Effect of a control programme on transmission of Schistosoma mansoni on an irrigated estate in Tanzania

Bull. Org. mond. Sante' 11972, 47, 325-330 Bull. Wld Hith Org. Effect of a control programme on transmission of Schistosoma mansoni on an irrigated estate in Tanzania A. FENWICK 1 Three methods were used to measure the level of transmission of infections of Schisto- soma mansoni on an irrigated sugar estate in northern Tanzania. The studies were carried out over a period of 3 years, during a programme for the control of the host snail Biom- phalaria pfeifferi. During the second and third years a mass diagnosis and treatment campaign against the infection was also carried out. Examinations for infection were made in newly employed subjects on arrival and after 6 and 12 months. Two studies were made in young children at an interval of 18 months, to determine age prevalence curves. In the third method, subjects were examined for infection 18 months after being found free from infection in a previous survey. Results are compared with data recorded in a previous study, made before snail control was commenced. The results suggest that the control programme has led to a great reduction in the incidence of S. mansoni on the estate. The Tanganyika Planting Company's 3 600-ha sugar estate is situated at Arusha Chini in northern Tanzania, in an area of low annual rainfall. Culti- vation of sugarcane is made possible by the use of water from the Weru Weru river to irrigate the land. However, the open canal irrigation system has created conditions that are highly favourable to Biomphalaria pfeifferi, the snail host of Schistosoma mansoni. In 1962 Foster (1967) showed that the incidence of S. mansoni infection was over 80% among field workers who had been employed for over 6 months and 56.9% in newly employed workers in the first 6 months. In young children, the incidence of infec- tion in December 1962 was 6% in children aged 0-1 years and over 50% in those aged 5 years. After several years experimentation with mollusci- cides (Crossland, 1963, 1967; Fenwick, 1970), a routine procedure for the control of B. pfeifferi was adopted from 1968 onwards (Fenwick, 1970). Several other measures were taken to reduce the prevalence of the infection including the provision of 1 Research Officer, East African Tropical Pesticides Research Institute, P.O. Box 3024, Arusha, Tanzania. Present address: London-Khartoum Bilharzia Project, P.O. Box 287, Khartoum, Sudan. piped water to the camps, a certain amount of health education, and, in 1969-70, a mass diagnosis and treatment campaign. Transmission of the infection to the residents on the estate during 1969 and 1970 was measured by three methods to determine whether the control programme had succeeded in reducing the incidence of infection from the high level of 1962. METHODS Snail control In the irrigation canals B. pfeifferi snails were controlled by the regular use of N-tritylmorpholine. A 16.5% emulsifiable concentrate formulation was applied at the headwaters of the irrigation systems to give a concentration of active ingredient in the main canal of 0.025 mg/litre 2 for 5 days every 7 weeks. Previous experiments (Fenwick, 1970) had shown that this regime would keep the canals and reservoirs snail free as indicated by the snail samp- ling technique used. The drainage ditches were treated with niclo- samide (approximately 8 (mg/l).h at about 12-week 2 Note: concentrations are expressed in SI units instead of ppm. 2903 325 326 A. FENWICK intervals'). The chemical was applied from knapsack sprayers into each ditch within 1 week of the de- tection of snails during routine sampling of selected stretches of ditches on the estate. When no irrigation wvas necessary because the rainfall was sufficient to water the crop, standing pools of water were sprayed with niclosamide to complete snail control within the estate boundaries. There are apparently few nearby sources of in- fection outside the estate boundaries, owing to the semi-arid nature of the area. No snails have been found in the Weru Weru River, though they may be present, but there may be sources of infection in Moshi, 22 km away, and at springs within a 20 km radius. As a check on the molluscicide programme, routine snail sampling was carried- out on sixteen 500-m stretches of canals and ditches every week during 1968, 1969, and 1970 by means of the mud- plug method described by Crossland (1962). Chemotherapy Throughout the 1960s, many of the inpatients at the estate hospital and any outpatients suffering from intestinal disorders were examined for S. mansoni and treated with either sodium D-[gluconato- (4-)]antimonate(I-) 2 or