Metrifonate in the control of urinary schistosomiasis in Zanzibar.

ResearchlRecherche Metrifonate in the control of urinary schistosomiasis in Zanzibar A.F. Mgeni,' U.M. Kisumku,2 F.S. McCullough,3 H. Dixon,4 S.S. Yoon,5 & K.E. Mott6 Selective population chemotherapy using three doses of metrifonate (7.5mg/kg body weight each time) at two-week intervals was assessed in an entire community in Kinyasini district in Zanzibar, United Republic of Tanzania. The objectives of the study were to (1) reduce the prevalence of heavy infections (defined as .50 S. haematobium eggs per 10ml of urine) by 75% in two years, and (2) reduce the overall prevalence of infection by 50% in two years. A total of 4113 people were examined at least once during the two-year period. In the initial survey the highest proportion of infected individuals was in the 10-14-year age group, and in all subsequent surveys in the 5-9-year age group. The age group with the highest proportion of heavily infected individuals was 5-9 years for all surveys. The overall reduction of prevalence of infection from survey 1 to survey 4 was 52.9% and the prevalence of heavy infection was reduced by 62.2%. The conversion rates (negative to positive in two consecutive surveys) were highest in the longest interval of 12 months and the rates of reversion (positive to negative in two consecutive surveys, without a history of treatment) were highest in the shortest interval of 4 months. Some statistically significant relationships were observed between the number of doses and the egg reduction rates. However, for the egg-negative rates, no statistically significant relationship was observed. In the 4-month interval a 67.6% egg-negative rate was observed among those who took at least one dose; with the 12-month interval a 48.3% egg-negative rate was observed. Thus, selective population chemotherapy with metrifonate was shown to reduce the prevalence and intensity of infection due to S. haematobium over a 24-month period. Introduction Advances in diagnostic techniques and chemotherapy now ensure that reduction in disease due to schis- Health Officer, Schistosomiasis Unit, Ministry of Health, Zanzibar, United Republic of Tanzania. 2 Director of Preventive Services, Ministry of Health, Zanzibar, United Republic of Tanzania. 3 Formerly Scientist, Division of Vector Biology and Control, World Health Organization, Geneva, Switzerland. 4 Chief, Epidemiological and Statistical Methodology, Epi- demiological Surveillance and Health Situation and Trend Assess- ment, World Health Organization, Geneva, Switzerland. 5 Scientist, Schistosomiasis and other Trematode Infections, Divi- sion of Control of Tropical Diseases, World Health Organization, Geneva, Switzerland. ' Chief, Schistosomiasis and other Trematode Infections, Division of Control of Tropical Diseases, World Health Organization, Geneva, Switzerland. Requests for reprints should be sent to Dr K.E. Mott at this address. Reprint No. 5127 tosomiasis is within the capabilities of the current health infrastructures of many endemic countries (1). The effectiveness of metrifonate, an organo-phos- phorus compound with specific anticholinesterase activity against Schistosoma haematobium, has recently been reviewed (2). Its relatively low cost, lack of toxicity, and high efficacy make it an attractive option, for large-scale treatment. However, most of the pub- lished studies on the efficacy of metrifonate for S. haematobium control have been limited to small groups of school-age children; this age group usually has the highest prevalence and the heaviest infections (3,4). Some studies have examined the efficacy of varying doses of metrifonate; others have used a single dose of 10mg/kg body weight and achieved a reduc- tion in egg output of at least 90% (4-6). A few studies have utilized the WHO-recommended schedule of three doses of 7.5 mg/kg body weight at two-week intervals and achieved a similar reduction in egg count (7,8). Zanzibar, an island off of the eastern coast of Bulletin of the World Health Organization, 66 (6): 721-730 (1990) (i) World Health Organization 1990 721 A.F. MgenI et al. Tanzania, was recognized as a highly endemic area of urinary schistosomiasis as early as 1885 (9). It was the site of the first clinical trials of metrifonate against S. haematobium infection 15 years before this present study (10, 11). This paper describes the results of selective popu- lation chemotherapy in an entire community in north- ern Zanzibar island. Metrifonate was administered in three doses (7.5 mg/kg body weight each time) at intervals of two weeks. The operational aim of the study was to (1) reduce the prevalence of heavy infections (defined as > 50 eggs per 10 ml of urine) by 75% in two years, and (2) reduce the overall prevalence of infection by 50% in two years. The effects of treatment, according