The American Journal of Tropical Medicine and Hygiene - Volume 62, Issue 2_suppl, February 2000

This supplement is a report on the Epidemiology 1, 2, 3 (EPI 1, 2, 3) investigation, its origins, evolution, and findings that were carried out over a period beginning in 1990 and ending in 1994 in Egypt. The large scope and size of the study, the largest to date on schistosomiasis in Egypt, was a rationale for publishing a supplement to document EPI 1, 2, 3 methods and results collectively in sufficient detail to serve as a reference for planning, designing, and analyzing future epidemiologic studies and evaluation of schistosomiasis control in Egypt. The 3 objectives of EPI 1, 2, 3 were to 1) determine the changing patterns of Schistosoma haematobium and S. mansoni, 2) investigate factors contributing to differences between villages in the Nile Delta, Middle Egypt, and Upper Egypt, and 3) investigate risk factors for morbidity. The objectives were addressed using standardized techniques, stool and urine examinations, clinical examinations (including abdominal ultrasound), and questionnaires on a selected sample of the populations of selected villages in 9 governorates in Egypt.

Bad sample designs and selection bias have plagued studies on schistosomiasis, and as a result some believe that schistosomiasis is too focal, making it difficult to draw reliable samples. The Epidemiology 1, 2, 3 (EPI 1, 2, 3) sample design, although complex, demonstrates that sampling theory is readily applicable to epidemiologic studies of schistosomiasis. The EPI 1, 2, 3 sampling scheme was designed to achieve the smallest feasible standard errors given EPI 1, 2, 3 objectives and certain logistical constraints. The sample design is a multi-stage selection of villages (ezbas, which were stratified by size) and households within each of 9 purposely selected Egyptian governorates. Villages and households were systemically selected from census frames. The sampling of ezbas was especially difficult because of the lack of complete sampling frames and their wide variation in population size. Ultimately, ezbas were stratified by size and then randomly selected from each stratum. Sample sizes for villages and ezbas and individuals within ezbas were calculated based on EPI 1 and 2 objectives, respectively. No re-selection was made for non-respondents. A 20% subsample of the full sample was drawn for clinical and ultrasonographic examinations. The sample selected from individual governorates closely parallel the age structure of the 1986 census of the respective rural populations. Details of the study design and related methods are given below.

Accuracy of data is of paramount concern for all research. The task of providing objective assurances of accuracy of parasitologic data for a large, multi-center epidemiologic research project in Egypt (Epidemiology 1, 2, 3 [EPI 1, 2, 3]) presented a unique set of challenges undertaken jointly by the Ministry of Health's Qalyub Center for Field and Applied Research with technical assistance from Tulane University (New Orleans, LA). The EPI 1, 2, 3 project was part of large bilateral research program, the Schistosomiasis Research Project, undertaken jointly by the governments of Egypt and the United States. This paper describes the nature of the quality control system developed to accomplish this task, presents results and discusses the findings.

Morbidity due to schistosomiasis was evaluated in a subpopulation of 14,000 of the randomized sample in the Epidemiology 1, 2, 3 Project. It was measured by using a standardized questionnaire for obtaining medical history and symptoms and by performing standardized physical and ultrasound examinations. Reported herein are descriptions of the methods and training and quality control efforts made to insure that the morbidity data was reliable and consistent when collected by 7 different teams in 9 different governorates.

