| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
 |
ABSTRACT |
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
INTRODUCTION |
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Hematuria and self-reported blood in urine have been shown to be simple and useful indicators for urinary schistosomiasis.12,13 However, no such indicator currently exists for intestinal schistosomiasis or STHs. Rather, microscopic examination of stool specimens continues to be the standard method for assessing prevalence and intensity of these infections. Routine screening of stool specimens, however, is labor-intensive and expensive and thus is not practical in many settings. Alternative methods or sampling schemes to identify high-risk populations are needed to improve the efficiency of mass treatment programs by targeting areas where the prevalence is high. Alternative indicators that have been suggested include self-reported blood in the stool and self-reported exposure to infective water bodies.
Geographic indicators may also be useful in identifying and targeting mass treatment programs for both schistosomiasis and STH infections. In a recent study in Tanzania, for example, the prevalence of Schistosoma mansoni was associated with proximity to Lake Victoria, while the prevalence of S. haematobium increased with distance from the lake.14 Brooker and others15 also found a strong association between proximity to Lake Victoria and prevalence of S. mansoni. While schistosomiasis tends to have a focal distribution, little is known regarding the spatial distribution of hookworm, roundworm and whipworm infections.
The objective of the present study was to determine the spatial distribution of intestinal and urinary schistosomiasis as well as STH infections in a well-defined population living near the shore of Lake Victoria in preparation for a combined treatment program.
|
METHODS |
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
This was the first large study of schistosomias in the Asembo area and no information was available on the prevalence of infection or the magnitude of morbidity. However, a study in nearby Kisumu town showed very high levels of S. mansoni infection among car washers working along the shore of Lake Victoria.17 These results suggested that rural communities along the lake might have high prevalences as well. The study was reviewed and approved by the Scientific and Ethical Review Committees of the Kenya Medical Research Institute and the Institutional Review Board of the Centers for Disease Control and Prevention.
To determine the geographic distribution of the infections with respect to Lake Victoria, a 10 x 13 km area bordering the lake was chosen and all of the primary schools in this area (n = 33) were selected for participation. One boarding school was removed from the list because the students came from outside the region. All others were included in the study. A list of registered students between 10 and 12 years of age was obtained from the headmaster at each school and age was confirmed with the parent while obtaining informed consent. Thirty-five to forty students were randomly chosen from each school list using a random number generator. Informed consent was obtained from the parent or guardian and assent was obtained from the student prior to enrollment in the study. Each student was provided a container and asked to provide an early morning stool on the following day. Approximately 30 mL of urine was collected from each student and stored in 50-mL polypropylene tubes. Urine specimens were collected between the hours of 10:00 AM and 2:00 PM on the day of the visit. All specimens were transported to The Kenya Medical Research Institute and processed within six hours of collection.
Each student also completed a short questionnaire, which was administered by trained interviewers in the local language. Students were asked to estimate the amount of water contact they had with Lake Victoria, categorized as none, some, or frequent contact, as well as contact with nearby streams and ponds during the past year. Water contact was further categorized by the type of activity such as swimming, fishing, bathing, or collecting water for household use. The students were also asked whether they noticed any blood in their urine or stool during the previous month and whether they had recently experienced any pain on urination.
Each specimen was analyzed in duplicate by the Kato-Katz method for eggs of S. mansoni, Ascaris lumbricoides, Trichuris trichiura, and hookworms (Necator americanus and Ancylostoma duodenal).18 A template was used that when filled contained approximately 41.7 mg of feces. Eggs were counted by two independent microscopists and any discrepancy in results was reconciled. The polypropylene tubes containing urine were shaken to resuspend the eggs and 10 mL was withdrawn with a syringe. Urine was filtered using Millipore (Billerica, MA) filters and analyzed microscopically for S. haematobium eggs. Urine specimens were also tested for hematuria using Ames Hemastix reagent strips (Miles Inc., Elkhart, IN). Egg counts for S. haematobium were expressed as eggs per 10 mL of urine and counts for S. mansoni were expressed as eggs per gram (epg) of feces. Intensity of infection was categorized as light (< 100 epg), moderate (100–399 epg), and heavy (
400 epg).19 Schistosoma haematobium and STH infections were categorized as either positive or negative.
