Evaluation of reagent strips in urine tests for detection of Schistosoma haematobium infection: a comparative study in Ghana and Zambia

Bulletin ofthe WorldHealth Organization, 63(1): 125-133(1985) C World HealthOrganization 1985 Evaluation of reagent strips in urine tests for detection of Schistosoma haematobium infection: a comparative study in Ghana and Zambia K. E. MoTT,1 H. DIXON,2 E. OSEI-TUTU,3 E. C. ENGLAND,3 K. EKUE,4 & A. TEKLE4 Thepresence ofhaematuria andproteinuria, detected by reagent strips, was compared with Schistosoma haematobium egg counts in the urines of human subjects from two epidemiologically distinct areas in Ghana and Zambia. In children and adults in both areas, the individual or combined semiquantitative levels of proteinuria and haematuria were related directly to increasing urinary egg counts. In both areas the presence ofblood in the urine was highly specific (greater than 85%o) and sensitive, being positive in 97%o of urine specimens with more than 64 eggs per 5-ml sample of urine. The sensitivity of the protein indicator was also high, but its specificity was less than the blood indicator. The specificity ofcombined proteinuria and haematuria was higher than either alone; on the other hand, the sensitivity was lower than either alone. At each level ofproteinuria and haematuria, the geometric mean urinary egg count was higher in Ghana than in Zambia. This study confirms the necessity to evaluate indirect diagnostic techniques in each endemic country, in order to establish criteria for their interpretation, before wide-scale use. Urinary schistosomiasis due to Schistosoma haematobium infection affects over 90 million persons, principally in Africa. Haematuria and proteinuria among infected children has long been recognized as an early sign of infection, and it has been shown that the degree of haematuria and pro- teinuria in children is related to the intensity of S. haematobium infection (1). New urine-filtration techniques for the detection of S. haematobium infection, although useful under field conditions, require microscopic examination by trained personnel. The lack of trained field personnel in endemic areas is a serious constraint on the devel- opment of schistosomiasis control programmes (2). Simple indirect diagnostic techniques that could be used efficiently by minimally trained health workers to identify heavily infected persons, particularly children, would therefore aid the implementation of 1 Chief, Schistosomiasis and other Snail-borne Trematode Infec- tions, Parasitic Diseases Programme, World Health Organization, 1211 Geneva 27, Switzerland. 2 Epidemiological and Statistical Methodology, World Health Organization, Geneva, Switzerland. 3 Schistosomiasis Unit, Division of Epidemiology, Ministry of Health, Accra, Ghana. 4 WHO Regional Office for Africa, Brazzaville, Congo. Formerly, Tropical Disease Research Centre, Ndola, Zambia. such control programmes. Rapid diagnosis of heavily infected persons followed by treatment with new, safe, and highly effective antischistosomal drugs (at a cost acceptable to endemic countries) could be expected to reduce the morbidity related to schistosomiasis. The prevalence, intensity, and related morbidity of urinary schistosomiasis in Africa vary according to the epidemiology, transmission patterns, and ecology of each endemic area. The present study describes an evaluation of reagent strips in the detection of urinary blood and protein in relation to S. haematobium infection in t%o geographically separated endemic areas that were epidemiologically distinct. MATERIALS AND METHODS Study areas Ghana. The Epidemiology Unit of the Ministry of Health is responsible for an applied field research project (3) on Lake Volta, 180 km from Accra. The present study was carried out in five settlements in Adawso, near this project area, where no previous diagnosis or treatment programmes had been under- 450 -125- 126 K. E. MOTT ET AL. Table 1. Prevalence and intensity of S. haematobium infection in the study population, according to age and sex, in Adawso, Lake Volta, Ghana