This is a descriptive report of the Epidemiology 1, 2, 3 project in Egypt that made use of large probability sampling methods. These results focus on Schistosoma mansoni infection in the northern Nile Delta governorate of Kafr El Sheikh. A probability sample of 18,777 persons, representing the rural population of the entire governorate, was drawn. The sample was designed not to exclude villages based on location or presence of health care facilities and to include representation of the smaller ezbas or hamlets. The objective was to obtain detailed estimates on age- and sex-specific patterns of S. mansoni infection, and to provide a baseline for prospective studies. Stool specimens were examined by the Kato method. The estimated mean +/- SE prevalence of S. mansoni infection in the rural population was 39.3 +/- 3.3% in 44 villages and ezbas after weighing for the effects of the sample design. The estimated mean +/- SE geometric mean egg count per gram of stool (GMEC) was 72.9 +/- 7.3. Prevalence and GMEC varied considerably by village and ezba, with ezbas having a significantly higher prevalence. Villages and ezba-specific prevalence was strongly associated with GMEC (r2 = 0.61, P < 0.001). The prevalence of S. mansoni infection increased by age to 55.4 +/- 3.2% at age 16 without a significant change in the adult ages. There were no gender differences until age 6, after which males were consistently higher until middle age, when the differences converged. The age- and sex-specific pattern of GMEC varied widely; however, when the GMEC data were collapsed into 5-year age groups, the GMEC peaked at 81.5 +/- 12.1 eggs/g in the 10-14-year-old age group. These estimates provide the basis for evaluating control measures for reducing prevalence, intensity of infection, and transmission.

Health questionnaires and parasitologic examinations of urine and stool were performed upon a stratified random sample of 10,899 individuals from 1,537 households in 27 rural communities in Menofia Governorate in Egypt in 1992 to investigate the prevalence of, risk factors for, and changing pattern of infection with Schistosoma sp. in the governorate. A subset, every fifth household, or 1,480 subjects, had physical and ultrasound examinations to investigate prevalence of and risk factors for morbidity. The prevalence of S. mansoni ranged from 0.3% to 72.9% and averaged 28.5%. The geometric mean egg count was 81.3 eggs/gram of stool. Age-stratified prevalence and intensity of infection was 30-40% and 60-80 eggs/gram of stool from the age of 10 onward; males had higher infection rates and ova counts than females in all age groups > 10 years old. Schistosoma haematobium was rare, being consequential in only 1 community. Risk factors for S. mansoni infection were male gender; age > 10 years; living in smaller communities; exposures to canal water; history of or treatment for schistosomiasis or blood in the stool; detection of splenomegaly by either physical or ultrasound; and ultrasound-detected periportal fibrosis (PPF). The more severe grades of PPF were rarely (21 of 1,450 examinations) detected. Risk factors for morbidity, i.e., ultrasound-detected PPF, were similar to those for infection. Schistosoma mansoni has almost totally replaced S. haematobium in Menofia. The prevalence of S. mansoni in rural communities remains high and average intensities of infection are moderate. The association of morbidity with schistosomal infection was variable and is obviously markedly influenced by both the frequent use of antischistosomal chemotherapy in communities in Menofia and by the prevalence of complications from chronic viral hepatitis in the population: hepatomegaly did not correlate with infection; PPF and splenomegaly, however, were related to S. mansoni infection in both individuals and communities.

A multi-stage stratified sample of 12,515 individuals from 1,941 households in 42 villages in the Ismailia governorate of Egypt were surveyed for schistosomal infection. A subset of 2,390 subjects were surveyed for morbidity by physical and ultrasonographic examination. The prevalence of Schistosoma mansoni infection in rural Ismailia was 42.9% and the geometric mean egg count (GMEC) was 93.3 eggs/gram of stool, with considerable variability between communities. Prevalence and intensity peaked in the 20-30-year-old age group and was higher in males than in females. Prevalence and intensity of S. haematobium was very low: 1.8% and 3.5 ova/10 ml of urine GMEC, respectively. Canal water exposure risk factors for S. mansoni infection were males bathing (odds ratio [OR] = 2.2), females washing clothes (OR = 1.9), and children playing or swimming (OR = 2.3). Presence of in-house piped water supply and latrine lowered infection rates (P < 0.001 and P = 0.002, respectively). Histories of S. mansoni infection (OR = 1.6) or treatment (OR = 1.5) and blood in feces (OR = 3.5) were associated with infection. Hepatomegaly (16.0%) was more frequently detected than splenomegaly (3.6%) by physical examination, with both being more frequent in older age groups. Splenomegaly, but not hepatomegaly, was associated with presence of S. mansoni ova in stools (OR = 1.4) and the community burden of infection (P = 0.02). Ultrasonographically detected hepatomegaly, splenomegaly, and periportal fibrosis (PPF) were detected in 43.0%, 17.4%, and 39.7% of the subjects, respectively. The higher grades of PPF were rare. Ultrasonographically detected splenomegaly, not hepatomegaly, was associated with S. mansoni infection, community burden of infection, and PPF. Risk factors for PPF were the same as for S. mansoni infection. There was a marginal association of PPF with infection and none (P = 0.33) with the intensity of infection in individuals or in the community. We conclude that in rural Ismailia, S. haematobium infection is rare but the prevalence and intensity of infection with S. mansoni is high. The risk of infection is associated with environmentally detected factors and behaviors. Hepatosplenic morbidity attributable to S. mansoni infection is low, presumably because of the favorable effect of wide application of praziquantel therapy.