to age group, sex, and initial intensity of infection were observed, and the efficacies of complete (three doses) treatment and incomplete treatment were compared. Methods Kinyasini district in Zanzibtar, United Republic of Tanzania, was selected as the study area. A detailed map (Fig. 1) showing the location of all houses in the district, as well as other relevant geographic features, was prepared and a complete census was undertaken in May 1981. All the residents in the district, who produced urine, were included in the study, each one being assigned a serial number. The four surveys in the study were carried out during a two-year period (July 1981 to July 1983), the census being updated at each survey. The examination and treatment team was composed of a health officer, two assistant health officers, three microscopists, two laboratory assistants and a driver, all from the Ministry of Health. The presence of S, haematobium infection was detected in a specimen of urine collected from each person. When the urine specimens were brought to the examination team, a few drops of 10% formal- dehyde were added. Using the Nytrel syringe filtration technique, a 10 ml random aliquot of each well- mixed specimen was examined by a trained micro- scopist either in a field laboratory or in the central laboratory. Either new or adequately washed Nytrel filters were used at each survey. No quality control was performed during the examinations. If a person was found to be positive, i.e., anyone with at least a single S. haematobium egg in the urine, treatment was given based on the WHO-recommended regimen (3 doses of 7.5 mg of metrifonate per kg body weight at 2-week intervals) (1). At each survey, the data on each individual included the district, school, house number, family order number, age, sex, urinary egg count, date of examination, weight, number of tablets of metrifonate per dose, and the number of doses received. Prelimi- nary data check and analysis using hand calculators were completed by the project staff. The data record forms were then forwarded to Geneva for final data processing and analysis by the WHO Epidemiology and Statistical Methodology unit. There are two distinct seasons for the trans- mission of schistosomiasis in Zanzibar. The high season, accompanying the rains, lasts from July to August, while the rest of the year can be considered as a low transmission season. The first survey was completed in August 1981, towards the end of the peak transmission season. The second survey was conducted 12 months after the first examination, in August 1982. The third survey was conducted 8 months after the second examination in April 1983, and the last survey was conducted 4 months later, in July 1983, at the beginning of the peak transmission season. At each survey, subsequent to the first, the total population examined included residents of the area who had participated in previous surveys, and those who had migrated into the study area since the last census. The results of the surveys, therefore, included data from individuals with previous treatments and those who were being treated for the first time. Although this obviates precise comparison of the results of successive surveys, nevertheless reasonable conclusions can still be reached when the analysis is performed on cohorts present during consecutive surveys. Other interventions were carried out concurrent- ly with chemotherapy including health education and information sessions in Kinyasini school, im- provements in the provision of safe water supplies, and mollusciciding. Information meetings with the village leaders and the community were held 2-3 times per year during the course of the project. Local religious leaders gave support to the activities of the project by encouraging participation of the com- munity in the surveys. Water improvement projects were carried out by the villagers and several wells were dug during the study period. Two molluscicide appli- cations, using niclosamide, at the main transmission sites (see Fig. 1) were undertaken in January 1983 before the third survey and in June 1983 before the last survey. Results A total of 4113 people were examined at least once during the two-year period. The age and sex distribu- tion of the total study population is shown in Table 1. It was assumed that the characteristics of people moving into the area were similar to those leaving the WHO Bulletin OMS Vol 68. 1990722 Metrifonate In the control of urinary schistosomiasis In Zanzibar Fig. 1. Map of Zanzibar and the Kinyasini project area. N BANDA MAJI 0 *a - _ MGAMBO o CCM Branch e House = Market A SchooI I Mosque a. Rice field /Swamp x Well _._ ___ - = * 0 Mainroad Borderline Footpath River Bridge Main transmission sites 1 mile 'sini INDIA N OCEAN ;.>.A...t.Dry or coral area ... Streams ..,..,.,.