Males Females Both Age No. Prevalence Egg No. Prevalence Egg No. Prevalence Egg (years) examined (%) count' examined (%) count' examined (%) count" 0-4 30 30.0 4.6 14 50.0 3.8 44 36.4 4.3 5-9 64 85.9 53.0 58 82.8 28.8 122 84.4 39.9 10-14 69 94.2 50.0 51 92.2 61.8 120 93.3 54.6 15-19 34 85.3 39.6 29 69.0 24.8 63 77.8 32.7 20-24 24 62.5 36.6 21 33.3 14.3 45 48.9 27.2 25-34 23 65.2 8.5 34 47.1 4.4 57 54.4 6.1 35-44 24 66.7 10.2 31 35.5 4.1 55 49.1 7.1 45-54 21 66.7 7.0 11 54.5 6.2 32 62.5 6.7 55-64 14 50.0 7.5 10 10.0 1.0 24 33.3 5.9 Total 303 74.3 29.0 259 62.9 21.2 562 69.0 25.4 a Geometric means of egg counts per random 5 ml urine sample from infected persons. taken. The prevalance and intensity of S. haemato- bium infection among residents of this area are among the highest reported in Africa. Zambia. The Tropical Disease Research Centre in Ndola, supported by the UNDP/World Bank/WHO Special Programme for Research and Training in Tropical Diseases, undertakes field testing of anti- schistosomal drugs in areas near Ndola. The present study in Mutenda was part of a screening programme to identify infected schoolchildren. None of the children examined had been treated previously. Methods In the study areas, the total bladder content of the participants was collected between 10 h 00 and 14 h 00 into polyethylene bags; at the same time, a simple questionnaire was completed (results will be reported elsewhere). a With the aid of reagent strips, b the protein content in the urine was recorded as negative (.< 10 mg/100 ml of urine), trace (10-30 mg), + (30-100 mg), + + (100-300 mg), + + + (300-1000 mg), or + + + + ( >1000 mg/ 100 ml); presence of blood in the urine was recorded as negative, +, + +, or + + +. After testing with the reagent strips, the urine specimen in the polyethylene bag was labelled and fixed with a few drops of 1O0o formaldehyde and transported to a designated central laboratory. In the laboratory two aliquots (5 ml each in the highly endemic area of Ghana and 10 ml each in Zambia) ' See article on pages 135-142 in this issue of the Bulletin. b Neostix-3 from Ames-Miles Laboratories, Elkhart, IN, USA. were examined microscopically using a Millipore filtration technique (3). The data from Ghana refer to 5 ml urine samples, i.e., half the volume of the urine samples examined in Zambia. Currently, 10 ml aliquots of urine are recommended for urine filtration techniques requiring the Nuclepore,C Nytrel,d or paper filters. However, because of the blockage of the paper filters with sediment and the high intensity of S. haemato- bium infections in the Lake Volta region, small volumes (5 ml) of urine were examined. Thus, in order to compare the egg counts between Zambia and Ghana, the data from Ghana should be multiplied by 1.5-2 times (3). The data from the questionnaires and laboratory records were processed independently and analysed in WHO headquarters, Geneva. The results from the two study areas are described below separately and are compared in the discussion. RESULTS Adawso, Ghana Study population. A total of 562 persons were interviewed and their urine specimens tested (Table 1). The age and sex distribution of the study population was similar to that reported before for the c Nuclepore polycarbonate filters, 12 or 14 jm pore size, from Nuclepore Corporation, Pleasanton, CA, 94566 USA. d Nytrel polyamide filters, 20 Am mesh, from Union des Gazes a Bluter, 42360 Pannisieres, France. DETECTION OF SCHISTOSOMIASIS BY REAGENT STRIPS entire project area (3). The peak prevalence and intensity of S. haematobium infection was observed in the 10-14-year age group. These results have been reported previously (1). Presence ofurinary blood andprotein in relation to egg count levels. Blood was detected in nearly all (97%) urine specimens that had more than 64 eggs per 5 ml of urine and in 86% of urine specimens with more than 16 eggs per 5 ml of urine. These specimens were from persons in both the 5-14-years (children) and > 15-years (adult) age groups. The specificity of the urinary blood determination was high; at least + blood was observed in 