Health questionnaires and parasitologic examinations of urine and stool were performed upon a stratified random sample of 14,344 individuals from 1,952 households in 34 rural communities in Gharbia Governorate of Egypt to investigate the prevalence of, risk factors for, and changing pattern of infection with Schistosoma sp. A subset, every fifth household, of 1,973 subjects had physical and ultrasound examinations to investigate prevalence of and risk factors for morbidity. Community prevalence of Schistosoma mansoni ranged from 17.9% to 79.5% and averaged 37.7%. The geometric mean egg count (GMEC) was 78.9 eggs/gram of feces. The prevalence and intensity of infection was 40-50% and 70-100 eggs/gram of feces in those > or =10 years of age. Schistosoma haematobium was detected in 5 of the 34 communities. The maximum infection rate was 2.8% and mean GMEC in the five communities was 2.1/10 ml of urine. The overall prevalence of S. haematobium in the governorate was 0.3%. Risk factors for infection with S. mansoni were male gender, an age >10 years, living in smaller communities, exposures to canal water, prior therapy for schistosomiasis, or blood in the stool (in children only). Morbidity detected by physical examination or ultrasonography did not correlate with S. mansoni infection in individuals with the exception of periportal fibrosis (PPF, odds ratio [OR] = 1.25). Periportal fibrosis was detected in more than half of the subjects by ultrasonography; 5.3% had grade II lesions and 1.0% had the most severe grade III changes. Risk factors for morbidity as manifested by ultrasonographically detected PPF were similar to those for infection. Periportal fibrosis had a negative relationship with abdominal pain (OR = 0.45) and hepatomegaly detected by physical examination and ultrasonography (ORs = 0.72 and 0.68), but it was associated with splenomegaly (ORs = 4.14 and 3.55). The prevalence of PPF, hepatomegaly, and splenomegaly increased with age. There was no relationship between community burden of schistosomiasis mansoni and any measurements of morbidity with the exception of splenomegaly detected by physical examination (r = 0.40). Schistosoma mansoni has almost completely replaced S. haematobium in Gharbia, which has a high prevalence and moderate intensity of S. mansoni infection. Periportal fibrosis was detected by ultrasonography in more than half of the subjects, and 1 in 16 had grade II and III lesions. The only relationship between PPF and other morbidity findings was its positive relationship with splenomegaly and negative association with hepatomegaly. Hepatic morbidity is common in communities in Gharbia but the role of schistosomiasis mansoni in this is uncertain.

The primary objectives of this study, carried out in Qalyubia Governorate in Egypt (south-central Nile Delta), were to continue tracking historical trends of infection prevalence of Schistosoma mansoni and S. haematobium, determine whether satellites (ezbas) of mother villages differed significantly with respect to schistosomiasis transmission, and to asses schistosomiasis-induced morbidity on a population basis using ultrasonography. Our study revealed that S. haematobium has virtually disappeared from Qalyubia governorate, and that S. mansoni prevalence continues to decline slowly (17% in 1991 compared with 19% in 1990). The prevalence of intestinal schistosomiasis was actually higher in the mother villages than in the ezbas of the same villages, indicating that prevalence based on surveys of villages alone did not (at least for Qalyubia) cause underestimates of true prevalence. (A mother village is the large village in an area that includes hamlets or ezbas. In many areas, the infection rate in ezbas is significantly higher than in the larger central village.) Ultrasonographic studies revealed that less than 3% of the population had stage 2 or stage 3 periportal fibrosis, commonly associated with chronic schistosomiasis mansoni. This low level of morbidity was consistent with earlier data from Qalyubia, which also showed a low level of S. mansoni-induced morbidity in this governorate.

