Schistosomiasis mansoni in Burundi: progress in its control since 1985.

Schistosomiasis mansoni in Burundi: progress in its control since 1985 D. Engels,1 J. Ndoricimpa,2 & B. Gryseels3 Described is the evolution of the schistosomiasis control programme in Burundi since 1985. A single round of selective population chemotherapy was carried out in the Rusizi Plain and the Bugesera focus from 1985 to 1990. The prevalences and intensities of infection as well as the number of symptomatic cases detected in general health services decreased considerably. Annual sample surveys in the treat- ed areas showed, however, that these improvements were rapidly reversed by reinfection of the demographically changing population. Since repeated selective population chemotherapy was not sus- tainable in the long term, a primary health care approach was adopted. In areas with good access to basic health services, approximately 10% of all schistosomiasis cases now receive treatment annually through this approach. Yearly selective chemotherapy in primary schools in suburban Bujumbura reduced the prevalence of schistosomal infection among pupils from 23% to 9% over the period 1984-90, and this programme has now been extended to highly endemic areas in Imbo-Sud. Focal snail control produced disappointing results, and emphasis has therefore shifted towards health educa- tion and environmental control of transmission. Introduction In Burundi intestinal schistosomiasis caused by Schistosoma mansoni is essentially a man-made problem: land reclamation, agricultural development, and human resettlement have largely contributed to its spread since the 1950s. The history of schistoso- miasis in Burundi and the development of a national control programme based on extensive epidemiologi- cal and operational studies have been reviewed elsewhere (1). The present article presents an over- view of this programme since its implementation in 1985 as well as the resulting shifts in control strat- egy. Burundi is a small (area: 28 000 kM2) but dense- ly populated country (1990 census population: 5.3 million). The centre of the country consists of high- lands (altitude: 1500-2000 m), which towards the west form mountains of height up to 2600 m. The highlands are surrounded by lowland areas (altitude: 700-1400 m): the Imbo Plain in the west, the Buge- sera depression in the north, and the Moso Plain in the east (Fig. 1). Schistosomiasis is currently con- fined to these lowlands. I Head, Schistosomiasis Control Programme, B.P. 337, Bujum- bura, Burundi. Requests for reprints should be sent to this author. 2 Technical Adviser, Schistosomiasis Control Programme, Bujumbura, Burundi. 3Senior Lecturer, Parasitology Laboratory, Faculty of Medicine, University of Leiden, Leiden, Netherlands. Reprint No. 5375 From an operational point of view, four distinct endemic areas can be considered: the Rusizi Plain; the suburban focus of Bujumbura; Imbo-Sud, border- Fig. 1. Maps of Burundi as a whole and of the three major rural areas endemic for schistosomiasis. Popula- tion at risk (1990): Rusizi Plain - 180 000; Bujumbura - 130 000; Imbo-Sud - 160 000; Bugesera - 70 000; and Moso - 200 000. Shaded circles: <5 km from a health centre. Rusizi Plain Bugesera Rugombo &; *b hAmn RWANDA Imbo-Sud sKma 1 tA_ _ -_ _Om _ ar~~~~~~5O5c __ >2000 m >,700m° 50m Bulletin of the World Health Organization, 71 (2): 207-214 (1993) © World Health Organization 1993 207 D. Engels et al. ing Lake Tanganyika south of the capital; and the Bugesera depression, where schistosomiasis is main- ly concentrated around Lake Cohoha and Lake Rwi- hinda. The number of people living in these endemic areas grew from 450 000 in 1985 to 520 000 in 1990. The characteristics of the endemic areas and popula- tions have been reviewed previously (1). Because of steady demographic pressure, human resettlement and agricultural development were ini- tiated recently in the eastem part of the Bugesera depression (around Lake Rweru) and the Moso Plain. Although not yet a public health problem, schistoso- miasis occurs in these areas and is likely to extend in the coming years. The number of people at risk is estimated to be 220 000. Administratively, Burundi is divided into 15 provinces and 114 communes, each made up of sev- eral collines, which, in tum are divided into sous- collines (dispersed habitat) or transversales (organi- zed habitat - Rusizi Plain). The ratio of health centres to population is approximately 1:14000 in the Rusizi Plain, 1:20000 in suburban Bujumbura, 1:35 000 in Imbo-Sud, and 1:40 000 in Bugesera. The annual recurrent budget for health in Burundi is US$ 1.6 per inhabitant, not including salaries. Drugs and reagents account for 27% of