As part of an ongoing Demographic Surveillance System, the geographic locations of all schools and individual households as well as roads, water sources, and the shoreline of Lake Victoria were determined using a differential global positioning system20 and mapped using ArcView version 3.2 software (Environmental Systems Research Institute, Inc., Redlands, CA). Distance was calculated as the shortest overland route between the school or house and the lake shoreline or other water bodies. The distance computations account for the curvature of the earth by computing arc length instead of linear distance and are the basis for spatial analysis.21 Questionnaire and laboratory data were entered into Epi Info version 6.04 (Centers for Disease Control and Prevention, Atlanta, GA). Correlation analysis, chi-square tests, and multivariate models were performed using SAS version 8.03 (SAS Institute, Cary, NC). Multivariate models controlled for clustering within schools using general estimating equations.
All children who tested positive for schistosomiasis were treated with praziquantel (40 mg/kg) and children testing positive for roundworms, whipworms, or hookworms were treated with albendazole (400 mg).
|
RESULTS |
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
). Only three cases of S. haematobium were detected, giving a mean prevalence of 0.2%. Schistosoma mansoni infections were predominately light (62.7%, 1–99 epg), with 27.4% considered moderate (100–399 epg) and 9.8% considered heavy infections ( 400 epg). The prevalence increased with each year of age (P < 0.003), which is consistent with typical age-prevalence curves that peak in early adolescence.22 Prevalence was slightly higher in females (18.5%) than males (14.5%; P = 0.06).
|
). School prevalence was strongly correlated with log distance to the shoreline (r = -0.89, P < 0.001). Three schools had a prevalence greater than 50% and all were located within 1.5 km of the shore. Eight schools had a prevalence greater than 20% and were all located within 4 km of the shore (Figure 2).
|
|
. Distance from each child’s home to the shoreline was also calculated. The relationship between prevalence of infection with S. mansoni and distance from the household to the lake, summarized by 1-km intervals, was similar to the relationship between prevalence and distance of the schools to the lake. Prevalence decreased from 60% among children living between 0 and 1 km from the shoreline to 22% for children living between 3 and 4 km from the shoreline. Beyond the 4 km line, the prevalence decreased further to 2.8% for children living beyond 9 km from the shoreline (Figure 2
). Prevalence was not associated with distance from the household to either ponds or streams in the study area, suggesting that Lake Victoria was the major source of S. mansoni infection.
|
). Children infected with S. mansoni were more likely to have had contact with lake water during the previous year (odds ratio [OR] = 5.9, 95% confidence interval [CI] = 4.1, 8.6) but less likely to have had contact with ponds (OR = 0.62, 95% CI = 0.43, 0.91) or rivers (OR = 0.5 95%, CI = 0.34, 0.74). Swimming, fishing, bathing, washing clothes, and collecting water from Lake Victoria were all associated with a higher risk of infection, while collecting water from ponds or rivers as well as bathing or washing clothes in rivers was associated with a lower risk of infection.
|
). Those who reported frequent contact with lake water had a prevalence of 52.4%, those with occasional contact had a prevalence of 24.5%, and those reporting no contact during the last year had a prevalence of 5.9% (P < 0.001, by chi-square test for trend).
|
). Specific activities associated with infection were swimming, fishing, and collecting water in Lake Victoria. Swimming in ponds was marginally associated with lower risk of infection (P = 0.08).
