Schistosomiasis in school-age children in Burkina Faso after a decade of preventive chemotherapy

Abstract Objective To assess the impact of a decade of biennial mass administration of praziquantel on schistosomiasis in school-age children in Burkina Faso. Methods In 2013, in a national assessment based on 22 sentinel sites, 3514 school children aged 7–11 years were checked for Schistosoma haematobium and Schistosoma mansoni infection by the examination of urine and stool samples, respectively. We analysed the observed prevalence and intensity of infections and compared these with the relevant results of earlier surveys in Burkina Faso. Findings S. haematobium was detected in 287/3514 school children (adjusted prevalence: 8.76%, range across sentinel sites: 0.0–56.3%; median: 2.5%). The prevalence of S. haematobium infection was higher in the children from the Centre-Est, Est and Sahel regions than in those from Burkina Faso’s other eight regions with sentinel sites (P < 0.001). The adjusted arithmetic mean intensity of S. haematobium infection, among all children, was 6.0 eggs per 10 ml urine. Less than 1% of the children in six regions had heavy S. haematobium infections – i.e. at least 50 eggs per 10 ml urine – but such infections were detected in 8.75% (28/320) and 11.56% (37/320) of the children from the Centre-Est and Sahel regions, respectively. Schistosoma mansoni was only detected in two regions and 43 children – i.e. 1 (0.31%) of the 320 from Centre-Sud and 42 (8.75%) of the 480 from Hauts Bassins. Conclusion By mass use of preventive chemotherapy, Burkina Faso may have eliminated schistosomiasis as a public health problem in eight regions and controlled schistosome-related morbidity in another three regions.


Introduction
Human schistosomiasis is endemic in 78 countries or territories. 1,2 It has been estimated that, in 2013, there were nearly 261 million people -including about 240 million in Africa -who required preventive chemotherapy because they were at risk of schistosome infection. 1 Following the 2001 World Health Assembly resolution WHA54.19, 3 several endemic countries in Africa launched national programmes for the control of schistosomiasis. 4,5 These programmes are largely based on preventive chemotherapy with praziquantel and are targeted at school-age children and adults at risk. 6 In resolution WHA65.21, the World Health Assembly called on all countries with endemic schistosomiasis to intensify their control programmes and, where appropriate, to initiate campaigns for the elimination of schistosomiasis. 7 The West African country of Burkina Faso is divided into 13 administrative regions (Fig. 1). Some form of human schistosomiasis is thought to be endemic in every one of the country's 63 health districts. [9][10][11] Although urogenital schistosomiasis -caused by Schistosoma haematobium -occurs throughout the country, intestinal schistosomiasis -caused by Schistisoma mansoni -is mainly confined to the southwest of the country. 9,11 Surveys conducted before the 1980s, showed that the prevalence of S. haematobium was very high, with focal prevalence up to 100% of people surveyed in the eastern part of the country. 9 Over the same period, S. mansoni infection was found in up to 79% of people surveyed in the Hauts Bassins and Sud-Ouest regions. 9 Burkina Faso established a national programme for the control of schistosomiasis and soil-transmitted helminths in 2004, with funding from the Schistosomiasis Control Initiative. 5,12,13 This programme's main objective was to use mass administration of praziquantel to prevent human schistosomiasis. National mapping surveys 14 led to the country being divided into a hyper-endemic zone -comprising the 19 health districts that make up the Boucle du Mouhoun, Nord, Sahel and Sud-Ouest regions -and a meso-endemic zone -comprising the country's other 44 health districts. In 2004, baseline data were collected from children attending 16 randomlyselected primary schools in the four regions of the hyperendemic zone. Depending on the study region, the observed prevalence of S. haematobium infection varied from 18.4% to 84.2% and the observed intensity of such infection -among all children investigated -varied from 39.4-126.9 eggs per 10 ml urine sample. 15 Biennial mass administration of praziquantel to school-age children began in the hyper-endemic zone in 2004 and in the meso-endemic zone in 2005. 13,15 Since 2006, adults who are considered to be at risk have also been targeted.
In 2007, Burkina Faso's national programme for the control of schistosomiasis and soil-transmitted helminths became part of a national integrated programme against neglected tropical diseases. At the beginning of 2013, four and five rounds of mass praziquantel administration were done in the mesoendemic and hyper-endemic zones, respectively. To assess the impact of these rounds and plan for the next phase, primary-school children at 22 sentinel sites were tested for schistosomiasis in 2013. Here we present the results of the assessment and discusses possible future strategies for the elimination of all forms of schistosomiasis from Burkina Faso.