this budget. Materials and methods Morbidity control In the initial attack phase of the programme, selec- tive chemotherapy was used as the strategy for mor- bidity control. Screening was based on examining from each inhabitant a single 25-mg Kato smear (2), prepared from a fresh stool sample and left to clear for 45 minutes (3). All schistosome eggs were count- ed; the presence of eggs of other common intestinal helminths (hookworm, ascaris, trichuris, and taenia) was recorded semi-quantitatively. Persons with at least one schistosome egg on a single slide were treated with a single dose of praziquantel (40 mg/kg body weight); individuals with hookworm, round- worm, or whipworm infections were treated with a single dose of 50-150 mg of levamisole or 500 mg of mebendazole; and individuals with tapeworm were treated with a single dose of praziquantel (20 mg/kg body weight). This strategy was applied to all communities and age groups in rural endemic areas. Chemotherapy campaigns were carried out once in the Rusizi Plain and Bugesera. In Imbo-Sud they were repeated sever- al times in highly endemic collines, since systematic coverage was not feasible owing to the dispersed habitat and unstable demography. In the endemic suburbs of Bujumbura the same strategy was used annually for the high-risk group of children aged 6-15 years, through a programme in primary schools. The impact of selective chemotherapy on infec- tion was monitored by means of annual sample sur- veys in the Rusizi Plain (Rugombo, Buganda and Gihanga, Fig. 1). The samples comprised every 10th household, starting with a random number, in each sous-colline or on each transversale. If the house- hold selected refused to participate, an adjacent household was included. In this manner a self- weighting 10% sample was obtained, taking into ac- count the highly focal distribution of schistosomiasis. Chemotherapy campaigns and surveys were car- ried out by two specialized mobile teams; one in the Rusizi Plain and Bugesera (6 microscopists) and the other in Bujumbura and Imbo-Sud (4 microscopists). Quality control was performed by systematic random re-examination of about an eighth of the daily num- ber of Kato slides examined. During the maintenance phase of the program- me, the diagnostic capacity of health centres in endemic areas was improved by introducing the Kato method and well-defined clinical indications for stool examination (direct examination + Kato slide). After the initial training of microscopists in the health centres, quality control was performed once a year and refresher courses were organized if neces- sary. The central mobile teams were integrated into the regional health services, and their services could be called upon to deal with particular problems. Snail control Snail populations were assessed using a man-minute method (4 collectors searching for 15 minutes). Niclosamide was applied using a knapsack sprayer for shallow stagnant waters (concentration of 1.5 x 103 mmol/l (0.6 ppm)) or using a constant-head applicator for flowing waters (concentration of 5.1 x 10-3 mmol/l (2.0 ppm) over 6 hours). Environmental control of transmission In Burundi safe water is commonly supplied to low- land areas by networks of piped water, collected on the (rainy) mountain crests. Standard laundry units consist of four basins, each with a tap and a simple draining system. Shower units consist of simple per- forated pipes, fed by a tap and adjusted to waist or shoulder level to allow people to clean themselves after coming back from working in the fields. Public toilets consist of blocks of four ventilated pit- latrines. Private latrines are promoted by selling concrete slabs at a subsidized price (ca. US$ 10); once the latrine is built by the owner, it is checked WHO Bulletin OMS. Vol 71 1993208 Control of schistosomiasis in Burundi by a health assistant before the vent piping is sup- plied. Flipcharts on the following were developed and distributed to health centres: schistosomiasis, the importance of safe water and sanitation, environmen- tal sanitation, and intestinal helminths. A 25-minute educational video-film was produced and is shown annually to the senior classes of all primary schools in areas of Burundi where schistosomiasis is endemic. More recently, simple information panels and posters, leaflets, and T-shirts promoting early diagnosis and prevention of schistosomiasis have been prepared for use in a more comprehensive large-scale information-education-communication (IEC) campaign. Data collection and analysis Standard procedures and record forms were used throughout the programme. Simple indices were cal- culated manually by each team leader to permit day- to-day