Approximately 23% of the children with intestinal schistosomiasis lived beyond 4 km from the shoreline, but seven (35%) of the 20 heavy infections occurred beyond this distance (Table 4). We included questions on self-reported contact with lake water and self-reported blood in the stool to see if responses to these questions would be helpful in detecting schistosomiasis-infected children in low prevalence areas. Of 748 children living beyond 4 km from the shore, 273 (36.5%) reported some or frequent contact with lake water during the last year and 34 of these children tested positive for infection (positive predictive value = 12.5%). For those living beyond 4 km from the shoreline, the sensitivity and specificity of self-reported contact with Lake Victoria as a means of predicting S. mansoni infection were 76.6% (34 of 45) and 66.0% (464 of 703), respectively. Among 25 children who reported frequent contact with Lake Victoria, nine were infected (positive predictive value = 36%). However, the sensitivity of self-reported frequent contact with the lake was low (20%, 9 of 45).
|
The mean school prevalence for infection with any of the STHs was 63%. Hookworm was the most common helminth (42.5%) followed by A. lumbricoides (22.9%) and T. trichiura (17.9%). Hookworm infections were homogeneously distributed throughout Asembo but roundworm and whipworm infections were clustered near the northern edge of the surveillance area (Figure 3). Ascaris lumbricoides infections were negatively associated with S. mansoni infections among children living beyond 4 km from the shoreline (OR = 0.11, 95% CI = 0.03, 0.45), but no association was seen among children living within 4 km of the shore (OR = 1.09, 95% CI = 0.61, 1.96).
A total of 291 of 1,215 children (24.0%) were infected with more than one STH. The most common co-infections were A. lumbricoides and hookworm (9.9%) followed by T. trichiura and hookworm (9.0%). Thirty-five children (2.9%) were infected with all three helminths. Of the 202 children infected with S. mansoni, 123 (60.9%) were infected with an STH. Hookworm was the most common co-infection (45.0%) followed by T. trichiura (19.3%) and A. lumbricoides (10.9%).
|
DISCUSSION |
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
These results support previous studies that have evaluated distance to Lake Victoria as a means of predicting infection levels of S. mansoni. Lwambo and others14 reported school prevalences ranging from 0% to 65% in Tanzania. All of the schools that had a prevalence greater than 20% were located within 5 km of Lake Victoria and schools that had a prevalence greater than 50% were located within 2 km. Brooker and others15 also found a strong inverse association between distance to Lake Victoria and prevalence of S. mansoni. They suggested a 5-km cutoff point for correctly predicting schools of high prevalence ( 50%). However, distance was less useful at predicting schools with moderate levels of infection (20–50%). These results suggest that distance from the shore may be a useful indicator of prevalence, although the specific distance from the shore that would trigger mass treatment will probably vary with location.
Only three children from two villages were infected with urinary schistosomiasis. It is not clear whether these infections were acquired elsewhere or indicate a focal point of transmission within Asembo. These two villages were located 3 km and 5.5 km from the lakeshore. Along the shoreline of Lake Victoria in Tanzania, urinary schistosomiasis was found in increasing prevalence with increasing distance from the lake.14 Since our study area extended only to 13 km from the shoreline, we cannot speculate on the prevalence of urinary schistosomiasis in villages more distant from Lake Victoria.
The heterogeneous distribution of intestinal schistosomiasis infections has important implications for potential mass treatment programs. Within 1.5 km of the shoreline, the school prevalence ranged from 58% to 80%. Based on WHO guidelines,23 all persons living in these villages, regardless of age, should be treated for schistosomiasis. From 1.5 to 4 kilometers from the shoreline, the prevalence remained greater than 20% but less than 50%. Accordingly, selective treatment of school-aged children in these villages would be recommended. Beyond the 4 km mark, the prevalence was considerably lower and averaged 5.7%. Treatment in these villages, therefore, would be based on a clinical or laboratory diagnosis of infection.
If a school-based mass treatment program were undertaken for all schools located within 4 km of Lake Victoria, 11 schools would receive treatment of schistosomiasis, including all 8 schools with a prevalence greater than 20% (sensitivity = 100%). Of the 24 schools having a prevalence less than 20%, 21 schools would have been correctly categorized as receiving no treatment (specificity = 87.5%). Three schools would have been included in the treatment group even though the prevalence was less than 20%. At an individual level, we would effectively treat 437 (35%) of 1,246 school children in the study area, 158 (77%) of 204 children infected with intestinal schistosomiasis, and 70% of those children with heavy infections. Thus, by selecting schools based on proximity to Lake Victoria, we would provide praziquantel to only one-third of all children in the study area, yet effectively treat more than three-fourths of those infected with schistosomiasis.