niridazole, if found to be infected. During 1968, 1969, and 1970 all hospital inpatients were examined and in addition a mass diag- nosis campaign was carried out involving about 90% of the estate residents. All new employees were screened for S. mansoni prior to employment and all infections found were treated with either niridazole or hycanthone (Fenwick & Jorgensen, 1972). Stool examination All stool examinations were carried out by the slow-speed centrifugation method described by Foster (1964). A small portion of the stool (2-3 g) was macerated in a test-tube with 10 ml of tap water and sieved through 4 layers of surgical gauze into a centrifuge tube. The suspension was then centri- fuged at 750-1 000 rev/min for 65 s and the super- natant liquid poured off. The sediment was taken up again in fresh water, shaken well, and the mix- ture centrifuged at the same speed for 20 s. Finally 1 I.e., the molluscicide was present at 8 mg/l for 1 hour, at I mg/l for 8 hours, or at some intermediate combination of concentration and time. 'Formerly known as " trivalent antimonyl sodium gluconate ". the sediment obtained was examined microscopi- cally by three different microscopists, each taking 1 drop and counting the number of eggs seen. This stool examination method was evaluated in a separate experiment (Fenwick, unpublished data, 1971) by comparing it with the quantitative filtration method of Bell (1963). It was shown that four consecutive examinations by the Foster method detected 50% of infections of 15 eggs/g detected by the Bell method, and a single examination by the Foster method detected 50% of infections of about 35 eggs/g. Incidence of S. mansoni in newly employed workers From March 1969 onwards every prospective employee at the estate was examined for S. mansoni infection on 4 consecutive days as part of his pre- employment medical examination. Those who were free from infection on arrival and were engaged by the company formed the subjects of the study, while those who were found to be infected with S. mansoni were treated but excluded from the study group. Workers found to be infected with S. haematobium were also excluded. The subjects were sought out and reexamined 6 months and 12 months after their initial exami- nation, in order to determine how many had contrac- ted S. mansoni during the two 6-month periods of their stay on the estate. Incidence of S. mansoni in young children In mid-1969, 18 months after the start of routine snail control, an age-prevalence curve for S. mansoni in young children was determined by studying 156 children, approximately 20 having been born in each of the years 1961-68 inclusive. Late in 1970, 3 years after the start of the snail control programme, a second age prevalence curve for S. mansoni was determined by studying 208 children, approximately 30 having been born in each of the years 1961-70 in- clusive. Four daily stool samples were examined by the Foster method to determine whether each child was infected. In this part of the study different children were examined each time data were collected, in contrast to the other two parts of the study in which the same individuals were followed for measurement of transmission. Since these determinations were carried out to measure transmission of S. mansoni on the estate, a correction factor had to be applied in drawing the curves to compensate for the effects of chemotherapy TRANSMISSION OF SCHISTOSOMA MANSONI ON AN IRRIGATED ESTATE IN TANZANIA administered during the 1969-70 mass diagnosis and treatment campaign. Several of the children who presented 4 negative stools in the 1970 survey had in fact been diagnosed and treated during the 1969-70 campaign. This did not mean that infec- tions were contracted during this period, but for the purposes of this transmission study these children were called positive on the assumption that in the absence of chemotherapy they would have been found infected during the 1970 prevalence survey. If this factor had not been applied the shift to the higher groups from the 1969 curve would not have reflected the true reduction in transmission but would have included the effect of chemotherapy, thus suggesting a greater reduction in transmission than had actually occurred. The children used in the surveys were selected at random from the housing cards kept for each indi- vidual resident on the estate by the camp clerks. The numbers were increased in the 1970 survey in order to obtain results that could be evaluated statistically. Incidence of S. mansoni in subjects previously free from infection In this part of the study, subjects who were chosen at random for the 1969 prevalence surveys