monitoring. Data were further processed and analysed by computer using dBase III+, BASIC, and SPSS PC+ software. Mean egg loads were calculated as geometric means of positive individual egg loads per gram of faeces (epg). Egg count categories were defined as follows: 1-100 epg (1-2 eggs per Kato slide); 101-400 epg (3-10 eggs per Kato slide); >400 epg (>10 eggs per Kato slide). The sequence of control measures in each of the four endemic areas is summarized in Table 1. Table 1: Sequence of control measures in the four areas of Burundi endemic for schistosomiasis, 1985-91 Area 1985 1986 1987 1988 1989 1990 1991 Rugombo Gihanga Rukana/Buganda Mutimbuzi Mpanda Rusizi Plain Single selective chemotherapy + monitoring Ea PHC programmeb Snail control: "intensive" Environmental control of transmission (southern part) Bugesera focus E "extensive" E Single selective chemotherapy PHC programme Annual selective chemotherapy of primary school children E PHC programme Imbo-Sud Repeated selective chemotherapy in highly endemic collines E "Extensive" snail control PHC programme a E = evaluation. b PHC = primary health care. WHO Bulletin OMS. Vol 71 1993 E E Bujumbura > ( 209 D. Engels et al. Results From 1985 to 1990 about 160 000 people participa- ted in the selective population chemotherapy cam- paigns in rural areas and about 70 000 stool exam- inations were performed during the school program- me in urban areas (Bujumbura, Rumonge). The cam- paigns in the Rusizi Plain took place between 1985 and 1989. The prevalences and intensities of schisto- somal infection measured during these campaigns are summarized in Table 2. The average participa- tion rate was 80%, and this varied from 67% to 97% in the various collines. The rate was highest in the northern part of the plain where the population was more stable. Bugesera was covered in 1989-90. The measured prevalences and intensities of infection are summarized in Table 2. Here, the participation rates varied from 54% to 97% in the various collines, with an average of 83% for the whole area. In Imbo- Sud repeated selective chemotherapy was used from 1985 to 1990 in the highly endemic collines just north of Rumonge. The prevalences of schistosomal infection measured during these campaigns are sum- marized in Table 3. The total cost of the selective treatment campaigns was approximately US$ 1 per person examined over the study period. The cost of praziquantel accounted for 50% of the running costs of the campaigns. We were able to evaluate the effect of this strat- egy on the passive detection of cases in general heal- th services in the northern part of the Rusizi Plain (Rugombo). The results are shown in Fig. 2. The evolution of the prevalence of infection after selective population chemotherapy in the three monitored communes in the Rusizi Plain is shown in Fig. 3. Two years after intervention, the prevalence had reverted to 65%, 100%, and 85% of its initial level in Rugombo, Gihanga and Buganda, respec- tively. The proportions of people treated, examined (but found negative), or not examined during the treatment campaigns are also shown. The proportion of infected people who were new to the area became important from the third post-intervention year onwards. Table 2: Prevalences and Intensities of schistosomal infection detected population treatment campaigns in the Rusizi Plain and Bugesera, Burundi during selective % of individuals % of individuals Commune Average Mean egg load excreting excreting or area prevalence (%) (epg) >100 epg >400 epg Rusizi Plain Rukana 15 (2-36) a 95 (40-136) 4 (0-9) 1 (0-2) Rugombo 32 (6-63) 122 (77-190) 13 (1-38) 4 (0-16) Buganda 13 (3-48) 114 (75-268) 5 (1-17) 2 (0-11) Gihanga 20 (6-31) 114 (72-160) 7 (2-15) 2 (1-7) Mutimbuzi 33 (8-63) 178 (86-308) 16 (3-38) 8 (1-23) Mpanda/Buringa 34 (5-56) 189 (81-294) 21 (2-36) 9 (0-19) Bugesera Area A 19 (10-28) 133 (82-199) 10 (4-17) 3 (0-6) Area B 7 (1-13) 82 (58-143) 2 (0-6) 0.5 (0-2) Area C 4 (0-6) 116 (40-128) 2 (0-3) 0.5 (0-1) a Figures in parentheses are the range, based on an analysis per sous-colline. Table 3: Prevalences of schistosomal infection detected during repeated selective population treatment campaigns in various collines in Imbo-Sud, Burundi, 1985-90 Overall detected % prevalences Colline 1985 1986 1987 1988 1989 1990 Magara - 19 (11-3) a 5 (2-0) - - Kagongo 17 (8-2) 16 (10-4) - 27 (15-6) 19 (9-4) 22 (12-4) Kizuka 33 (20-8) 18 (10-2) 22 (11-4) - 31 (18-8) Mwange 22 (11-4) 10 (4-1) - - 18 (12-8) Rukinga - - 21 (13-3) 27 (15-6) 23 (12-4) a The first figure in parentheses is the % of infections with >100 epg, while the second figure shows the % of infections with >400 epg. WHO Bulletin OMS. Vol 71 1993210 Control of schistosomiasis in Burundi Fig. 2. Effect of selective population treatment on pas- sive detection of schistosomiasis cases in general health services in Rukana/Rugombo (Rusizi Plain). 