspot IUs were invited for training at PHL-IdC. In the training, teachers learned about the transmission, prevention and symptoms of urogenital schistosomiasis, and methods for classroom-based participatory teaching, such as the application of blood fluke pictures, creating snail boards with students and joint drawing of S. haematobium life cycles [28]. Moreover, they practised approaches and safe play methods with educational messages about urogenital schistosomiasis for the implementation of school-based “Kichocho” (Kiswahili word for schistosomiasis) outreach days. After the training, the teachers were provided with blood fluke pictures, with educational flipcharts detailing the life cycle of S. haematobium and information about prevention of transmission and infection, with snail boards and with equipment to conduct “Kichocho days”, such as ropes and balls. The teachers were encouraged to apply the participatory methods they had learned in their classes and asked for an appointment for a Kichocho day in their schools, which they organised together with the behaviour team from PHL-IdC. During the intervention period, Kichocho days were implemented in the main public primary school and all madrassas in each hotspot IU. Teachers and students, together with the behaviour team, created and practised songs, dramas and games with messages about S. haematobium transmission and prevention, which were then performed by the students on a Kichocho day, in front of all students and teachers of the school and additional visitors, such as parents and community members.

In the first intervention period in 2021, the following BCC interventions were implemented: Two washing platforms were constructed per hotspot IU within a period of 5–6 days. On average, three community meetings were held in each hotspot IU. In each hotspot IU, the behaviour team used 2 days to deliver books about schistosomiasis, educational flipcharts, blood fluke pictures and snail boards to school and madrassa teachers. On average, six Kichocho days were implemented in the primary schools and madrassas of each hotspot IU. The preparation for Kichocho days by the behaviour team required an average of 2.8 days per IU. A total of 7334 children and 419 adults participated in the 30 Kichocho days during the intervention period in 2021.

Considering a 20% dropout resulting from non-consenting parents or students’ absenteeism or inability to produce a sufficiently large urine sample, we aimed for a final sample size of 60 children per school and 960 children in all 16 schools in each of the annual cross-sectional surveys.

Assuming a prevalence of “good KAP” of 67%, the prevalence could be estimated with a precision of 7.4% points [precision = one-half length of the 95% confidence interval (CI)]. The intra-cluster correlation coefficient assumed for the sample size calculation of this study was 0.1.

The KAP data collected in ODK were transferred directly to a secure ODK server at the Swiss Tropical and Public Health Institute (Swiss TPH; Allschwil, Switzerland). The statistical software R versions 3.5.1−4.3.2 (R Foundation for Statistical Computing, Vienna, Austria) were used to clean the double-entered electronic records of the laboratory data. Any discrepancies between the entries were traced back to the original paper records and corrected accordingly. After cleaning, the KAP and laboratory data sets were merged and analysed using R version 4.3.2.

S. haematobium egg counts were stratified into light-intensity (1–49 eggs/10 ml of urine) or heavy-intensity (≥ 50 eggs/10 ml of urine) infection, based on WHO guidelines [30]. The results were reported as the percentage of schoolchildren with positive test results.

All our questions about knowledge and attitude were multiple-choice questions, where several responses could be provided per question. A wrong response was scored as 0, a reasonable response was scored as 0.5 and a correct response was scored as 1 (Additional file 2). Knowledge and attitude were graded based on the participant’s final score for the questions under the respective outcome variable.

Four questions pertained to the participant’s knowledge about the cause of S. haematobium infection, where and how it is transmitted and the animal serving as an intermediate host in the transmission. Based on the minimum and maximum total knowledge scores (0 and 8) an individual could obtain, using different intervals of scoring points as cut-offs, knowledge was reported as a single outcome, with the following stratifications: no knowledge (0), poor knowledge (0.5 to 2.5), moderate knowledge (3 to 5.5), and good knowledge (6 to 8).

There were two questions pertaining to the participant’s attitude and knowledge about behaviour to prevent S. haematobium infection and transmission. Based on the minimum and maximum attitude scores, using intervals of 1 scoring point as cut-offs, attitude was reported as a single outcome, with the following stratifications: poor attitude (0 to 1), moderate attitude (1.5 to 2.5) and good attitude (3 to 4).

Ten questions pertained to the participant’s reported risky and protective practices related to S. haematobium exposure and infection. Risky practices included the crossing of a waterbody or rice field when going to school or a farm, and using natural open waterbodies for specific washing practices or for playing. Protective practices included playing at home, in the village, football field or in the bush. Here, no scoring system was used, but percentages of the total number of schoolchildren who carried out these practices were calculated for each practice.

Descriptive statistics (proportions and means) were used to assess the difference in schistosomiasis-related KAP in schoolchildren who received BCC interventions versus schoolchildren who did not receive BCC interventions, both at baseline in late 2020 and at the 1-year follow-up in early 2022.

Boxplots were used to indicate the distribution of knowledge and attitude scores of students in hotspot and low-prevalence schools at baseline and follow-up. A linear mixed-effect model accounting for clustering within schools was employed to determine the difference in mean knowledge scores and mean attitude scores between children from hotspot and low-prevalence schools at follow-up.