Ethical considerations
The assessment survey formed part of the monitoring and evaluation activities of the programme. It was conducted by the national monitoring and evaluation team and was authorized by the Ethics Committee of the Ministry of Health of Burkina Faso. Before the survey, written informed consent was obtained from the head teacher of each study school and verbal informed consent was obtained from a parent or guardian of each child. Each child was given a unique identification number so that data could be analysed anonymously.

Mass drug administration
Although the national strategy included biennial praziquantel rounds, the amalgamation of the national programme for schistosome control into the integrated programme for the control of neglected tropical diseases led to some scheduled administrations being missed (Table 1).
In each round of praziquantel administration, trained health workers treated children of school age either in schools or -for the children who were not attending any school -in communities. 13 A dose pole was used to measure children's height and determine the required dose. 13

Baseline data
For our analyses, we used baseline data that were collected for the national programme for schistosomiasis control. These data were collected from 16 randomly selected primary schools in the hyper-endemic zone, in 2004 -before the first mass administrations of praziquantel. 15,17 Stool and urine samples were collected from about 100 randomly selected children aged 7-14 years -half of them girls -at each of the 16 schools and checked for the eggs of S. mansoni and S. haematobium, respectively.

Impact surveys
In 2008, the national Ministry of Health designated 22 sentinel sites for the monitoring and evaluation of the schistosomiasis programme in Burkina Faso: three in Hauts Bassins, two each in Boucle du Mouhoun, Centre-Est, Centre-Nord, Centre-Ouest, Centre-Sud, Est, Nord, Sahel and Sud-Ouest and one in Cascades. These sites, all of which were schools, were purposefully selected across 11 of the country's 13 health regions to give a fairly even geographical distribution across the country (Fig. 1). Cross-sectional surveys in each sentinel site were done in 2008 and 2013. In each of these surveys, stool and urine samples were collected and examined for schistosome

Parasitological examination
One urine sample and one stool sample from each child were collected in separate containers with unique identification numbers, and sent to a laboratory for examination on the day of their collection. Urine samples were filtered through a nylon filter (pore size 12 μm; Merck Millipore, Billerica, United States of America) and the number of eggs counted under a microscope. For specimens of less than 10 ml, the volumes were measured before filtration and the number of eggs per 10 ml calculated. Intensity of S. haematobium infection was expressed as the number of eggs per 10 ml of urine examined. The Kato-Katz method was used to check stool samples for S. mansoni eggs. On the day that the sample had been collected, duplicate slides were prepared from each sample and examined. Eggs were counted and intensity of infection was expressed as the number of eggs per gram of faeces.