> > St% V% 0 5 mili TMtende WHO Bulletin OMS Vol. 68. 1990. SCALE... . ... .. . . | 723 A.F. Mgenl et al. Table 1: Age and sex distribution of the study population: cumulative summary of four surveys, Zanzibar, 1981-83 Age group (years) Male Female Total 0-4 334 325 659 (16.0)' 5-9 548 563 1111 (27.0) 10-14 404 287 691 (16.8) 15-24 289 319 608 (14.8) 25-44 297 303 600 (14.6) ,45 238 206 444 (10.8) Total 2110 2003 4113 a Figures in parentheses are percentages. Table 2: Prevalence of S. heematoblum, Zanzibar, 1981-83 Egg- Heavy No. examined positives infection' Survey 1 2685 1324 (49.3)b 518 (19.3) Survey 2 1887 751 (39.8) 254 (13.5) Survey 3 2458 848 (34.5) 269 (10.9) Survey 4 1719 398 (23.2) 126 (7.3) ' Heavy infection: >50 eggs/10 ml of urine. bFigures in parentheses are percentages. area, i.e., the population structure stayed essentially the same. However, due to migration, the total number of people examined changed from survey to survey. Table 2 shows the number of people examined during each of the surveys, the overall prevalence rates and the prevalence rates for heavy infection (i.e., individuals with >,50 eggs/lOml of urine); the number of people includes those who were new to the surveys. Surveys Rrst survey (August 1981). During the first survey 2685 people were examined, of whom 1324 (49.3%) were positive and 518 (19.3%) were heavily infected. The highest prevalence of heavy infections was in the 10-14-year-old age group (203/672: 30.2%). There was no difference in infection between males and females (P >0.05). Second survey (August 1982). Twelve months after the first survey, 1887 persons were examined; 1420(75.3%) had been present during the previous survey (survey 1) while 467 (24.7%) were new to the survey. Of the 1887 examined, 751 (39.8%) were infected, and 254 (13.5%) were found to be heavily infected. In this survey the group with the highest percentage of heavy infection was the 5-9-year-old group (104/383: 27.2%). Third survey(April 1983). Eight months after the second survey 2458 individuals were examined, of whom 1342 (54.6%) had been present during the previous survey (survey 2) and 1116 (45.4%) were either new to the survey, or not present during the previous survey. Of the 2458 examined, 848 (34.5%) were found to be positive, and 269 (10.9%) were found to be heavily infected. The 5-9-year-old age group had the highest percentage of heavily infected individuals (147/687: 21.4%). Fourt survey (August 1983). A total of 1719 individuals were examined during the fourth survey, held 4 months after the third survey. A total of 1416 (82.4%) had been present during the previous survey (survey 3) and 303 (17.6%) were either new to the survey or not present during the previous survey. Of the 1719 examined, 398 (23.2%) were found to be positive, and 126 (7.3%) were found to be heavily infected. Again, the 5-9-year-old group had the highest proportion of heavily infected individuals (78/541: 14.4%). With assumptions about the study population stated previously, the overall reduction from survey 1 to survey 4 was 52.9% for the prevalence of positive (from 49.3% at survey 1 to 23.2% at survey 4, Table 2), and 62.2% for the prevalence of heavy infection (from 19.3% at survey 1 to 7.3% at survey 4, Table 2). Similar reductions in prevalence among men and women were observed (P>0.05). Changes between the surveys People who were present at one survey, but absent at the previous survey were not included in the further analysis. Thus the data from only those individuals who were present at least on two consecutive surveys were analysed (Table 3) in terms of the overall prevalence rates, the prevalence rates for heavy infec- tion, and the percentage reduction for the two preva- lence rates. Table 3: Prevalence of S. haematoblum In cohort present at consecutive surveys: reduction in overall prevalence and prevalence of heavy infection, Zanzibar, 1981-83 Egg positive Heavy infection No. examined' No. Reduction No. Reduction Survey 2 1420 570 (40.1)b 17.8% 191 (13.5) 28.9% Survey 3 1342 391 (29.1) 26.0% 117 (8.7) 31.0% Survey 4 1416 303 (21.4) 40.9% 90 (6.4) 45.3% ' Number examined includes only those who were present at two consecutive surveys, i.e., the number for survey 2, 1420, includes those who were present at both survey 1 and survey 2. b Figures in parentheses are percentages. WHO Bulletin OMS. Vol. 68. 