15% of urine specimens from children and 11 % of urine specimens from adults in which S. haematobium eggs were not found. The sensitivity of urinary protein determinations was high. Protein was detected in urine specimens from 94% of infected children and 84% of infected adults. Protein levels of at least 100 mg/100 ml of urine were found in 62%o of infected children and 39%o of infected adults. On the other hand, the specificity of urinary protein determinations was low. Proteinuria was detected in the majority of urine specimens that had no S. haematobium eggs- from children (81 %o) and from adults (65%). Protein concentrations of at least 100 mg/100 ml of urine were observed in 26% of egg-negative specimens from children and 12% of egg-negative specimens from adults. The combined criteria of trace proteinuria or more and + blood was more specific than either indicator alone. The combined criteria of at least + + protein (> 100 mg/I00 ml of urine) and + blood showed even higher specificity only in adults (Table 2). However, the sensitivity of the combined criteria was lower than that for either indicator alone. Urinary blood levels and S. haematobium prevalence and egg counts. Among both children and adults, a positive correlation was observed for males and females between semiquantitative urinary blood levels and the prevalence of infection and geometric mean S. haematobium egg counts. The geometric mean egg counts in children were generally higher than those in adults for the same urinary blood levels. Urinary protein levels and S. haematobium prevalence and egg counts. Among both children and adults, a positive correlation between urinary protein levels and prevalence of infection and geometric mean S. haematobium egg counts was observed for males and females. In this highly endemic area, children had higher mean egg counts than adults with similar urinary protein levels, but the differences were not statistically significant (P > 0.14). Table 2. Relationship between S. haematobium egg counts and results of urinary protein and blood determinations by reagent strips in the study population in Adawso, Ghana Percentage of specimens with Protein reading:b >Trace > + + neg. trace > ++ Age group Blood reading:b and egg count' Number examined neg. neg. > + > + > + 5-14 years 0 27 81 26 15 11 11 1-4 20 80 25 35 35 15 5-64 106 93 57 74 72 49 65-256 53 98 68 94 92 66 > 257 36 100 89 100 100 89 Total infected 215 94 62 80 78 57 o 15 years 0 119 65 12 11 9 4 1-4 53 70 19 34 23 8 5-64 65 89 40 54 52 29 65-256 29 93 66 90 86 66 > 257 10 100 70 100 100 70 Total infected 157 84 39 57 52 31 ' Egg counts per random 5 ml urine sample from infected persons. b See text (page 126) for equivalent values of trace, + +, etc. 127 128 K. E. MOTT ET AL. Table 3. Prevalence and intensity of S. haematobium infection in the study population, according to age and sex, in Mutenda, Zambia Males Females Both Age No. Prevalence Egg No. Prevalence Egg No. Prevalence Egg (years) examined (%) count' examined (%) count' examined (%) count' 0-4 0 - - 5 60.0 20.0 5 60.0 20.0 5-9 77 59.7 17.7 77 55.8 11.1 154 57.8 14.1 10-14 200 66.5 14.5 124 66.1 19.1 324 66.4 16.1 15-19 58 63.8 13.4 22 95.5 14.1 80 72.5 13.6 20-24 8 100 4.6 8 100 18.1 16 100 9.1 25-34 4 100 22.3 11 100 55.2 15 100 43.4 35-44 4 100 7.1 14 100 30.9 18 100 22.3 45-54 6 83.3 4.6 24 100 41.2 30 96.7 28.2 55-64 6 83.3 10.8 8 100 61.2 14 92.9 31.4 Total 363 66.7 13.8 293 73.0 20.6 656 69.5 16.7 ° Geometric means of egg counts per random 10 ml urine sample from infected persons. Table 4. Relationship between S. heematobium egg counts and results of urinary protein and blood determinations by reagent strips in the study population in Mutenda, Zambia Percentage of specimens with Protein reading:b Trace + + neg. > trace ++ Age group Blood reading:b and egg count' Number examined neg. neg. > + > + > + 5- 14 years 0 174 39 7 5 5 3 1-4 71 89 34 55 55 20 5-64 186 95 73 86 86 68 65-256 33 100 94 100 100 94 > 257 14 100 93 100 100 93 Total infected 304 94 67 81 81 60 > 15 years 0 24 29 17 0 0 0 1-4 34 79 35 76 74 35 5-64 78 99 