Data analysis
The data collected in 2013 were entered into spreadsheets and double checked by biomedical technicians. As we could not access the full data set from the 2008 assessment, we compared the data collected in 2013 with a descriptive summary of the data collected in 2008 8 and the data collected in the 2004 baseline survey. 15 Prevalence and intensities of infection -and their corresponding 95% confidence intervals (CI) -were calculated using SPSS version 19 (IBM, Armonk, USA). When calculating the overall values for prevalence and intensity of infection across the 11 regions with sentinel sites, the samples were adjusted with weighting according to the proportion of the national population represented by each regional population in 2013 -as projected from the results of the 2006 census. The complex-samples module of the SPSS package was used -with regions as the strata and schools as clusters -to take account of the clustering of the investigated school children. In general, our comparisons of the intensity of infection were based on the arithmetic mean egg counts for all subjects. Children were considered to have heavy S. haematobium infections if they had at least 50 eggs per 10 ml of urine. 18 Children with more than 399 eggs per gram of faeces were considered to have heavy S. mansoni infections. 18 Prevalence and intensities were compared using χ 2 and Kruskal-Wallis tests, respectively. The geographical coordinates of each sentinel site, as determined in Google Maps (Google, Mountain View, USA), were used to plot the site's approximate position on national maps drawn in ArcMap version 10 (ESRI, Redlands, USA). Costs of the mass praziquantel administration were estimated using financial data collected in 2013 and 2014 (G Liebowitz, Helen Keller International, unpublished data, 2015). Before 2013, the relevant financial data were either incomplete or unavailable.  Table 2 summarize the prevalence of the S. haematobium and S. mansoni infections observed among the 3514 school children -1748 boys and 1766 girls -aged 7-11 years who provided stool and urine samples at the 22 sentinel sites. Table 2 also summarizes the mean egg counts. Although the adjusted overall prevalence of S. haematobium infection was 8.76%, the prevalence of such infection ranged from 0.0% (0/160) to 56.3% (90/160) according to sentinel site (median: 2.5%). The children from Centre-Est, Est and Sahel had significantly higher prevalence of S. haematobium infection than the children from the other eight regions (P < 0.001). After adjustment for the sex distribution of the national population, the proportions of the boys (9.90%) and girls (7.65%) found infected with S. haematobium were similar (P > 0.05).

Situation in 2013
The adjusted arithmetic mean intensity of S. haematobium infection -among all children investigated -was 6.0 eggs per 10 ml urine. The mean egg counts for the children from Boucle du Mouhoun, Centre-Est and Sahel were significantly higher than those for the children from the other eight regions (P < 0.001). Boys were generally more heavily infected than girls ( P = 0.013). The adjusted overall prevalence of heavy S. haematobium infection was 2.82%. The Centre-Est (8.75%; 28/320) and Sahel regions (11.56%; 37/320) had the highest percentages of children infected. In six of the regions included in the as-  sessment, less than 1% of the children investigated had S. haematobium infection. Overall, 3.83% of the boys investigated and 1.8% of the girls were found heavily infected with S. haematobium (P > 0.05).
S. mansoni was only detected in the Hauts Bassins region -with a prevalence of 8.75% (42/480) and an arithmetic mean egg count of 7.7 per gram of faeces -and the Centre-Sud region -with a prevalence of 0.31% (1/320) and an arithmetic mean egg count of 0.15 per gram of faeces.

Data for 2004 and 2008
The prevalence of S. haematobium recorded in the 22 sentinel sites during the national survey in 2008 was, in general, markedly higher than that recorded in 2013 (Fig. 1). Table 3 shows the baseline data collected in 2004 from the Boucle du Mouhoun, Nord, Sahel and Sud-Ouest 15 and the corresponding data, from the same four regions, from the assessment in 2013. As these two sets of data were collected in different sites and different numbers of sites -and the exact locations of the sites surveyed in 2004 could not be determined -we made no direct statistical comparisons between the two data sets and could not produce a map of the baseline data to match our other figures. However, the data in Table 3

Drug distribution costs
At the beginning of the national programme for schistosomiasis control, the cost of a round of mass treatment with praziquantel was estimated to be 0.32 United States dollars (US$) per child treated. 13