1990.724 Metrifonate In the control of urinary schistosomiasis in Zanzibar Survey 1 to survey 2. A total of 1420 persons were examined during both the first survey and the second survey 12 months later. Of these 693 (48.8%) were positive and 270 (19.0%) were infected heavily at the time of the first survey. Of the 693 positives from the first survey, 480 (69.3%) had a reduction in intensity of infection, of whom 330 (47.6%) had negative egg output at the second survey. These numbers include the population that was treated as well as those who were not treated after the first survey. Survey 2 to survey 3. The number of people examined during both the second survey and the third survey 8 months later totalled 1342. Of these, 528 (39.3%) were positive and 169 (12.6%) were heavily infected at the time of the second survey. Of the 528, the intensity of infection was reduced in 395 (74.8%) and 326 (61.7%) had negative egg output by the third survey. Survey3to survey 4. In all, 1416 persons were examined during both the third survey and the final survey 4 months later, of whom 513 (36.2%) were positive and 165 (11.7%) were heavily infected at the third survey. Ofthose infected, the intensity of infection was reduced in 411 (80.1 %), of whrom 345 (67.3%) were egg negative at the last survey. Changes over 24 months (from survey 1 to survey 4). A total of 673 persons were examined in all four surveys; their age distribution is given in Table 4. The cohort was made up of 465 persons who were treated at least once during this period, most of them with at least 1 dose of metrifonate out of the recommended regimen of 3 doses. A total of 208 persons did not receive any treatment; these persons were negative or, if positive, refused treatment. Out of the 673 persons, 32 (4.8%) were positive at each examination. All 32 individuals received at least one dose of metrifonate during the course of the study. The 5-9-year-old age group had the highest Table 4: Age distribution of cohort examined in all four surveys Age group (years) 0-4 5-9 10-14 15-24 25-44 >45 No. positive in Initially positive, No. all 4 surveys then always negative 112 152 118 70 124 97 6 (5.4) 15 (9.9) 9 (7.6) 0 (0.0) 1 (0.8) 1 (1.0) Total 673 32 (4.8) 6 (5.4) 8 (5.3) 13 (11.0) 11 (15.7) 23 (18.5) 24 (24.7) 85 (12.6) ' Figures in parentheses are percentages. proportion of individuals who remained positive (15/152: 9.9%); only two persons, out of the 32, were 15 years or older. It was noted that 85 individuals (12.6%) were positive in the first survey, became egg-negative after treatment, and then remained egg-negative through- out the rest of the study. The older age groups () 15 years of age) had a higher proportion of individuals who were egg-negative and not tested positive com- pared to the younger age groups. In addition, the intensity of infection among these 85 individuals showed that while 19 (22.4%) of those initially positive were heavily infected, 66 (77.6%) were mild to mod- erately infected. Compliance and conversion/reversion rates Between 55% and 65% of those eligible for treatment took three doses of metrifonate at each survey (Table 5). The compliance rates at each survey were similar among persons with different levels of infection (P>0.05). Also, there was no statistical difference in the compliance rate between the sexes or among different age groups (P>0.05). The rates of "conversion" (negative to positive in two consecutive surveys) and "reversion" (positive to negative, without a history of treatment) are sum- marized in Table 6. The conversion rates were highest (207/727: 28.5%) between survey 1 and survey 2. On the other hand, the reversion rates were highest (13/22: 59.1 %) between survey 3 and survey 4. Effect of treatment and age Egg reduction rates (positive egg output that was changed to a reduced or negative egg output due to metrifonate chemotherapy), by dose, are summarized in Table 7. Those who took at least one dose of metrifonate during the first survey had a reduction rate of 78.4% (482 of 615), which includes those that became egg-negative. For the second and the third surveys, the reduction rates were 70% (343 of 491) and 80.4% (395 of 491), respectively. The highest frequency of egg-negative counts after treatment were observed in the older age groups (over 15 years of age) and were found to be statistically significant (Table 8). Interestingly, similar trends in egg-negative rates were observed for the untreated group as well between surveys 2 and 3. Other interventions The data collected during the study did not include information on interventions other than chemother- apy which may have affected the level of morbidity in the study population. The impact of the health education activities was not assessed during this study. WHO Bulletin OMS Vol 68. 