69 90 90 64 65-256 28 100 86 100 100 86 > 257 9 100 89 100 100 89 Total infected 149 95 66 89 89 63 ° Egg counts per random 10 ml urine sample from infected persons. b See text (page 1 26) for equivalent values of trace, + +, etc. DETECTION OF SCHISTOSOMIASIS BY REAGENT STRIPS Combined urinary blood and protein levels and S. haematobium prevalence and egg counts. Among both children and adults, a positive correlation between combined urinary blood and protein levels and the prevalence of infection and geometric mean S. haematobium egg counts was observed. About a 10-fold difference was noted in geometric mean egg counts between urine specimens without detectable protein and blood and urine specimens with at least both + blood and trace protein in children (respec- tively, 5.7 vs 72.5 eggs per 5 ml of urine) and in adults (3.7 vs 34.7 eggs per 5 ml of urine). Mutenda, Zambia Study population. A total of 656 persons were interviewed and their urine specimens tested (Table 3). This population was mainly school-age children; a few adults were also examined. The data from urine examinations of five children below 5 years of age were excluded from this analysis. Presence ofurinary blood andprotein in relation to egg count levels. In Mutenda, the presence of either blood (+ or more) or protein (trace or more) was detected in all the urine specimens that had more than 64 eggs per 10 ml of urine (Table 4). The specificity of the trace protein determination was low, particularly in children (61 %/). The specificity of the blood determination was high, particularly in adults (100%). Blood was detected in 81%o of urine speci- mens containing S. haematobium eggs from 5-14- year-old children and in 89% of egg-positive urine specimens from those over 14 years of age. Urinary blood levels and S. haematobuium preva- lence and egg counts. A positive correlation was observed between urinary blood levels and the geometric mean S. haematobium egg counts in both children and adults (Table 5). Prevalences of infection for each semiquantitative urinary blood level were similar between both age groups and between men and women. The geometric mean egg counts at specific blood levels were similar between children and adults. Urinary protein levels and S. haematobium preva- lence and egg counts. Among the children a positive Table 5. S. haematobium prevalence and egg counts,' by age group and sex, in relation to semiquantitative urinary blood readings of reagent strips in Mutenda, Zambia Presence of blood in the urine Total Negative + + + + + + 5-14-year age group: Males: No. of subjects 122 80 36 39 277 % infected 24 96 97 97 65 Egg count 4.6 9.7 26.6 57.4 15.3 Females: No. of subjects 102 46 22 31 201 % infected 28 98 96 97 62 Egg count 5.0 11.2 26.9 56.8 15.8 > 15-year age group: Males: No. of subjects 32 27 11 16 86 % infected 28 100 100 100 73 Egg count 4.3 7.2 17.3 23.1 10.5 Females: No. of subjects 8 18 28 33 87 % infected 88 100 100 100 99 Egg count 4.7 11.0 42.6 58.0 30.2 a Geometric means of egg counts per random 10 ml urine sample from infected persons. 129 K. E. MOTT ET AL. Table 6. S. haematobium prevalence and egg counts,a by age group and sex, in relation to semiquantitative urinary protein readings of reagent strips in Mutenda, Zambia Presence of protein in the urine Total Negative Trace + + + + + + + + + + 5-14-year age group: Males: No. of subjects 70 29 47 66 50 15 277 % infected 14 38 68 94 98 100 65 Egg count 7.3 6.5 5.4 14.0 27.1 96.5 15.3 Females: No. of subjects 54 28 35 42 28 14 201 % infected 15 50 74 91 93 93 62 Egg count 3.8 5.2 9.2 18.9 29.8 65.0 15.8 > 15-year age group: Males: No. of subjects 21 12 10 23 16 4 86 % infected 24 75 100 87 94 100 73 Egg count 2.0 5.0 7.0 14.7 19.1 24.4 10.5 Females: No. of subjects 4 4 20 31 19 9 87 % infected 75 100 100 100 100 100 99 Egg count 5.0 4.1 19.2 25.5 62.2 140.9 30.2 a Geometric means of egg counts per random 10 ml urine sample from infected persons. correlation was also observed between the urinary protein levels and the prevalence of infection (Table 6). No difference in geometric mean S. haematobium