Discussion
After a decade of preventive chemotherapy, progress has been made in Burkina Faso in the control of schistosomiasisat a modest cost. In the 2013 assessment, the prevalence of schistosome infection among school children was found to be below 5% in five of the 11 included regions -and below 10% in eight of the regions. In the two regions not included in the 2013 national assessment -i.e. Centre and Plateau Central -the Ministry of Health also found the prevalence of S. haematobium infection to be below 5% in 2013. 19 In 2013, therefore, recorded prevalence of S. haematobium infection remained high -i.e. above 18% -in only three regions: Centre-Est, Est and Sahel. In addition, the heavy S. haematobium infections that are associated with most of the morbidity of urogenital schistosomiasis were only rarely detected -i.e. in less than 1% of the children checked in eight regions included in the 2013 national assessment or the smaller ministry of health study. 19 According to the criteria of the World Health Organization (WHO), 20 by 2013, these eight regions had eliminated urogenital schistosomiasis as a public health problem. By the same year, another three regions -i.e. those in which 1-5% of children surveyed were found to have heavy S. haematobium infections -had reached the target of controlling the morbidity caused by such schistosomiasis. 20 Despite the generally encouraging trends revealed by our analyses, there were some causes for concern. For example, the Centre-Est and Sahel regions appeared to have failed to control urogenital schistosomiasis by 2013. At one Centre-Est sentinel site, the prevalence of S. haematobium infection was much higher in 2013 (56.3%) than in 2008 (14.4%). Similarly, in a Hauts Bassins sentinel site, the prevalence of S. mansoni infection recorded in 2013 (26.3%) was higher than that recorded in 2008 (17.9%). At several sites in the Centre-Est and Est regions, the prevalence of S. haematobium infection recorded in 2013 was similar to that recorded in 2008. There are at least three possible reasons for an increase or persistence in the prevalence of infection. First, the frequency of treatment may be inadequate, especially in areas with particularly high levels of infection and transmission. Second, even though the overall coverage of mass administration may appear adequate, focal treatment coverage may not be satisfactory. Third, there may be particular social or environmental factors that are supporting focal transmission despite the benefits of the preventive chemotherapy. The re-sults of ongoing research in the Centre-Est region may help to explain the local persistence of schistosomiasis foci.
After studying the results of the 2013 assessment and the relevant WHO recommendations, 20,21 the managers of the national programme against neglected tropical diseases have recently reviewed the progress achieved, set objectives for the next phase of the programme and increased treatment frequency in some areas. The objectives are now to use mass drug administrations: (i) biennially, to interrupt the transmission of S. haematobium and S. mansoni in the Cascades, Centre, Centre-Nord, Centre-Ouest, Centre-Sud, Nord, Plateau Central and Sud-Ouest regions; (ii) annually, to control schistosomerelated morbidity or eliminate schistosomiasis as a public health problem in the Boucle du Mouhoun, Est, Hauts Bassins and Sahel regions; and (iii) biannually, to control schistosome-related morbidity or eliminate schistosomiasis as a public health problem in the Centre-Est region. 8 At the same time, schistosomiasis surveys are to be extended to nonsentinel areas to check that the trends seen at the sentinel sites are nationally representative and identify any foci of transmission that have not been recognized previously.
Although it has long been known that regular treatment with praziquantel can prevent both the severe and milder morbidity associated with schistoso-  6,17,22,23 there is an indication in the data from Burkina Faso that it may also lead to the elimination of schistosomiasis in certain transmission settings. Burkina Faso is a land-locked country that is usually divided into three climate zones: the north-Sudanese in the south, the sub-Sahelian in the centre and the Sahelian in the north -with annual rainfall of 900-1200, 600-900 and 400-600 mm, respectively. 24 In much of the country, water is a scarce resource. Surface water consists of two main rivers that carry water all year around -i.e. the Mouhoun and Nakambe rivers -several perennial water reservoirs and some seasonal water bodies. 25 In previous studies in Burkina Faso and Niger, the prevalence of S. haematobium infection was found to be reduced by one round of mass drug administration and then to remain low for another 2-3 years in the absence of further chemotherapy. 15,26,27 In these countries, which have relatively little perennial surface water, the risks of re-infection after mass drug administration with high coverage are relatively low. This may explain why repeated biennial mass drug administration in Burkina Faso appears to have effectively eliminated schistosomiasis as a public health problem in at least eight regions.
Apart from increasing treatment frequency where necessary, other complementary public health interventions may need to be considered in Burkina Faso -particularly in the persistent foci with high prevalence of infection. WHO has recommended comprehensive measures for eliminating neglected tropical diseases 28 and complementary measures that could be introduced in a phased approach to the control of schistosomiasis 20 -e.g. health education, improved sanitation and access to clean water, environmental snail control and focal use of molluscicides. [29][30][31][32] In Burkina Faso, snail management and operational research on molluscicide use are needed. Closer collaboration between the inte-grated programme for the control of neglected tropical diseases and the education and communications sectors are needed to support behavioural change communications to change watercontact behaviour and minimize the risk of infection. The integrated programme and the water, sanitation and hygiene sectors also need to work together to reduce transmission. ■