Study area and villages
The study was carried out in the area of the Senegal river basin in five villages (Mbane, Ndiawara, Dioundou, Guia, and Khodit). The Senegal river has three distinct regions which are characterised by distinct environmental conditions: a mountainous upper valley which runs from the source in Guinea to Bakel; a middle valley which runs from Bakel to Dagana, including Podor, and a lower valley which runs from Dagana to the delta, where the Diama dam is located (Figure 1). The study villages belong to two sites and represent three different means of accessing water. The village of Mbane is located along the Lac of Guiers in the lower valley of the Senegal river basin. The villages of Ndiawara, Dioundou, Guia, and Khodit are located in the middle valley. Ndiawara and Dioundou are located along a tributary of the Senegal river (Doue), while Guia and Khodit mainly depend on an irrigation canal.
The village of Mbane
The village of Mbane (16°16′15′′N 15°48′07′′W) is located on the eastern shore of the Lac de Guiers, about 30 km to the south of the town of Richard Toll. The population is mainly composed of Wolof, Toucouleur, and Moor ethnic groups. The village has a population of 2,650 inhabitants. Water for domestic and recreational activities is mainly supplied from the directly from the lake. Livestock and agriculture are the main income-generating activities, and trade is only subsidiary. Fishing is artisanal and is focused exclusively on the lake. Market gardening is the main agricultural activity and takes place close to the lakeside. The village has two primary schools, a French school and a French-Arab school, with a total of 598 pupils. There are taps and latrines in the village and pit latrines at the schools. The lake is a permanent transmission site.
The villages of Ndiawara and Dioundou
The villages of Ndiawara (6°35′05′′N 15°51′2′′W) and Dioundou (6°36′4′′N 14°54′31′′W) are located between the Taredji-Podor road and the Doue river on the right. The Doue river is a secondary branch of the Senegal river, which forms an island known as Morfil Island. The villages are about 6 km from one another and are mainly inhabited by the Toucouleur ethnic group. The other minority ethnic groups are the Wolof and the Moors. Economic activity was mainly based on pastoralism, but the inhabitants have now become agro-pastoralists with farms in the irrigated areas. Market gardening is the main economic activity along the Doue river. The river contains water throughout the year, and it is used for laundry, dishes, swimming and recreational activities. It constitutes the main transmission site for the populations of the two villages.
Ndiawara is a big village with a population of about 3,842 inhabitants. In Ndiawara, there is a big school with eleven classes with about 367 pupils and a health centre. There are four pit latrines in the school and the village also has a water tower and two wells which are accessible to all villagers. Public taps are generally not operational, although some households have running tap water.
Dioundou is a small village with about 620 inhabitants and one school with four classes and a total of 50 pupils. There are three public taps which are served by the Ndiawara water tower. There is also one well and latrines in the village, and some houses have tap water. There is no health centre in the village and the populations consult at the medical centre at Ndiawara.
The villages of Guia and Khodit
Guia (16°35′53.′′N 14°55′23.2′′W) and Khodit (16°35′44.2′′N 14°56′41.1′′W) are located to the left of the Taredji-Podor road, 10 km from Podor between the Doue river and an irrigation canal that runs through the two villages. The Toucouleur ethnic group is the main group in the villages. The villages include several farms mainly producing food which includes winter crops (market gardening near the canals), livestock and fishing. The villages are about 2 km from one another. These two villages depend on a pumping station that drains water from the river into a large irrigation canal. Several gates exist along the irrigation canal where people come into contact with the water and which that constitute the major source of transmission.
Guia has a population of about 4,672 inhabitants. The village hosts a weekly market where traders from surrounding villages meet to sell their goods or agricultural or market gardening products. Guia has two schools with a total intake of 640 pupils. One is large (465 pupils) and is located not far from the irrigation canal. There are four pit latrines in the two schools. There is also a water tower and two wells. Public water taps were previously numerous, but are no longer functional, since tap water is now supplied to most houses. The health centre is located at Guia and serves other villages, including Khodit.
Khodit is less populated than Guia with about 2,336 inhabitants and one school which had 216 pupils in 2020. There are four pit latrines, one well and one water tap in the school and one in the village. The public tap is supplied by the Guia water tower and also populations consult at the Guia health centre.
Sample selection and study design
This study is based on a prospective cohort of children aged between 5 and 11, before and after a single treatment with PZQ (40 mg/kg) in five villages selected based on previously reported S. haematobium prevalence [30]. The minimum sample size for the baseline prevalence for this study was calculated based on the normal distribution N = ε2pq/i2, with N being the size of the sample; ε the normal Z score corresponding to the risk of error α = 5%; p the mean prevalence of S. haematobium (90%) in the villages; q = (1—p); and i the precision, fixed at 5%. N = (1.96) x 2 x (0.9) x (0.1) / (0.05) 2 = 138. We increased this number to account for the possible loss of patients during the study that could be due to failure to sign the informed consent paperwork, the absence of children during the baseline screening, failure to provide a sample, and other reasons, by adding 5% of the minimum sample size: N = 138 + (138 x 5 / 100) = 145. Each village was considered as a cluster and since, at the village level, children living in the same household incur the same risks for S. haematobium infection, the sample size in each village (N=145) was multiplied by two in order to limit this cluster effect. We therefore calculated the sample size for the study at 1.450 children, although only 777 children were examined for S. haematobium at the baseline. As the previous prevalence was high (over 89%), we expected to have at least 50 positive children in each village (250 in total) at baseline for PZQ efficacy and re-infection monitoring, but only 226 positive children were monitored (Figure 2).