of adults and children and found to be free from in- fection at that time were reexamined after 18 months to find out how many had contracted the infection in the intervening period. They were deemed to have become infected if, on 18-month follow-up, S. mansoni ova were found in one of four consecutive stool examinations by the Foster method, or if they had been found infected in that period during the mass diagnosis campaign, and had been treated. RESULTS Snail control During the 3 years 1968-70, 18 applications of N-tritylmorpholine were made to each of the irri- gation systems and 12 niclosamide treatments were applied to the drains. During the rains, temporary pools were sprayed with niclosamide as they were found. Weekly sampling of canals and reservoirs did not reveal any B. pfeifferi snails in the irrigation systems except on one occasion, when application in one large area was delayed for 3 weeks in January 1969 while fresh stocks of N-tritylmorpholine were awaited. B. pfeifferi snails were detected in the drains immediately prior to every treatment, but this was the criterion for treatment, and their pre- sence was a calculated risk in the scheme. Incidence of S. mansoni in new workers Between March 1969 and November 1969, 220 new arrivals were found to be free from infection and constituted the subjects of this study. Of these, 165 were reexamined 6 months after their initial examina- tion and 16 were found to be infected (Table 1). Of the 55 not reexamined, 29 had been discharged or had left the company before completing 6 months service, 20 were living outside the estate boundary, 5 refused to cooperate, and 1 was excluded because he had been treated for a S. haematobium infection. Of the 149 who were eligible for the second follow-up 12 months after arrival, 105 were examined and 5 were found to be infected (Table 1). Of the 44 missing, most had left the company's employment. The overall incidence of new infections was 9.7 % in the first 6-month period and 4.8% in the second. In field workers the incidence was 8.4% and 3.9% respectively. In other workers the incidence was 13.0% in the first 6 months and 7.4% in the second 6 months. It should be noted that the measurements relate to the first and second 6-month periods subse- Table 1. Follow-up examinations for S. mansoni made at 6 and 12 months on new workers free from infection on arrival Follow-up examinations 6 months, 6 months 12 months 1962-63 Foster, 1 967) no. eligible 220 149 no. examined 165 105 no. infected 16 5 - % infected 9.7 4.8 field workers no. examined 119 78 58 no. infected 10 3 33 % infected 8.4 3.9 56.9 other workers no. examined 46 27 no. infected 6 2 % infected 13.0 7.4 327 A. FENWICK Table 2. Age prevalence survey of S. manson in children, 1 July 1969 Year of birthAge group No. Infected childrenYear of birth g(years) studied 1968/69 0-1 24 0 0 1967 1-2 20 0 0 1966 2-3 20 0 0 1965 3-4 15 1 6.7 1964 4-5 21 2 9.5 1963 5-6 1 8 4 22.2 1962 6-7 1 8 8 44.4 1961 7-8 20 1 2 60 total 156 27 17.3 quent to the arrival of the subjects on the estate and not to a given 6 calendar months. The arrival of the 165 subjects was spread over the period March-November 1969 so that their follow-up examinations were from September 1969 to May 1970 (6 months) and from March-November 1970 (12 months). The possibility of seasonal variation in the rate of transmission on the estate has never been studied but it is believed that there was little while the study was in progress. Incidence of S. mansoni infection in young children The results of the mid-1969 prevalence survey in young children are shown in Table 2 and those of the 1970 prevalence survey in Table 3. The age prevalence results from Foster's 1962 survey and from the 1969 and 1970 surveys are summarized in Table 4. The significant aspects of the results are that in 1969 and 1970 there was a shift in the position of the curve towards the higher age groups and there was a zero infection rate in all children born after the start of the molluscicide trials in 1966. Incidence of S. mansoni in subjects previously free from infection Of the 258 subjects followed up 24 (9.3%) were found to be infected (Table 5). In fact 3 others were followed up and found to be free from infection, but they were excluded from the study because within Table 3. Age prevalence survey of S. mansoni in children at the end of 1970 Year Age No No. inf cted°but Infected children of birth grouP studied treated in total %(years) diedfected 1970 toa 1970 0-1 9 0 0 0 0 1969 1-2 20 0 0 0 0 1968 2-3 36 0 0 0 0 1967 3-4 32 0 0 0 0 1966 4-5 24 0 2 2 8.3 1965 5-6 29 3 1 4 13.8 1964 6-7 23 2 2 4 17.4 1963 7-8 19 4 2 6 31.6 1962 8-9 