egg counts in the urine specimens from children without detectable, trace, or + levels of protein was noted (Table 6), whereas in the urine specimens from adults (over 14 years old) the egg counts were higher in urines with at least + protein compared to negative urines or with trace protein (Table 6). Among adults, a high prevalence (75%) of infection was observed at all levels of proteinuria. Combined urinary blood and protein levels and S. haematobium prevalence and egg counts. Among both children and adults a positive correlation was observed between combined urinary blood and protein levels and the geometric mean S. haemato- bium egg count. No differences in prevalence of infection and geometric mean egg counts at the specific different combined levels of urinary blood and protein were observed between children and adults (Table 7). In children, a 4-fold difference in egg count was observed between negative urines and urines with at least + blood and trace protein (5.5 vs 20.4 eggs per 10 ml). In adults the difference was nearly 9-fold. DISCUSSION In both endemic areas of the present study, the levels of proteinuria and haematuria, as detected by reagent strips, were directly related to urinary S. haematobium egg counts in children and adults. These findings are in agreement with observations made in the endemic areas of the Gambia (4), Sudan (5) and Nigeria (6). Some differences in the results between the two study areas were noted. In both children and adults with similar levels of S. haematobium egg output, for example, the rates of proteinuria and haematuria were higher in Zambia than in Ghana. Furthermore, in children, the geometric mean S. haematobium egg count associated with a specific level of proteinuria or haematuria was generally higher in Ghana than in Zambia. These differences were even more pro- nounced when combined criteria of proteinuria and haematuria were considered. In children and adults, the geometric mean egg counts for similar levels of proteinuria and haematuria were higher in Ghana than in Zambia. The inverse relationship between age and the geometric mean S. haematobium egg count at each 130 DETECTION OF SCHISTOSOMIASIS BY REAGENT STRIPS Table 7. S. haematobium prevalence and egg counts,' by age group, in relation to combined semiquantitative urinary blood and protein readings of reagent strips in Mutenda, Zambia Blood reading: b Negative > + ) + > + ) + Protein reading: b Negative Trace > + > + + +++ 5-14-year age group: No. of subjects 124 254 237 189 101 % infected 15 97 97 97 98 Egg count 5.5 20.4 21.4 28.5 41.2 175-year age group: No. of subjects 24 132 121 94 46 % infected 29 100 100 100 100 Egg count 2.7 23.3 26.9 31.5 47.9 a Geometric means of egg counts per random 10 ml urine sample from infected persons. b See text (page 126) for equivalent values of trace, +, etc. specific level of proteinuria and haematuria has been discussed elsewhere (1). In Ghana the same levels of haematuria were associated with a higher egg count in children than in adults. This was not observed in Zambia in relation to haematuria or to proteinuria. Genetic differences in the populations of Ghana and Zambia may have contributed to these findings. Sickle cell trait, for example, is estimated to occur in 101o of the general population in Ghana (7), and in 20% of the population in the study area in Zambia (8). Painless haematuria is frequently associated with sickle cell trait and may have contributed to the rates of haematuria discovered in these endemic areas. No increased risk of morbidity in concomitant sickle cell trait and S. haematobium infection has been reported, as compared with the risk in either clinical state alone. Differences in the pathogenicity of the S. haemato- bium strains in Ghana and Zambia may also have influenced these observations but this is less likely. Some differences in morbidity associated with different S. haematobium strains have been shown experimentally (9, 10), but it was felt that host vari- ations rather than strain differences were more signi- ficant (9). Wilkins et al. (4) observed