Sampling and data collection
The baseline sample collection took place in August 2020, nine months after the last mass drug administration by the NSCP. Class registers were used to randomly select the study participants, taking into account the sex ratio in each school, and an identifier number was randomly allocated to each child throughout the follow-up period. Urine samples were collected in sterile 150 ml containers between 10 am and 2 pm. A urine container with the corresponding identifier was given to each child, and they were asked to go individually to the latrine and fill the container with their urine. The containers were checked to ensure that at least ten millilitres could be obtained for the quantification of eggs. If the volume collected was too low, the child was asked to drink water and wait until urinating again. The container was then placed directly in a cooler to protect it from sunlight and to prevent the eggs from hatching. Once collection was completed in one village, the samples were transported to the laboratory for microscopic examination on the same day.
Baseline microscopic examination
In the laboratory, each urine sample was gently shaken to ensure homogenisation of the eggs before filtering ten millilitres of the liquid through a Swinnex® filter. Microscopic examination of each sample to detect S. haematobium eggs was performed using the filtration method [44]. The positivity of S. haematobium and the number of eggs per ten millilitres of urine (infection intensity) were recorded. A child was considered to be infected by S. haematobium at baseline if at least one egg was found in their urine sample during microscopic observation. The baseline prevalence and intensity of infection of S. haematobium were calculated considering all participants in each village.
PZQ treatment
The product used was Praziquantel, an existing product manufactured by Merck, SA de CV, Mexico. The drug is also used by the NSCP in MDAs in school-aged children in Senegal. Treatment was administered at the standard dose of 40 mg/kg at the health centre or at the school in each village, in accordance with Senegalese NSCP guidelines and in the presence of community relays from the health centre. To make sure that the drug was swallowed correctly, each child was asked to open their mouth for verification. If the drugs had not been swallowed, another glass of water was given to the child. For children who vomited the drug, a second full dose was given. Children who continued to eliminate schistosome eggs (either viable or not) one month after the initial treatment received a second dose of 40 mg/kg.
PZQ efficacy assessment
PZQ efficiency was controlled four weeks after the treatment. Urine samples were collected and microscopically analysed in search of residual eggs and infection intensity was recorded, as previously described. In all samples in which calcified or non-calcified eggs were found, a hatching test to verify the egg’s viability was performed. The hatching test consisted of filtering out the remaining urine and rinsing the filter then placing the eggs in a Petri dish containing fresh water. The Petri dish was then exposed to artificial light for thirty minutes to allow hatching, and then checked under a dissecting microscope to detect swimming miracidia. The assessment of PZQ efficacy was measured by determining CR and ERR, as defined by previous studies [16,17]. The ERR was calculated using the following formula: [1 − AMEC after treatment/ AMEC before treatment)] x 100 [16].
Monitoring re-infection
Re-infection was monitored seven months after treatment, in April 2020. At this time, any child emitting parasite eggs was considered as being re-infected with S. haematobium. Urine samples were collected and examined as described above. Re-infection was only monitored in children who were positive at baseline, received treatment and who were negative at the treatment assessment time point. Each individual was considered as being re-infected if at least one miracidium was found after the eggs had hatched.
Data analysis
The prevalence of S. haematobium was calculated as the number of children who were positive over the total number screened. Intensity of infection was classified as “light” if the number of eggs/10 ml was between 1–49 eggs and “heavy” if it was more than 50 eggs [45].
At the baseline (S0), we aimed to test the effect of the type of water access, namely, irrigation canal (Guia and Khodit), lake (Mbane) or river (Dioundou and Ndiawara), and also the age and the sex of the child with regards to parasite prevalence and parasite egg load. Since our sampling design displayed an underlying spatial structure in villages, we used generalised linear mixed models (glmm) with villages set as random effect. The first model aimed at explaining the parasite prevalence (measured at the individual level, i.e. 0 or 1) and, hence, a binomial error term distribution was used. The second model aimed at explaining the parasitic load (measured at the individual level and only for children who had been parasitised), and a negative binomial error term distribution was used, since over-dispersion was detected. The two glmm were adjusted using the R package glmmADMB [46] and the significance of the results was examined using a type II Anova. The level of re-infection between water accesses types were compared using the Chi-squared test. In the case of a p-value < 0.05, the difference was considered as statistically significant.