19907 725 A.F. Mgeni et al. Table 5: Compliance rates for three doses of metrifonats In the study population No. of doses received No. positive No. who refused treatment 1 2 3 Survey 1 1324 195 (14.7)' 97 (7.3) 167 (12.6) 865 (65.3) Survey 2 751 73 (9.7) 99 (13.2) 165 (22.0) 414 (55.1) Survey 3 848 69 (8.1) 83 (9.8) 209 (24.6) 487 (57.4) Survey 4 No treatment records are available a Figures in parentheses are percentages. Table 6: Conversion/reversion of Individuals who did not take any treatment Survey 2 Survey 3 Survey 4 Conversion 207/727 (28.5%) 189/814 (23.2%) 135/903 (15.0%) Proportion with heavy infection" 101/207 (48.8%) 42/189 (22.2%) 26/135 (19.3%) Reversion 33/78 (42.3%) 16/37 (43.2%) 13/22 (59.1%) Proportion with mild/moderate infectionb 24/33 (72.7%) 14/16 (87.5%) 12/13 (92.3%) ' Heavy infection: > 50 eggs/10 ml of urine. b Mild/moderate infection: less than 50 eggs/10 ml of urine. Table 7: Egg-reduction rates according to dosage of Table 8: Egg-negative rates for those treated with at least metrifonate: all Infections and heavy Infectlons (,>50 one dose of metrifonats and those untreated "getslu mi urine) Survey 1-2 Survey 2-3 Survey 3-4 (12 months) (8 months) (4 months) Doses No. % No.a % No.' % A/l infections: 0 33 42.3 22 59.5 16 72.7 1 30 57.7 45 70.3 30 78.9 2 80 79.2 93 78.2 93 75.6 3 372 80.5 205 66.6 272 82.4 P< 0.05b N. S.b N. S.b Heavy infections: 0 19 24.4 4 10.8 2 9.1 1 11 21.2 11 17.2 7 18.4 2 30 29.7 29 24.4 33 26.8 3 152 32.9 89 28.9 85 25.8 p< 0.05b P< 0.05b N. S.b 'Numbers include those that are egg-negative. bChi-square test for binomial trends. N. S. = not significant. No data were collected on the use of the new wells or the change in behaviour attributable to the availability of wells in the study area. However, at the outset of the project there were only three wells in the study area and by August 1983 there were eight. All the new wells were constructed by their owners, who recog- nized the benefits of washing clothes and drawing Age Survey 1-2 Survey 2-3 Survey 3-4 group (12 months) (8 months) (4 months) (years) No. % No. % No. % Treated: 0-4 24 55.8 30 58.8 25 65.8 5-9 63 34.4 66 43.7 142 58.9 10-14 62 43.4 46 53.5 89 69.0 15-24 47 53.4 61 78.2 35 85.4 25-44 51 63.8 60 82.2 20 100.0 > 45 50 64.1 47 90.4 21 95.5 Total 297 48.3 310 63.1 332 67.6 P< 0.001 P< 0.001' P< 0.001' Untreated: 0-4 0 0.0 2 33.3 2 '66.7 5-9 7 43.8 1 14.3 1 25.0 10-14 11 36.7 3 33.3 1 25.0 15-24 3 30.0 3 50.0 6 100.0 25-44 9 69.2 3 75.0 2 66.7 > 45 3 50.0 4 80.0 1 50.0 Total 33 42.3 16 43.2 13 59.1 N. S.' P<0.025' N. S.' ' Chi-square test for binomial trends. N. S. = not significant. water at a distance from the streams. In addition, although mollusciciding was carried out twice in 1983, observations on snail host population density, prior to and after application, were not specifically recorded. WHO Bulletin OMS Vol, 68. 1990726 Metrifonate in the control of urinary schistosomiasis In Zanzibar Discussion In this study, selective population chemotherapy with metrifonate was shown to reduce the prevalence and intensity of infection due to S. haematobium over a 24-month period in a community on the island of Zanzibar. The operational target of at least 50% reduction in prevalence was achieved. However, heavy infections were reduced by 62% rather than the projected 75%. In all previous reports, metrifonate has been shown to be less efficacious against heavy infections than light infections. The explan- ations for these observations have ranged from a partial effect of the drugs on the adult worm to the variation in transmission intensity. As noted from the initial trials (12) of metrifonate some persons con- tinued to be heavily infected after treatment. Using selective population chemotherapy at 10mg/kg body weight, selective mollusciciding, and modification of the snail habitats by a self-help programme, Rugemalila & Eyakuze observed even higher egg- negative rates (69% at 6 months and 67% at 12 months) in mainland Tanzania (13). The data collected during the study do not lend themselves to rigorous analysis. The study was not designed to obtain information on a cohort in dif- ferent surveys. The detailed information needed on individuals is difficult to collect and is often beyond the objectives or capabilities of a control programme. Certain comparisons, however, can be made, bearing in mind the limitations of the study design. The assumption of a high level of reinfection in the study population is supported using the conver- sion/reversion ratio suggested by Jordan (14). We observed an increasing proportion of negative-to- positive conversion as the interval of evaluation increased. This is further substantiated by the parallel increase in proportion of those acquiring heavy infec- tions. On the other hand, the recognized fluctuation in low egg counts over short periods of time is confirmed by the highest rate of reversion at the 4-month interval. As the period of observation got longer, the rates of conversion and reversion became closer. However, this observation must be qualified by stating that the observation periods were conduc- ted at different times of the year, i.e., there may have been an effect on the rates of conversion and reversion due to seasonality of transmission. The 40% rate of spontaneous occurrence of negative egg counts in untreated persons was similar to that observed by Rugemalila & Eyakuze (13). As suggested by them, we observed that these "sponta- neous cures" were associated with initially low egg counts and hence were probably due to the short- term variability of urinary egg excretion. In our study area the high rate of reversion to negative egg counts after the 4-month interval may also have been due to low transmission during that period (April to July) and the higher proportion of persons with initially low counts. The timing of chemotherapy to corres- pond with the projected transmission cycle has been advocated (i.e., treatment should be given ideally immediately after peak transmission) to ensure the longest possible period without reinfection before retreatment. This hypothesis has never been fully tested. We also observed that the age group with the highest prevalence changed during the course of the study. At the first survey the expected peak in the 10-14-year-old age group was observed. In subsequent surveys, the lowest egg-negative rates and the highest proportion of the heavy infections were observed in the 5-9-year-old age group. This change may be due, in part, to the age-dependent immunity proposed by Wilkins et al., who observed that the prevalence in the 5-9-year-olds in their study had a higher preva- lence at 3 and 15 months after treatment with praziquantel and that the geometric mean egg counts (log1o (n + 1)) were proportionally higher after treat- ment at the same intervals (15). This reinforces the necessity to plan the most appropriate interval to treat children in the early phases of large-scale treatment programmes. The egg-negative rates, resulting after treatment with metrifonate, increased as the intensity of infection decreased in the course of the study and also as the interval of evaluation decreased. These observations are consistent with those made in endemic areas of S. haematobium in Tanzania and elsewhere. Among those that were treated at least once, the egg-negative rates were found to be related to age (P<0.001 for surveys 2 to 4, using the test for binomial trend). Among those that were not treated at all, only survey 3 had a statistically significant relationship between the egg- negative rate and age (P<0.05). Furthermore, even with the full regimen of three doses of metrifonate, some people became more intensely infected in the subsequent survey. The increase in intensity of infec- tion was mostly in the group that was mildly infected (1-9 eggs/lOml urine) and in the younger age group (between 5 and 14 years of age). In the cohort which participated throughout all surveys, it was observed that those whose infections were of low intensity at the first survey, tended to remain egg-negative throughout the period of obser- vation. Since most of these persons were in the older age groups, this supports the operational approach of treating adults at intervals of two or more years and targeting school-aged children for more frequent screening and treatment. However, individuals with infections, initially of high intensity, had a less pre- WHO Bulletin OMS Vol 68. 1990 727 A.F. Mgeni et al. dictable outcome. This observation has public health relevance in that when operationally and logistically feasible, heavily infected individuals or communities with a high proportion of heavily infected persons (probably greater than 10%) should be reexamined at relatively short intervals and retreated if necessary in the early stages of the control activities. A number of studies have utilized single-dose metrifonate chemotherapy with success (6,16), while others have reported a relationship between the num- ber ofdoses given and the rate of cure; higher cure rates usually correlate with an increased number of doses. In our study, only a weak correlation between efficacy (egg-negative rate and egg-reduction rate) and dosage was found. Davis & Bailey (12) stated that there was a non-linear relationship between the number of doses and cure rates. Sato et al. (7) have shown that the cure rate and percentage reduction of mean egg output were significantly lower in groups with fewer doses. In this study, however, a statistically significant rela- tionship between the number of doses and efficacy existed only during some of the surveys (Table 7). The number of doses was significantly related to the egg-reduction rates between survey 1 and survey 2 (P <0.05), for both the "all infected" cohort and the "heavily infected" cohort. The "heavily infected" co- hort also had a statistically significant relation be- tween doses and egg-reduction rates between survey 2 and survey 3 (P <0.05). There was no statistically significant relationship between the number of doses and the egg-negative rates. Metrifonate has been widely used because of its low cost and relative lack of adverse side-effects. However, because the recommended treatment with