the same overall rates of haematuria and proteinuria corresponding to similar egg count levels between populations in Egypt (Nile delta) and the Gambia. Differences in the rates of haematuria and proteinuria between children and adults in these two areas were reported. The rates of haematuria among children and adults with low egg counts were higher in Zambia than among similar age groups in Ghana. The rate of haematuria in Zambia was low in uninfected children (5%) and was not observed in adults. In contrast to the findings from Ghana, proteinuria at + + or greater levels was both a sensitive and a specific indicator of S. haematobium infection among children in Zambia. These observations in Zambia are similar to recent findings in Zimbabwe (11). Low levels of proteinuria (below 100 mg/100 ml of urine) were not specifically related to S. haematobium infection in both study areas. This finding is in agreement with observations in Gambia (4) and Nigeria (6). The present study has shown that regional differences must be considered before establishing criteria for screening for S. haematobium infection. Haematuria was present in 80% of infected children in both Zambia and Ghana, as reported elsewhere (12). On the other hand, haematuria was observed in only 57% of infected adults in Ghana and in nearly 900/o of infected adults in Zambia. In both areas, combined + + proteinuria and haematuria was present in less than 63% of infected persons. It has been suggested that higher levels of proteinuria should be included in the screening procedures to identify heavily infected persons (4, 6). However, the reading of the protein portion of the reagent strips requires additional training for field personnel, and this does not proportionally increase either the sensitivity or the specificity. Haematuria detected by reagent strips identified a high proportion of infected children and adults, compared with those diagnosed by microscopic examination using a filtration technique. Moreover, 97% of heavily infected children (with more than 64 eggs per 5 ml urine) were found to have haematuria by the reagent strips, the use of which has therefore been suggested as an indirect diagnostic technique to identify heavily infected persons in public health programmes, particularly among school-age children (4-6, 11, 12). Because of variations between countries, preliminary evaluation at a country level is necessary to establish the criteria for their use. 131 132 K. E. MOTTET AL. ACKNOWLEDGEMENTS The authors gratefully acknowledge the advice and encouragement of Dr A. Davis, Director, Parasitic Diseases Programme, World Health Organization, Geneva, Switzerland. This study received financial support from the Parasitic Diseases Programme of the World Health Organization and from the UNDP/World Bank/WHO Special Programme for Research and Training in Tropical Diseases. RtSUMt EVALUATION DES BANDELETTES DE PAPIER REACTIF DANS LES EPREUVES URINAIRES DE DETECTION DE L'INFECTION A SCHISTOSOMA HAEMATOBIUM: ETUDE COMPARATIVE AU GHANA ET EN ZAMBIE La pr6valence, l'intensit6 et la morbidit6 de la schisto- somiase urinaire en Afrique varient selon l'6pid6miologie, les modes de transmission et l'ecologie de chaque zone d'end6mie. La pr6sence d'h6maturie et de prot6inurie, d6cel6e au moyen de bandelettes de papier r6actif, a ete compar6e avec les num6rations d'aeufs de Schistosoma haematobium dans les urines de sujets provenant de deux r6gions 6pid6miologiquement distinctes du Ghana et de Zambie. A Adawso, Ghana, 562 personnes ont 6te examin6es. La pr6valence et l'intensit6 maximales de l'infection ont ete observ6es dans le groupe d'age 10-14 ans. On a trouve du sang dans 97% des 6chantillons d'urine comptant plus de 64 oeufs par 5 ml, et dans 86% des 6chantillons d'urine comptant plus de 16 ceufs par 5 ml. La recherche des prot6ines etait d'une bonne sensibilit6 mais d'une faible sp6cificit6. En associant les criteres de prot6inurie et d'h6maturie, on obtenait une meilleure sp&cificit6 qu'avec chacun d'entre eux, mais la sensibilit6 6tait alors plus faible qu'avec l'un ou l'autre pris