700_700 selective chemotherapy | co 500 4) 6 11 A0 198019811982198319841985 198619871988 Fig. 3. Endemic situation after selective chemotherapy in three communes in the Rusizi Plain: evolution of pre- valences and proportions of people (A) treated; (B) exam- ined but found negative; and (C) not examined during the selective treatment campaigns. T = selective population chemotherapy. '85 '86 '87 '88 '89 MA EBMC Yearly selective chemotherapy of primary school children in the schistosomiasis-endemic sub- urbs of Bujumbura reduced the overall prevalence of schistosomiasis among such children by 61% (from 23% to 9%) over a period of 6 years (1984-90). The number of children examined increased from 7462 in 1984-85 to 15 886 in 1989-90. The effective cover- age of school-age children (6-15 years) in these sub- urbs by the programme increased from 35% in 1984-85 to 61% in 1989-90. To establish the maintenance phase of the pro- gramme, equipment and training for early diagnosis and treatment were provided to health centres in the Rusizi Plain in 1989, in the Bugesera focus in 1990, and in Bujumbura and Imbo-Sud in 1991. This considerably increased the number of cases detected in general health services: in the Rusizi Plain from 970 in 1988 to 3584 in 1991; and in Bugesera from 63 in 1988 to 180 in 1991. Of these cases, 93% pre- sented with gastrointestinal symptoms, 6% with vague complaints, and only 1% with symptoms in- dicative of hepatosplenic disease. Quality control monitoring of the microscopic diagnosis in health centres in 1990 in the Rusizi Plain found reproduci- bility levels of 83% and 71%, respectively, for quali- tative and quantitative diagnosis, compared with the results obtained by a specialized team. The extra cost for improved diagnosis and treatment of schistoso- miasis and intestinal helminths in endemic areas was estimated to be US$ 0.02 per inhabitant per year, the cost of praziquantel accounting for 85% of this. Since the programme led to an improvement in the diagnosis and treatment of infectious intestinal pathology in general, part of this extra cost would presumably be offset by more appropriate use of other drugs, e.g., metronidazole and antibiotics. In 1987-88, intensive focal mollusciciding was performed in three communes of the Rusizi Plain (Rugombo, Buganda, Gihanga). All important sites of intense man-water contact (n = 130) in these communes were examined systematically every 2 months and treated if live Biomphalaria snails were found. Although on average 97% of the treated sites no longer had live snails immediately after treat- ment, most positive sites were rapidly recolonized, and 2 months after each treatment considerable populations of adult snails were again found in most of these sites. Although the recolonized snail popula- tions never reached the initial levels again, they were sufficient to maintain schistosomiasis transmission. The cost of this type of snail control was US$ 0.20 per inhabitant per year, expenses for niclosamide accounting for 70% of this. Supplies of safe water and sanitation were intro- duced to the southem part of the Rusizi Plain (Gihanga, Mpanda, Mutimbuzi communes-80 000 people in 1990). From mid-1986 to mid-1991 a new network of piped water was installed and about 280 community standpipes were repaired or installed, increasing the ratio from 1 standpipe per 300 fami- lies to 1 per 60 families. Also, the following were constructed: 27 laundry units, near traditional washing places; 11 shower units, at the exits of rice fields; 62 blocks of public toilets, at schools and markets; and 12 footbridges, at frequently used crossing sites. About 4400 concrete latrine slabs WHO Bulletin OMS. Vol 71 1993 211 D. Engels et al. were distributed to 25% of all families in the area. The cost of this programme was US$ 32 per inhabi- tant (US$ 22 for the new piped water system and US$ 10 for the additional infrastructure), representing about 2-3% of the planned investment in new irrigation schemes in the area. Surveillance surveys were carried out in the new risk areas in the eastern part of Burundi. In Nyanza-Lac (1987) a population-based survey showed that the overall prevalence of schistosomia- sis was 6% (range per sous-colline: 2-11%), almost identical to the prevalence in 1984. Among the population living along the shores of Lake Rweru (Fig. 1), 4% (range per sous-colline: 2-8%) was infect- ed in 1989. A health-centre-based survey in the Moso Plain (1989) found few cases (6 out of 1005 people examined) scattered over the area, which is