Children infected with STHs were homogeneously distributed in the study area, although this was not true of the parasites individually. Infections with A. lumbricoides, and to a lesser extent T. trichiura, were clustered in the northern villages away from the lake. We are unable to explain this clustering because indicators such as soil conditions, altitude, and latrine coverage were similar throughout the study area. Nearly two-thirds of children were infected with one or more helminthes, indicating that according to WHO recommendations, a mass treatment of all school-aged children should be conducted throughout the Asembo area. A recent review of studies of schistosomiasis and STHs concluded that they were independently distributed and suggested that mass treatment programs for schistosomiasis and helminths would have to be tailored to the local conditions.24 Our results agree with this conclusion. Based on our findings, we recommend that all school-aged children in Asembo should be treated for STHs, and those living within a specified distance of a known source of transmission of schistosomiasis (in this case 4 km from Lake Victoria) should be treated with praziquantel.
In conclusion, distance to Lake Victoria was the most significant predictor of infection for S. mansoni among school children in Asembo, and a GIS proved to be a useful tool in suggesting the source of infection. We recommend the use of distance from known sources of transmission of S. mansoni as a useful means of identifying high-risk populations for mass treatment and other interventions. Questionnaires regarding water-related behaviors or time spent in the lake may be useful in identifying additional infected children who live beyond the cut-off point for mass treatment. However, the questions must be carefully structured to maximize both the sensitivity and specificity of the responses.
Received November 30, 2002. Accepted for publication May 27, 2003.
Acknowledgments: We thank the headmasters, teachers, and students in each of the schools that participated in this study. We also thank Alfred Okoth and Matunda Kennedy for their assistance in the field and laboratory, and Maurice Ombok for his assistance in the collection of GIS data. This paper is published with the permission of the Director of the Kenya Medical Research Institute.
Authors’ addresses: Thomas Handzel, Division of Emergency and Environmental Health Services, National Center for Environmental Health, Centers for Disease Control and Prevention, Mailstop F-48, 4770 Buford Highway NE, Atlanta, GA 30341, Telephone: 770-488-4466, Fax: 770-488-7829, E-mail: tnh7{at}cdc.gov. Diana Karanja and Julius Andove, Center for Vector Biology and Control Research, Kenya Medical Research Institute, PO Box 1578, Kisumu, Kenya. David G. Addiss, Allen W. Hightower, Daniel H. Rosen, Laurence Slutsker, and W. Evan Secor, Division of Parasitic Diseases, Centers for Disease Control and Prevention, Mailstop F-13, 4770 Buford Highway NE, Atlanta, GA 30341. Daniel G. Colley, Center for Tropical and Emerging Global Diseases, Department of Microbiology, 623 Biologic Sciences Building, University of Georgia, Athens GA 30602.
|
REFERENCES |
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  C. J. Lin, E. Katongole-Mbidde, T. Byekwaso, J. Orem, C. S. Rabkin, and S. M. Mbulaiteye Intestinal Parasites in Kaposi Sarcoma Patients in Uganda: Indication of Shared Risk Factors or Etiologic Association Am J Trop Med Hyg, March 1, 2008; 78(3): 409 - 412. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. R. DESAI, D. J. TERLOUW, A. M. KWENA, P. A. PHILLIPS-HOWARD, S. K. KARIUKI, K. A. WANNEMUEHLER, A. ODHACHA, W. A. HAWLEY, Y. P. SHI, B. L. NAHLEN, et al. FACTORS ASSOCIATED WITH HEMOGLOBIN CONCENTRATIONS IN PRE-SCHOOL CHILDREN IN WESTERN KENYA: CROSS-SECTIONAL STUDIES Am J Trop Med Hyg, January 1, 2005; 72(1): 47 - 59. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