isol6ment. Chez I'adulte comme chez l'enfant, on a observ6 une correlation positive entre les taux semi-quantitatifs de sang et de proteines dans les urines et la pr6valence de l'infection, de meme qu'avec la moyenne g6om6trique du nombre d'oeufs de S. haematobium. A Mutenda, Zambie, parmi 656 sujets principalement d'age scolaire, on a d6cele du sang et des proteines dans tous les 6chantillons d'urine comptant plus de 64 ceufs par 10 ml. La recherche de l'hematurie 6tait d'une bonne sp6cificite, surtout chez les adultes (10001.). Comme au Ghana, on a observ6 une corr6lation positive entre les taux urinaires de sang et de proteines et la prevalence de l'infection ainsi qu'avec la moyenne g6ometrique du nombre d'oeufs de S. haematobium. Chez les enfants et les adultes ayant une oviurie analogue, I'hematurie et la prot6inurie 6taient plus 6lev6es en Zambie qu'au Ghana. De plus, les nombres d'oeufs associ6s a un taux sp6cifique d'h6maturie ou de proteinurie 6taient g6n6ralement plus 6lev6s au Ghana qu'en Zambie. La fr6quence du trait dr6panocytaire dans la zone d'6tude etant, semble-t-il, plus 6lev6e en Zambie (2001) qu'au Ghana (100o), cela pourrait expliquer cette diff6rence. Bien qu'il existe aussi des diff6rences entre les souches de S. haematobium en pr6sence, elles ne sont probablement pas responsables des differences entre les taux de morbidite observ6s. Avant d'6tablir les crit&es de d6pistage de l'infection a S. haematobium, il faut examiner les differences regionales. L'h6maturie d6celee par les bandelettes de papier r6actif permet d'identifier une proportion importante d'enfants et d'adultes infect6s, par comparaison avec le diagnostic microscopique apres filtration de l'urine. Du fait des differences entre pays, une evaluation preliminaire au niveau du pays est necessaire pour etablir les criteres d'infection avant de proc6der au depistage a l'aide de bandelettes de papier r6actif. REFERENCES 1. Morr, K. E. ET AL. The relationship between intensity of Schistosoma haematobium infection and clinical haematuria and proteinuria. Lancet, 1: 1005-1008 (1983). 2. WHO Technical Report Series, No. 643, 1980 (Epidemiology and control of schistosomiasis: report of a WHO Expert Committee). 3. Scorr, D. ET AL. Epidemiology of human Schistosoma haematobium infection around Volta Lake, Ghana, 1973-75. Bulletin of the World Health Organization, 60: 89-100 (1982). 4. WILKINS, H. ET AL. The significance of proteinuria and haematuria in Schistosoma haematobium infection. Transactions of the Royal Society of Tropical Medi- cine, 73: 74-80 (1979). 5. FELDMEIER, H. ET AL. Simultaneous use of a sensitive filtration technique and reagent strips in urinary schis- tosomiasis. Transactions of the Royal Society of Tropical Medicine and Hygiene, 76: 416-421 (1982). 6. PUGH, R. H. H. ET AL. Malumfashi Endemic Diseases Research Project, XV. The potential medical impor- tance of bilharzia in northern Nigeria: a suggested DETECTION OF SCHISTOSOMIASIS BY REAGENT STRIPS 133 rapid, cheap and effective solution for control of Schistosoma haematobium infection. Annals of tropical medicine and parasitology, 74: 597-613 (1980). 7. RINGELHANN, B. ET AL. A survey of haemoglobin variants, thalassemia, glucose-6-phosphate dehydro- genase deficiency in northern Ghana. Ghana medical journal, 7: 120-128 (1968). 8. BARCLAY, G. P. T. & SPLAINE, M. The distribution of sickle cell trait in Zambia. Tropical and geographical medicine, 24: 393-400 (1972). 9. WEBBE, G. & JAMES, C. A. A comparison of two geographical strains of Schistosoma haematobium. Journal of helminthology, 45: 271-284 (1971). 10. WRIGHT, C. A. & KNOWLES, R. J. Studies on Schisto- soma haematobium in the laboratory. III. Strains from Iran, Mauritius and Ghana. Transactions of the Royal Society ofMedicine and Hygiene, 66: 108-118 (1972). 11. TAYLOR, P. Proteinuria as a simple diagnostic test for urinary schistosomiasis in schoolchildren in the rural areas of Zimbabwe. Central African journal of medicine, 28: 216-219 (1982). 12. BRIGGS, M. ET AL. Screening with reagent strips. British medical journal, 3: 433-434 (1971).