not much different from the situation in 1984 (1). Discussion Almost 5 years of fieldwork were required to carry out the selective chemotherapy campaigns in all the endemic rural areas. Monitoring surveys showed that communities rapidly became reinfected. In the first 2 years following the interventions, the majority of schistosomiasis cases in each of the study samples had participated in selective population treatment campaigns, but were found to be negative upon screening. These findings are consistent with the conclusions of preliminary operational studies (4) and can, at least in part, be explained by the low sen- sitivity of the screening method for relatively light infections. Under operational conditions this is com- pounded by the demographic changes in the study area. Indeed, despite the satisfactory participation rates only 60% of the 1990 population in the rural areas covered took part in the chemotherapy cam- paigns. Repeated interventions are thus necessary to ensure continuity of morbidity control. Finan- cial, logistic and organizational limitations are major constraints to the long-term sustainability of such a strategy. Therefore, an attempt was made to improve the capacity of morbidity control using a primary health care approach. The results have been encouraging. At the current rate of detection of schistosomiasis in the Rusizi Plain, approximately 10% of all cases (detectable with a single 25-mg Kato slide) receive treatment every year; comparison with selective population treatment (participation rate, 80%; 16% of cases treated yearly; demographic changes not taken into account) showed that the annual yield of cases detected by basic health services was 60%, at an eighth of the cost per inhabitant. Although the signs and symptoms of intestinal schistosomiasis are non-specific, and mixed intestinal infections are common, we can assume selection in favour of symptomatic cases. An additional advantage of a primary health care strategy is its ability to keep pace with demographic changes. A further increase in the demand for early diagnosis and treatment can be expected once the IEC campaign becomes fully operational. In some endemic areas of Burundi (Bugesera, Imbo-Sud) the impact of this type of strat- egy is still slight because of poor access to health services; rapid improvement of this situation can be expected in view of the strong commitment of nation- al health authorities and funding agencies to increasing primary health care coverage. Some aspects need to be further investigated; for example, the importance of schistosomiasis-related hepatic morbidity in Burundi and the impact of chemotherapy on it. Community-based surveys have shown increased rates of (mild) hepato- and spleno- megaly related to schistosomiasis, but few or no cases of clinical portal hypertension (1). Health centres detect few cases of severe hepatosplenic disease. On the other hand, operational studies have not found a significant reduction in liver and spleen enlargement rates, even after repeated chemotherapy (5).a Recent ultrasonographic studies in the Gezira region of the Sudan have reported a high frequency of liver disease, its reduction by chemotherapy, and the unreliability of indirect assessment of schistoso- miasis morbidity (6, 7). The planned use of ultra- sonography in Burundi should clarify the situation and reveal whether additional control action is needed. In view of the results of the school programme in Bujumbura, the frequency of the selective chemo- therapy in two of the four endemic suburbs was reduced from 1990 onwards by restricting its use to only the first, fourth, and sixth classes in each school. This maintenance strategy has permitted extension of the school programme to risk areas in Imbo-Sud, where an unstable demographic and endemic situation, intensive reinfection, and rela- tively poor access to health services justify this extra intervention. Despite considerable efforts and its high costs, the results of focal snail control have been dis- appointing. This is largely because of poor weed control, lack of maintenance, and the operational difficulty of covering entire water networks, leading to rapid recolonization of treated sites by snails. a Gryseels, B. Morbidity and morbidity control of schistosomia- sis mansoni in sub-Saharan Africa. Ph.D. thesis, University of Leiden, 1990, pp. 137-216. WHO Bulletin OMS. Vol 71 1993212 Control of schistosomiasis In Burundi Therefore, the emphasis in snail control has shifted to proper engineering and maintenance, directed to- wards farmers and irrigation managers. Some oper- ational research in the use of plant mollusci- cides has also recently been initiated. The most satisfactory