The study was carried out at Niakhar (14°30′ N, 16°30′ W), a Demographic Site Survey (DSS) located in the region of Fatick (Sine-Saloum), 135 km east of Dakar, the capital of Senegal, West Africa. The study zone (Figure 1a) was approximately 15 km long and 15 km wide and covered 230 km². The climate is continental Sudan-Sahelian, with temperatures ranging from 24°C in December-January to 30°C in May-June 11 . The rainy season spans four months (July-October), and the mean annual precipitation was 506 mm from 1962 to 2010 12 . The district includes 30 villages with a total of 43,000 inhabitants 13 , mainly of the serer ethnic group (96.4%) 14 . An environmental survey of this area has mapped all of the water sources (backwater, ponds, wells and supply taps), and it showed that the two-thirds of the population had no access to tap water 11 . A more detailed description of the Niakhar district has been given elsewhere 15 .

All the studies in the Niakhar district were conducted with the approval of the National Ethics Committee of Senegal. Oral consent was obtained after meeting with parents of schoolchildren and chiefs of the villages. A cross-sectional descriptive and analytical study was carried out from February to June 2009, and it involved children aged 7 to 15 years. The methodology used was cluster sampling, as recommended by the World Health Organization, with samples of at least 210 subjects each, which were divided into 30 clusters of at least 7 subjects. Because we had 20 primary schools, we decided to work on 15 clusters of 14 subjects each, giving us the same sample size of 210. The sampling interval (SI) was determined by the formula SI = M/n, with M as the total number of schoolchildren in 2009 and n as the number of clusters (SI = 6.873/15 = 458). To determine the first cluster for investigation, a random draw of numbers was made between 1 and the number corresponding to the sampling interval (458). The school with the closest number of schoolchildren to the SI number was chosen first. Other schools were selected as follows: E, E + SI, E + 2 SI, E + 3 SI … (E, number of children in the first selected school). The 15 clusters were distributed over 14 villages (out of the 20 villages that had at least a primary school). A total of 210 schoolchildren were enrolled in the study.

Urine samples were collected between 10:00 a.m. and 2:00 p.m., and their macroscopic aspect in transparent urine jars was examined. All samples that did not have a more or less dark yellow and a translucent aspect, without any particulates, were considered abnormal. We selected three macroscopic aspects: clear urine (urine of normal appearance and translucent), cloudy urine (urine of abnormal appearance, not translucent, with clots or suspension elements) and hematic urine (abnormal urine, non-translucent and red). Data were collected using a questionnaire whose reliability has been demonstrated in several countries for the detection of high-risk areas for schistosomiasis 16 17 . This questionnaire was given to each urine-providing child to collect information on the child’s sex, age, water contact, symptoms of urinary schistosomiasis, knowledge about the disease, and previous praziquantel treatment. A plastic urine jar of 50 ml was then given to each child with a number corresponding to the questionnaire. Urine samples were transported to the Niakhar laboratory for analysis.

Microscopic examination of each urine sample for detection of S. haematobium eggs was performed using the method by Plouvier et al. 18 . In brief, 10 ml from each sample of urine was passed through a Millipore filter (12 μm polycarbonate filter), and this filter was analyzed with a 10 × objective. The number of eggs per filter was counted, and the infection intensity was classified as light (<50 eggs/10 ml of urine) or heavy (≥50 eggs/10 ml of urine), as defined by the World Health Organization 19 .

Data were recorded using the Epi-Info, version 3.5.1 (August 13, 2008) and analyzed using STATA 11.1. The relationships between characteristics of human infection (prevalence and intensity) and other variables, such as the location of villages, sex and age of children, water contact, or access to tap water, were tested. Differences in percentages were analyzed using the χ² test. The Geometric Means of Williams (GMW) was used for egg counts. The Mann–Whitney test was used for comparisons between the characteristics of human infection and the children’s sex, water contact or access to tap water, while the Kruskal-Wallis test was utilized for comparison between infection characteristics and the ages of children. Confidence intervals (CI) at 95% for GMW were estimated according to Kirkwood and Sterne 20 . A value of p < 0.05 was considered significant.

All the studies in the Niakhar district were conducted with the approval of the National Ethics Committee of Senegal, followed by the local health authority. Before the study began, the chiefs of the villages and parents were fully briefed on the objective of the study. The study was explained to each participant for their understanding and cooperation. Parents and participants were asked to consent verbally to participate in the study; only those who provided consent were enrolled and requested to prove urine samples. At the end of the study, all the children registered were treated with praziquantel according to their weight (40 mg/kg).

The sex ratio (M/F) was 0.9, with 101 boys and 109 girls. The average age was 10.4 years (SD: 2.1 years). The individuals were divided into three age groups (7–9, 10–12 and 13–15 years), representing 37.2%, 41.4% and 21.4% of the children, respectively. Hematuria (52.4%) and dysuria (54.3%) were the most commonly reported symptoms by both infected and uninfected children. Abdominal pain (34.7%) was the least reported symptom. Macroscopic urine examination showed that clear urine (49%) was more frequent than cloudy urine (40%) or hematic (11%) urine. A few children (4.8%) had knowledge of schistosomiasis, and 5.5% had received previous treatments for this disease. A question on travel outside the Niakhar area showed that only 12% of children traveled to other regions during vacations. The others, especially the youngest, remained in their villages (Table 1).

Of the 210 urine samples examined for S. haematobium, 121 (57.6%) were infected, with a mean geometric count (EMGC) of 185 [95% CI, 130–264] eggs/10 ml of urine, and 72.8% of the infected urine samples exceeded 49 eggs/10 ml of urine. The highest prevalence (> 90%) was found in the village of Kothiokh, and the lowest prevalence (< 25%) was found in the village of Ngonin. A significant relationship (p < 0.001) in the prevalence of infection between villages was noted. The greatest heavy infection intensities (> 90%) occurred in the villages of Ngayokhem, Ngonin and Poultok-Diokhine (Figure 1b).

The prevalence of infection was greater in persons who used backwater (69.6%) than those who used ponds (53.3%). A highly significant relationship (p < 0.01) was found between the prevalence of infection and the type of water contact. The highest EGMC, 202 (95% CI, 134–314) eggs/10 ml of urine, was found in persons who use ponds. People who did not have access to tap water had the highest prevalence and EGMC. Significant differences were also noted between prevalence and access to tap water (p < 0.005), and between intensity of infection and access to tap water (p < 0.005). If only the villages with tap water are considered, the prevalence and intensity of infection were significantly higher in boys than in girls (p < 0.01). In contrast, in villages without tap water, the prevalence and intensity of infection were higher in girls, but these differences were not significant (Table 2).

Boys had a higher infection rate (66%) and EMGC (253 [162–396] eggs/10 ml of urine) than girls (49.5% and 126 [95% CI, 130–264] eggs/10 ml, respectively). Significant differences in prevalence values and intensities of infection according to sex (p < 0.05) were noted (Table 3). Boys were heavily infected regardless of age, but a significant difference (p < 0.05) was only noted between boys and girls in the 7–9 year group.

The prevalence of infection increased with age, and statistical analysis indicated a significant relationship (p < 0.05) between these two variables. The EMGC decreased in the 10–12 year group, before increasing again in the 13–15 year group. No significant difference between the age of children and the infection intensity was noted (Table 3). However, if boys and girls are grouped, the intensity of infection decreased with age in girls and increased in boys. The difference between boys and girls was also significant in the 7–9 year group (p < 0.05). The variables age, gender and access to water were confronted to villages with medium and high intensity but statistical analysis showed no significant difference (p > 0.05).