Health questionnaires and parasitologic examinations of urine and stool were performed upon a stratified random sample of 7,710 individuals from 1,109 households in 21 rural communities in Fayoum Governorate, Egypt in 1992 to investigate the prevalence of, risk factors for, and changing pattern of, infection with Schistosoma sp. in the governorate. A subset, every fifth household, or 1,038 subjects, had physical and ultrasound examinations to investigate prevalence of, and risk factors for, morbidity. The prevalence of S. haematobium ranged from 0% to 27.1% and averaged 13.7%. The geometric mean egg count (GMEC) was 10.0 eggs/10 ml of urine. Age-stratified prevalence and intensity of infection were 18-25% and 10-15 eggs/10 ml of urine in those 5-25 years of age. Schistosoma mansoni were detected in inhabitants of 13 communities, but 78.5% of the infections were focally present in a group of 4 satellite hamlets around a single village. The overall prevalence of S. mansoni in the governorate was 4.3% and the GMEC was 44.0 ova/g of stool. Risk factors for infection with either species were male gender, an age <20 years, living in smaller communities, and exposures to canal water by males. Histories of burning micturation, blood in the urine, or prior schistosomiasis and reagent strip-detected hematuria and proteinuria were risks for S. haematobium, but not for S. mansoni. Both urinary tract and higher grades of hepatic morbidity were rare. Obstructive uropathy was present in 6.3% of the subjects and was more common in males and older people. Ultrasonography-detected bladder lesions were present in 5.2% and correlated with S. haematobium only in younger subjects and in those with hematuria and proteinuria. The prevalences of hepatomegaly, splenomegaly, and periportal fibrosis (PPF) were associated with each other and increased with age and in males. Ultrasonography-detected hepatomegaly and splenomegaly were weakly associated with S. mansoni infections only in children. The prevalence of PPF was greater in the 4 communities with >25% S. mansoni infection rates in comparison with the 17 other villages and ezbas. Transmission of S. mansoni is focally well established in Fayoum, which also has the highest prevalence of S. haematobium in the governorates surveyed by the Epidemiology 1, 2, 3 Project. However, both urinary tract and hepatic morbidity are relatively rare in the governorate. This probably results from the long-standing schistosomiasis control program in Fayoum, which suppressed intensity more than prevalence of infection, leading to less community morbidity.

Risk factors, prevalence, and intensity of infection with Schistosoma sp. and prevalence and magnitude of morbidity caused by schistosomiasis was assessed in a stratified random sample of 16,433 subjects from 2,409 households in 33 rural communities in Minya Governorate, Egypt. The prevalence of S. haematobium ranged from 1.9% to 32.7% among the communities and averaged 8.9%. The average intensity of infection was a geometric mean egg count (GMEC) of 8.5 per 10 ml of urine and ranged from 1.6 to 30.9. Prevalence was maximum (18-20%) in those 10-20 years of age and higher in males than in females. Intensity of infection followed the same pattern. Infection with S. mansoni was present almost exclusively in a single village, confirming spread of this species up the Nile River and its focality in Minya. Risk factors for S. haematobium infection were an age from 11 to 20; male gender; males bathing in, women washing clothing or utensils in, and children swimming or playing in canals; and a history of, or treatment for, schistosomiasis. Recent history of burning micturition was associated with infection in children but not in adults, while a history of blood in urine correlated with S. haematobium infection in both age groups. Reagent strip-detected hematuria and proteinuria were highly associated, particularly in children, with S. haematobium infection. The presence of hepatomegaly or splenomegaly on physical examination was not associated with S. haematobium ova in the urine. Hepatomegaly, as measured by ultrasonography in the midclavicular line or the midsternal line, or ultrasonography-detected splenomegaly were not present more frequently in infected subjects than in uninfected subjects. Schistosoma ova were not detected more frequently in urine of subjects with ultrasonography-detected periportal fibrosis than in the urine from subjects without this finding. Ultrasonography-detected urinary bladder wall lesions were detected in only 6 (0.3%) subjects and obstructive uropathy was observed in 54 (2.7%) subjects. The absence of an association between prevalence of urinary tract morbidity and S. haematobium infections was surprising. Two possible explanations are 1) that repeated chemotherapy has reduced the prevalence of urinary tract morbidity and 2) that morbidity was not being detected by the ultrasonographic operators.