outcome of the program- me is perhaps the progress that has been made in the supply of safe water. Donor as well as national agen- cies have been persuaded to increase the quantity of water supplied to areas endemic for schistosomiasis. In the southem part of the Rusizi Plain community standpipes are now available to all 80 000 inhabi- tants within 500-700 m of their homes. The rest of the Rusizi Plain and part of Imbo-Sud (Nyanza-Lac) will be covered in the next few years. Additional facilities (such as shower units, laundry units and footbridges) are being provided near traditional sites of intense man-water contact. Such facilities are very popular when they are in good working order. However, maintenance is still a problem and local authorities are trying to find adequate solutions to this difficulty; families have recently started paying an annual tax (US$ 0.40 per head) for this purpose. The private latrine programme started off well. Sub- sidized prices and credit facilities are a prerequisite for its success, although they may hinder the pro- gramme's viability after withdrawal of external funds. Although it has so far not been possible to evaluate the effect on transmission of schistosomia- sis of this environmental control programme, its general impact on well-being is clear. The health education programme has been limit- ed to specific target groups such as schistosomiasis patients and primary school children (8). Large-scale health education programmes indeed serve no purpose if alternatives to traditional water contact activities are not available. In the southern part of the Rusizi Plain, where everything has now been set up to meet increased demand in all aspects of schistosomiasis prevention, an extended IEC programme is planned. Its objective is to increase awareness of risk behaviour and to create simple responses favouring early diagnosis and treatment and the use of alterna- tive water sources and latrines. Surveys in the eastern lowlands of Burundi, an area where rice cultivation is widely promoted, have shown that the progression of schistosomiasis in recent years has been slower than initially feared. However, further surveillance is necessary. In conclusion, it can be stated that the strategy for schistosomiasis control in Burundi has evolved considerably since 1985. Morbidity control has shifted from systematic treatment of infections to early treatment of intestinal disease. The available health infrastructure has allowed the control strategy to be integrated to a high degree into basic health services, enabling it to be sustained and making it affordable for the national health budget. Whether hepato- splenic morbidity can also be reduced by this approach should be answered by the results of forth- coming ultrasound studies in Burundi. Efforts to reduce transmission of schistosomiasis by snail control met with limited success, and the costs were high. Emphasis has thus shifted towards environ- mental control of transmission, which is more funda- mental and beneficial, not only to health, but also to the quality of life in general. Acknowledgements We are grateful to members of the schistosomiasis control teams and peripheral health personnel. We also acknowl- edge the local communities and administrative authorities for their collaboration. The Schistosomiasis Research and Control Programme in Burundi was initiated by the Institu- te of Tropical Medicine, Antwerp, Belgium, and is suppor- ted by the Burundi Ministry of Health, the Belgian Techni- cal Cooperation Fund, and the European Development Fund. The programme is also funded by the UNDP/World Bank/WHO Special Programme for Research and Training in Tropical Disease (TDR). Provision of safe water and sanitation in the southern part of the Rusizi Plain was financed by the European Development Fund and the Kreditanstalt fur Wiederaufbau. Resume La bilharziose A Schistosoma mansoni au Burundi: progrbs de la lutte depuis 1985 La bilharziose intestinale a Schistosoma mansoni est presente au Burundi dans les regions de basse altitude. Un programme de lutte a ete ins- taure dans quatre de ces regions: la Plaine de la Rusizi, la zone semi-urbaine entourant Bujumbu- ra, lImbo-Sud et le Bugesera. Au d6but de sa mise en ceuvre en 1985, ce programme de lutte 6tait base sur: a) la chimioth6rapie s6lective, appliquee a toute la population en milieu rural et aux enfants scolarises (6-15 ans) en milieu semi- urbain; b) la lutte cibl6e contre l'h6te intermediaire et c) I'amelioration des conditions d'hygiene. L'effet favorable de la chimiotherapie s6lective (cout: 1 $US par personne examin6e) sur la pr6- valence et l'intensit6 des infestations ainsi que sur le nombre de cas de