metrifonate requires three doses for the complete regimen, the cost of drug delivery, compared to a single-dose drug (e.g., praziquantel), has been shown to be rather high (17). As a method of minimizing the cost of delivery, many researchers have been looking at the efficacy of a reduced number of metrifonate doses given at more frequent intervals. In Burkina Faso, two doses of 7.5 mg/kg body weight were given 2-weeks apart three times in one year to school- children with a resulting 85% cure rate (18). This approach has not yet been tested on a large scale and the possible changes in the cost of drug delivery have not been estimated, but they would probably be considerable. Although the improvements in the water supply were welcomed by the residents, the incorporation of safe water projects within schistosomiasis control programmes need to be analysed in greater detail. Economic as well as epidemiological benefits and costs of water projects must be examined carefully to maximize the benefits of schistosomiasis control pro- grammes. The effectiveness of mollusciciding could not be assessed during this study. As emphasized by pre- vious consultants to the Ministry of Health, mollus- ciciding was not feasible until the geographical limits of transmission sites could be determined.8 It was assumed that this might be possible after chemo- therapy had reduced both prevalence and intensity to low levels and the water contact patterns of those remaining infected or those acquiring new infections could be elucidated. This approach is now being tested on the island of Pemba, north of Zanzibar. The results of this study have shown that selective mass chemotherapy, using metrifonate, is an effective method of reducing morbidity due to S. haematobium infection. Furthermore, the fact that there was a lack of correlation between the number of doses taken and the rate of reduction in intensity in some of the surveys suggests that other factors may be influencing the rate of reduction; if so, they remain to be determined with rigour. a McCullough, F.S. & Kraft, R.J.G. Site visit to Zanzibar. Unpub- lished report to the WHO Regional Office for Africa, AFR/SCHIST/34, 1976. Acknowledgements The authors are grateful for the support of the Ministry of Health of Zanzibar, especially to Mr Ali Mohamed Shoka, Deputy Minister of Health at the time the project was initiated. All the field work was undertaken by Mr Ameir Khamis, Health Officer, Mr Abdul-Rahaman, Assistant Health Officer, and Mr Issa Haji Ameir, Mr Hassan Ali and Mrs Maulid Soud, microscopists; Mr Hassan Ali and Mrs Dude Mwatabu, laboratory assistants; and Mr Abdullah Shamte and Mr Ame Mussa, drivers. Hamad A.J. Haji of the Ministry of Health and Miss Patricia A. Tschida, U.S. Peace Corps volunteer, assisted the team during some of the surveys. Financial support was provided by the World Bank/UNDP/WHO Special Programme for Research and Training in Tropical Diseases (TDR). Special thanks are due to Dr A.O. Lucas, Director of TDR at the time, who innovatively and actively supported the idea of this oper- ational project. Resume Le m6trifonate dans la lutte contre la bilhar- ziose schistosomiase urinaire a Zanzibar Cette etude d6crit l'efficacit6 d'une chimiothe- rapie selective appliquee a la totalite d'une communaute du district de Kinyasini, a Zanzibar, R6publique-Unie de Tanzanie, en utilisant trois doses de metrifonate (7,5 mg/kg de poids corporel chaque fois), a deux semaines d'intervalle. Le but WHO Bulletin OMS Vol 68. 1990.728 Metrifonate In the control of urinary schistosomiasis In Zanzibar de l'etude etait de reduire la pr6valence des fortes infestations (,> 50 ceufs de S. haematobium pour 10 ml d'urine) de 75% en deux ans, et de reduire la pr6valence globale de l'infestation de 50% en deux ans. Au total, 4113 personnes ont ete examinees au moins une fois pendant la periode de deux ans. Lors de la premiere enquete (aout 1981), 1324 personnes sur 2685 (49,3%) etaient infestees, et 518 (10,3%) etaient fortement infestees. La prevalence la plus elevee d'infestation a ete observee dans le groupe d'age 10-14 ans. Lors de la seconde enquete (aout 1982), 12 mois apres la premiere, 751 per- sonnes sur 1887 (39,8%) etaient infestees, et 254 (13,5%) etaient fortement infestees. Dans cette enquete, le groupe ayant le plus fort pourcentage d'infestation etait le groupe 5-9 ans. Lors de la troisieme enquete (avril 1983), huit mois apres la deuxieme, 848 individus sur 2458 (34,5%) etaient infestes et 269 (10,9%) etaient fortement infestes. Le groupe d'age 5-9 ans etait celui ayant le pour- centage le plus eleve d'individus infestes (21,4%). Lors de la quatrieme enquete (aout 1983), quatre mois apres la troisieme, 398 individus sur 1719 (23,2%) etaient positifs et 126 (7,3%) etaient forte- ment infestes. A nouveau, le groupe d'age 5-9 ans avait la proportion