In the Assiut, Egypt Epidemiology 1, 2, 3 investigation, a sample of 14,204 persons in 10 villages, 31 ezbas (satellite communities), and 2,286 households was drawn from a rural population of 1,598,607. Parasitologic examination of urine and stool were made for Schistosoma haematobium and S. mansoni, and physical and ultrasound examinations were made on a 20% subsample. The overall estimated prevalence of S. haematobium was 5.2 +/- 0.5 (+/- SE). This varied considerably by village and ezba, ranging from 1.5% to 20.9%, with ezbas having a slightly higher overall prevalence than villages. The overall estimated geometric mean egg count was 6.6 +/- 0.5 eggs per 10 ml of urine and was consistently low throughout the communities. Infection with S. haematobium was associated with age (peak prevalence of 10.6 +/- 1.5% in 15-19-year-old age group) males, children playing in the canals, a history of blood in the urine, and reagent strip positivity for hematuria and proteinuria. The prevalence of either hepatomegaly or splenomegaly detected by physical examination was low (4.0% and 1.5%, respectively). The prevalence of hepatomegaly determined by ultrasonography was substantially higher, 24.1%. The prevalence of periportal fibrosis (PPF) was 12.0%, but grade II or III PPF was present in less than 1%. Ultrasonography-determined hepatomegaly, in both the midclavicular line and the midsternal line, increased by age to more than 30%. Periportal fibrosis was more common in the age groups in which infection rates were the highest. At the village and ezba level of analysis, the prevalence of hepatomegaly, splenomegaly, and PPF tended to be higher in communities having the highest prevalence of infection with S. haematobium.

Qena is the southernmost governorate of Egypt included in the Epidemiology 1, 2, 3 national study. A probability sample selected 17,822 individuals from 2,950 households in 34 ezbas and 10 villages from a total rural target population of 1,731,252 (based on the most recent 1986 census of the population by the Egyptian Central Agency for Public Mobilization And Statistics). Parasitologic examination of urine and stool were made for Schistosoma haematobium and S. mansoni, respectively, and physical and ultrasound examinations were made on a 20% subsample. The overall estimated prevalence of S. haematobium was 4.8 +/- 0.7% (+/-SE) and geometric mean egg count (GMEC) was 7.0 ova per 10 ml of urine. Considerable variation in prevalence was observed between the villages and ezbas, ranging from 0.0% to 20%, with the smaller ezbas having a slightly higher overall prevalence. The age- and sex-specific patterns of S. haematobium showed typical peak prevalence in early adolescence, with males having a higher prevalence than females. A history of hematuria was associated with current infection (odds ratio = 3.6, 95% confidence interval = 2.32-5.63). Hepatomegaly and splenomegaly determined by physical examination present in 7.9% and 3.0%, respectively. Ultrasonography-determined hepatomegaly of the left liver lobe was found in 10.1%. Ultrasonography-detected hepatomegaly in both the left and right lobes increased in prevalence from approximately 5% in children to 15-20% in adults. The prevalence of ultrasonography-detected splenomegaly increased slightly with age. Grade III periportal fibrosis was detected in only 2 individuals in the sample. Bladder wall lesions and obstructive uropathy were also very infrequent. Other associations with these measures are given. Most villages and ezbas had an S. mansoni prevalence of less than 1%. The exception was Nag'a El-Sheikh Hamad, where the prevalence was 10.3 +/- 0.5% (GMEC = 57.4 +/- 2.6). Two other communities also had a prevalence >1% (Ezbet Sarhan and Kom Heitin).