bilharziose d6tectes dans les services de sante g6n6raux, a et6 d6montr6. Pourtant, la reinfestation rapide dans une popula- tion changeante, associee a l'effet de la sensibilite r6duite du test diagnostique (Kato 25 mg) en cas d'infestation l6gere, impose la r6p6tition r6guliere de ce genre d'intervention afin d'assurer une WHO Bulletin OMS. Vol 71 1993 213 D. Engels et al. continuite dans le contr6le de la morbidite. Les chances de perennisation de cette strategie etant jugees faibles, une approche plus integr6e a ete testee. Les capacit6s de diagnostic precoce des services de sante de base ont ete ameliorees, entre autres par l'introduction de la m6thode Kato. Dans les regions ou I'accessibilite a ces services est bonne, on a pu ainsi multiplier par 4 le nombre de cas detectes. En d'autres termes, environ 10% des cas presumes y sont actuelle- ment detectes et traites chaque annee par cette strategie int6gr6e. Par rapport aux traitements de masse selectifs cela represente un potentiel de d6tection de 60% pour 1/8 du coOt. La morbidite ainsi detectee et trait6e 6tant essentiellement intestinale, l'importance de la morbidite hepato- splenique et son 6volution apres traitement meri- tent d'etre analys6es plus en dMtail a l'aide de 1'6chographie. A Bujumbura, le programme scolai- re a permis de r6duire la prevalence chez les 6co- liers de 23% en 1984-1985 a 9% en 1989-1990. Suite a ces bons resultats, la fr6quence du traite- ment s6lectif a pu etre r6duite dans deux des quatre quartiers semi-urbains. Le temps ainsi eco- nomise a permis d'6tendre ce programme aux zones fortement touch6es de l'Imbo-Sud ou des campagnes r6p6t6es de chimioth6rapie selective n'avaient pas donn6 les resultats escomptes et ou l'accessibilite aux services de sante de base est encore relativement faible. Malgre des efforts consid6rables associes a un coOt 6leve (0,2 $US par habitant par an), les resultats de la lutte contre l'h6te intermediaire ont ete decevants. Par consequent, une plus grande attention a 6te accordee a I'amelioration des conditions d'hygiene des populations concern6es. Dans ce domaine, beaucoup de progres ont ete realis6s dans la par- tie sud de la Plaine de la Rusizi, tant en ce qui concerne I'approvisionnement en eau potable que l'am6nagement d'infrastructures sanitaires favori- sant plus sp6cifiquement la pr6vention de la bil- harziose (lavoirs, rampes lave-pieds, passerelles). Bien que tres 6lev6 en valeur absolue (32 $US par habitant), I'investissement dans ce domaine ne represente que 2 a 3% des depenses pr6vues pour 1'extension des r6seaux d'irrigation. Grace a I'amelioration des conditions d'hygiene, les activi- tes d'education pour la sant6 du programme de lutte contre la bilharziose ont pu etre etendues a la population generale. Ces activites etaient jusque-la limit6es a des groupes-cibles comme les malades et les eleves des ecoles primaires. La surveillance de l'endemie dans les nouvelles regions a risque de l'est du pays a montre une extension moins rapide que celle initialement redoutee. References 1. Gryseels, B. The epidemiology of schistosomiasis in Burundi and its consequences for control. Trans- actions of the Royal Society of Tropical Medicine and Hygiene, 85: 626-633 (1991). 2. Katz, N. et al. A simple device for quantitative stool thick-smear technique in schistosomiasis mansoni. Revista do Instituto de Medecina Tropical de Sao Paulo, 14: 397-400 (1972). 3. Peters, P. et al. Quick Kato smear for field quantifi- cation of Schistosoma mansoni eggs. American journal of tropical medicine and hygiene, 29: 217-219 (1980). 4. Gryseels, B. et al. Repeated community-based chemotherapy for the control of Schistosoma man- soni: effect of screening and selective treatment on prevalences and intensities of infection. American journal of tropical medicine and hygiene, 45: 509-517 (1991). 5. Gryseels, B. & Nkulikiyinka, L. Two year follow-up of infection and morbidity with Schistosoma mansoni after treatment with different regimens of oxamni- quine and praziquantel. Transactions of the Royal Society of Tropical Medicine and Hygiene, 83: 219-228 (1989). 6. Mohamed-Ali, 0. et al. Ultrasonographical investi- gation of periportal fibrosis in children with Schisto- soma mansoni infection: reversibility of morbidity seven months after treatment with praziquantel. American journal of tropical medicine and hygiene, 44: 444-451 (1991). 7. Homeida, M. et al. Effect of antischistosomal chemotherapy on prevalence of Symmers' periportal fibrosis in Sudanese villages. Lancet, 2: 437-440 (1988). 8. Engels, D. & Mpitabakana, P. Schistosomiasis control and health education in Burundi. Tropical medicine and parasitology, 40: 226-227 (1989). 214 WHO Bulletin OMS. Vol 711993