16 3 3 6 37.5 total 208 12 10 22 10.6 2 months of being found negative in 1969 they had been diagnosed as infected while in hospital and were treated. These were obviously light infections missed in the 1969 surveys. The 258 subjects were divided into occupational groups, from which it can be seen that the incidence was 15.0% in field workers, 16.3 % in other workers, 7.8% in wives, and 3.3% in children. The final column in Table 5 shows the annual incidence rate Table 4. Results of the age prevalence surveys of S. mansoni in children in 1962 (Foster, 1967), 1969, and 1970 Age group Percentage infected (years) in 1962 in 1969 in 1970 0-1 6 0 0 1-2 8.7 0 0 2-3 16.2 0 0 3-4 25.0 6.7 0 4-5 27.5 9.5 8.3 5-6 56.0 22.2 13.8 6-7 50.0 44.4 17.4 7-8 60 31.6 8-9 37.5 328 TRANSMISSION OF SCHISTOSOMA MANSONI ON AN IRRIGATED ESTATE IN TANZANIA Table 5. Results of 18-month follow-up examinations of subjects free from infection in the 1969 prevalence studies Number Infected subjects Annual Subjects followed incidence up No. % rate (%) field workers 60 9 15.0 10.2 other workers 43 7 16.3 11.0 wives 64 5 7.8 5.4 children 91 3 3.3 2.2 total 258 24 9.3 - calculated from the formula I = 1 -X12/y, where x is the proportion remaining negative for y months (y = 18). For this calculation it was assumed that there was no seasonal variation in transmission. The results given include at least 4 subjects who had very light infections on reexamination and could have been missed in the 1969 survey. DISCUSSION For 3 years the population of B. pfeifferi was con- trolled by the careful use of molluscicides, with the result that snails were detected in the irrigation systems on only one occasion and were detected in the drains only during the week before treatment. The absolute level to which the population was reduced as a result of the periodic treatments is unknown, as is the absolute level the population had reached before molluscicide was applied. However, it is known that for these 3 years the snail population was held continually at a level lower than 0.1 % of the level that would have been expected in the absence of snail control. This is deduced from comparison of results prior to control, when over 2 500 Biomphalaria snails per week were detected in routine sampling, with results of sampling studies continued during the campaign. It is interesting to note that the incidence of new infections in newly employed workers was higher in the first 6 months than in the second 6 months. It is possible that not all of the 16 found after 6 months were new infections, but that some were light infections missed in the initial examination by the Foster method. Soon after the 6-month follow-up, 7 of the 16 were called to the hospital and it was found that when examined by the Bell (1963) quantitative method of examination their egg output was low; for 3 of the 7 it was only 5 eggs/g. This level can be missed in 25 % of cases examined by the Foster method and it is therefore suggested that the true incidence may well have been lower than the 16 found, by a factor of at least 3/7. This would reduce the incidence of infection in field workers for the first 6 months from 10/119 to 6/119 (5%) and that in other workers to 4/46 (9 Y.). Another possible explanation is that one would expect any new infections to be light as there had been a dramatic reduction in the snail population. However, one would still expect the same incidence in the second 6 months as in the first 6 months and that was not the case. The evidence suggests that the true incidence of infection in field workers was 5% in the first 6 months after arrival and 3.9% in the second 6 months-a great reduction from the 56.9% prevailing in 1962/63. A second point of interest is that apparently there was a higher incidence of S. mansoni infection in other workers than in field workers. However, with the number of observations included in the com- parisons, it is not possible to establish that the dif- ference is statistically significant. If transmission was occurring on the estate one would expect field workers, who come into daily contact with water, to contract the infection more readilythan otherworkers. A possible explanation for the unexpected level of infection in other workers is that they may have had a greater tendency to leave the estate occasionally and that they contracted the infection outside the estate boundaries. If one assumes that some of the new infections seen were acquired outside the estate, the transmission rate on the estate may have been very low during the period of the study. The results of the two surveys carried out in young children in 1969 and 1970 suggest a complete break in transmission during that time. The second curve is shifted significantly to the right and no children born after 1966 were found to be infected. From a comparison of the annual incidence of infection shown in Table 5 with the annual incidence of 81.3% in the group of new workers followed earlier by Foster (1967), before control was at- tempted, it seems clear that the control measures taken resulted in a drastic reduction in the transmis- sion of S. mansoni infection on the estate. Throughout this study chemotherapy was used to reduce the intensity of infection and by late 1970 the total egg output was estimated at only 15% of that recorded in 1969 (Fenwick & Jorgensen, 1972). 