Health questionnaires and parasitologic examinations of urine and stool were evaluated from a stratified random sample of 89,180 individuals from 17,172 households in 251 rural communities in 9 governorates of Egypt to investigate the prevalence of, risk factors for, and changing pattern of infection with Schistosoma sp. in Egypt. A subset, every fifth household, or 18,600 subjects, had physical and ultrasound examinations to investigate the prevalence of and risk factors for morbidity. Prevalence of S. haematobium in 4 governorates in Upper Egypt in which it is endemic ranged from 4.8% to 13.7% and averaged 7.8%. The geometric mean egg count (GMEC) ranged from 7.0 to 10.0 ova/10 ml of urine and averaged 8.1. Age stratified prevalence of infection peaked at 15.7% in the 10-14-year-old age group and decreased to 3.5-5.5% in all groups more than 25 years of age. Age-stratified intensity of infection peaked at approximately 10.0 ova/10 ml of urine in the 5-14-year-old age groups and was about half that in all groups more than 25 years of age. Males had higher infection rates and ova counts than females in all age groups. Schistosoma mansoni was rare in Upper Egypt, being consequential in only Fayoum, which had a prevalence of 4.3% and an average intensity of infection of 44.0 ova/g of stool. Risk factors for S. haematobium infection were male gender, an age <21 years old, living in smaller communities, exposures to canal water; a history of, or treatment for, schistosomiasis, a history of burning micturition or blood in the urine, and reagent strip-detected hematuria or proteinuria. The more severe grades (II and III) of ultrasonography-detected periportal fibrosis (PPF) were rare (15 of 906) in these schistosomiasis haematobia-endemic governorates. Risk factors for morbidity (ultrasonography-detected urinary bladder wall lesions and/or obstructive uropathy) were similar to those for infection, with the exception that risk progressively increased with age. Subjects with active S. haematobium infections were 3 times as likely as those without active S. haematobium infections to have urinary tract morbidity. The prevalence of S. mansoni in 5 governorates in Lower Egypt, where it is endemic, ranged from 17.5% to 42.9% and averaged 36.4%. The GMEC ranged from 62.6 to 93.3 eggs/g of stool and averaged 81.3. Age-stratified prevalence of infection peaked at 48.3% in the 15-19-year-old age group, but averaged 35-45% in all groups more than 10 years of age. The intensity of infection was highest in the 10-14-year-old age group, and showed a range of 70-85 eggs/g of stool in those > or =5 years of age. Males had higher infection rates and ova counts than females in all age groups. Schistosoma haematobium was rare in these governorates; Ismailia (1.8%) had the highest infection rate. Risk factors for S. mansoni were male gender, an age >10 years old, living in smaller communities, exposures to canal water, a history of, or treatment for, schistosomiasis or blood in the stool, detection of splenomegaly by either physical examination or ultrasonography, and ultrasonography-detected PPF. The more severe grades (II and III) accounted for 463 (13.3%) of the 3,494 having ultrasonography-detected PPF. Risk factors for morbidity (ultrasonography-detected PPF) were similar to those for infection except that inhabitants of smaller communities were not at increased risk. Active S. mansoni infection increased the odds ratio (OR) of having PPF by 1.37. In comparison with others with normal-size livers, subjects having hepatic enlargement in either the midclavicular line or the midsternal line detected by physical examination or ultrasonography had a reduced risk (ORs = 0.64-0.72) of PPF. The prevalences of lesions detected by ultrasonography were 23.7% for enlargement of right lobe of the liver, 11.3% for enlargement of left hepatic lobe, 20.6% for splenomegaly, and 50.3% for PPF. Schistosoma mansoni has almost totally replaced S. haematobium in Lower E