la plus elevee d'individus in- festes (14,4%). Dans toutes les enquetes, le groupe d'age ayant la proportion la plus elevee d'individus fortement infestes etait le groupe 5-9 ans. La reduction globale de la prevalence de l'infestation entre l'enquete 1 et l'enquete 4 etait de 52,9% et la prevalence des fortes infestations a ete reduite de 62,2%. Les taux de conversion (negatifs a positifs dans deux enquetes consecutives) etaient les plus eleves pendant la periode la plus longue, de 12 mois, et les taux d'inversion (positifs a negatifs dans deux enquetes consecutives, sans antecedent de traite- ment) etaient les plus eleves dans l'intervalle de 4 mois. Sur 673 personnes suivies pendant la duree de l'tude, 32 (4,8%) sont restees positives pendant l'tude entiere; 85 personnes (12,6%) qui etaient positives dans la premiere enquete sont devenues negatives dans la seconde enquete et le sont restees pendant tout le reste de l'etude. Certaines relations, statistiquement significa- tives, ont ete observees entre le nombre de doses et les taux de reduction des ceufs. Cependant, pour les taux negatifs, aucune relation statistiquement significative n'a ete observee. Entre 55% et 65% des sujets justiciables du traitement ont pris trois doses de metrifonate lors de chaque enquete. Chez les personnes qui avaient pris au moins une dose, et pour l'intervalle de 12 mois, un taux de negativite de 48,3% a ete observe. Les taux les plus eleves de numeration negative d'aeufs ont toujours ete ob- serves chez des personnes de plus de 15 ans, dans les deux cohortes de personnes traitees et non traitees. Dans cette etude, une chimiotherapie se- lective appliquee a 1'ensemble d'une population avec du metrifonate s'est montree efficace pour reduire la prevalence et l'intensite de l'infestation a S. haematobium pendant une periode de 24 mois. References 1. WHO Technical Report Series No. 728, 1985. (The control of schistosomiasis: report of a WHO Expert Committee). 2. Feldmeler, H. & Doehring, E. Clinical experience with metrifonate. Acta tropica, 44: 357-368 (1987). 3. Tswana, S.A. & Mason, P.R. Eighteen-month follow- up on the treatment of urinary schistosomiasis with a single dose of metrifonate. American journal of tropi- cal medicine and hygiene, 34: 746-749 (1985). 4. King, C.H. et al. Chemotherapy-based control of schis- tosomiasis haematobia. I. Metrifonate versus praziquantel in control of intensity and prevalence of infection. American journal of tropical medicine and hygiene, 39: 295-305 (1988). 5. Arap Siongok, T.K. et al. Quantification of infection with Schistosoma haematobium in relation to epi- demiology and selective population chemotherapy. II. Mass treatment with a single oral dose of metrifonate. Journal of infectious diseases, 133: 856-858 (1978). 6. Kholy, A.E. et al. The effect of a single dose of metrifonate on Schistosoma haematobium infection in Egyptian schoolchildren. American journal of tropical medicine and hygiene, 33: 1170-1172 (1984). 7. Sato, K. et al. Efficacy of metrifonate in a highly endemic area of urinary schistosomiasis in Kenya. American journal of tropical medicine and hygiene, 38: 81-85 (1988). 8. Diallo, S. et al. Efficacite de 3 cures de metrifonate dans le traitement de la bilharziose urinaire au Senegal. Dakar medical, 28: 67-76 (1983). 9. Petrie, J. Bilharzia haematobia. Lancet, 2: 133-134 (1903). 10. Forsyth, D.M. & Rashid, C. Treatment of urinary schistosomiasis: practice and theory. Lancet, 1: 130- 133 (1967). 11. Forsyth, D.M. & Rashid, C. Treatment of urinary schistosomiasis with trichlorophane. Lancet, 2: 909- 912 (1967). 12. Davis, A. & Bailey, D.R. Metrifonate in urinary schis- tosomiasis. Bulletin of the World Health Organization, 41: 209-224 (1969). 13. Rugemallla, J.B. & Eyakuze, V.M. Use ot metrifonate for selective population chemotherapy against urinary schistosomiasis in an endemic area of Mwanza, Tanzania. East African medical journal, 58: 37-43 (1981). 14. Jordan, P. Schistosomiasis-the St. Lucia project. Cambridge, Cambridge University Press, 1985. WHO Bulletin OMS Vol 68. 1990 729 A.F. Mgeni et al. 15. Wilkins, H. et al. Resistance to reinfection after treat- ment of urinary schistosomiasis. Transactions of the Royal Society of Tropical Medicine and Hygiene, 87: 29-35 (1987). 16. Pugh, R.N.N. & Teesdale, C.H. Long-term efficacy of single-dose oral treatment in schistosomiasis haema- tobium. Transactions of the Royal Society of Tropical Medicine and Hygiene, 78: 55-59 (1984). 17. Korte, R. et al. Cost and effectiveness of different approaches to schistosomiasis control in Africa. Tropical medicine and parasitology, 37: 149-152 (1986). 18. Drulhle, P. et al. Essai de controle de la bilharziose urinaire par 3 cures annuelles de metrifonate. Annales de la Societe be/ge de M6decine tropicale, 61: 99-109 (1981). 730 WHO Bulletin OMS. Vol. 68. 1990.