329 330 A. FENWICK The relative contributions of chemotherapy and snail control to the reduction in transmission rate are unknown, but it is felt that snail control may have been the larger factor because in 1962-63, when transmission of infection was at a high level, the extent of chemotherapy was as great as during the present control programme. It must be admitted, however, that the drugs in use at that time were probably less efficient than the niridazole and hycanthone that were used later. ACKNOWLEDGEMENTS The author is grateful to the management of the Tanganyika Planting Co. for permission to carry out this work and for their cooperation. Particular thanks are due to Dr T. A. Jorgensen for his encouragement, advice, and cooperation at the Company hospital. For help in preparing the final manuscript thanks are due to Dr G. Webbe and to Dr D. R. Bell, Dr W. W. MacDonald, and Dr M. E. A. Materu. Mr W. M. Foster gave invaluable help in the day to day running of the work. The Director, East African Tropical Pesticides Research Institute, Arusha, Tanzania, gave permission for this study to be carried out and the Secretary General of the East African Community gave permission for publication. RI SUMmt INFLUENCE D'UN PROGRAMME DE LUTrE SUR LA TRANSMISSION DE SCHISTOSOMA MANSONI DANS UNE PLANTATION DE TANZANIE POURVUE D'UN SYST-ME D'IRRIGATION Dans une plantation de cannes it sucre du nord de la Tanzanie, le reseau de canaux d'irrigation constitue un habitat extremement favorable a la proliferation de Biomphalaria pfeifferi, hote intermediaire de Schisto- soma mansoni; aussi la pr6valence et l'incidence de l'infection humaine par ce parasite etaient-elles parti- culierement elevees. De 1968 a 1970, on a mis en ceuvre un programme de lutte contre la schistosomiase en associant a l'emploi judicieux de molluscicides des operations de depistage et de traitement de masse. Le present article decrit les m6thodes utilisees pour mesurer l'influence de ces mesures sur l'incidence et la transmission de l'infection. En premier lieu, sur 220 nouveaux travailleurs reconnus indemnes de schistosomiase a leur arrivee 'a la plan- tation, 165 ont ete r6examin6s apres 6 mois et 105 apres 12 mois. La fr6quence globale des nouvelles infections a Wt6 de 9,7% durant le ler semestre et de 4,8% pendant le second, alors qu'anterieurement (en 1962/63), le taux d'infection nouvelle atteignait 56,9 %. On a aussi 6tabli des courbes de prevalence de la schisto- somiase selon l'age chez les jeunes enfants lors de deux enquetes effectuees en 1969 et en 1970. Dans l'intervalle de 18 mois separant les deux examens, la courbe de frequence s'est deplacee vers les groupes d'age superieur. A la fin de 1970, on n'a trouve aucun sujet infecte parmi les enfants ages de moins de 5 ans, alors qu'en 1962 la prevalence atteignait 25% dans le groupe d'age de 34 ans. Enfin, 258 personnes reconnues indemnes d'infec- tion 'a S. mansoni au debut de 1969 ont ete reexaminees a la fin de 1970. Les resultats ont permis de fixer les taux annuels d'incidence de la schistosomiase a 10% environ chez les hommes adultes, 'a 5,4% chez les femmes et 'a 2,2% chez les enfants, taux de loin inferieurs aux taux constat6s avant I'application des mesures de lutte. Pour l'auteur, ces resultats montrent 1'efficacite de la lutte entreprise, la r6duction brutale de la transmission etant due principalement i la destruction systematique des populations de mollusques. REFERENCES Bell, D. R. (1963) Bull. Wld Hlth Org., 29, 525 Bell, D. R. et al. (1967) Ann trop. Med. Parasit., 61, 422 Crossland, N. 0. (1962) Bull. Wld Hith Org., 27, 125 Crossland, N. 0. (1963) Bull. Wid Hlth Org., 29, 515 Crossland, N. 0. (1967) Bull. Wld Hlth Org., 37, 23 Farooq, M. et al. (1966) Bull. Wld Hlth Org., 35, 369 Fenwick, A. (1970) Bull. Wid Hith Org., 42, 589 Fenwick, A. & Jorgensen, T. A. (1972) Bull. Wld Hlth Org. (in press) Foster, R. (1964) J. trop. Med. Hyg., 67, 271 Foster, R. (1967) J. trop. Med. Hyg., 70, 159