| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
 |
ABSTRACT |
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
INTRODUCTION |
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
The Safe Water System (SWS), a household-based water quality intervention developed by the Centers for Disease Control and Prevention (CDC) and the Pan American Health Organization, consists of water treatment using locally produced sodium hypochlorite solution and safe water storage in a narrow-mouth container (Figure 1
).5,6 The SWS reduces the risk of water-borne diarrheal diseases in developing country settings.7–9 However, because the causes of diarrhea and the proportion of diarrhea associated with water-borne pathogens may be different among persons with and without HIV, we evaluated the effectiveness of SWS among persons with HIV and their family members living in rural Uganda.
|
|
METHODS |
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Participants and enrollment.
Persons with HIV-1 infection who were clients of The AIDS Support Organization (TASO) in the rural Tororo district in Uganda and without access to chlorinated municipality water were consecutively enrolled in the study between January and March 2001 (Figure 2). After written, informed consent was provided, study staff visited participants’ homes to conduct a census and obtain consent from household members. Consent forms and questionnaires were translated into six local languages and back-translated into English. A household was defined as persons who shared a hearth and slept in the same house or cluster of houses for at least five days of the week for the preceding three months. A finger stick sample of blood was collected from household members on filter paper for HIV testing. All persons were encouraged to receive HIV test results and counseling at home or at the project clinic at Tororo Hospital. The HIV test result counseling was provided to study participants alone, or with partners. For participants 10–17 years of age, counseling included both the child and parent or legal guardian, and for those 0–9 years of age, only the parent or legal guardian was counseled. Consent for HIV testing or receiving test results were not requirements for enrollment, but only persons for whom HIV test results were available were included in analyses.
|
Data collection and main outcome measures. At baseline, participants were interviewed about household demographic and socioeconomic characteristics, and water, hygiene, and sanitary practices. All HIV-positive participants received an assessment by a physician and provided blood samples for viral load and CD4 cell count.
During weekly visits, study staff administered a questionnaire to all household members regarding diarrhea episodes, days with diarrhea, days of school or work lost, and hospitalization or death of a household member in the previous seven days. Seriously ill participants were encouraged to come to the study clinic and/or hospital and be treated free-of-charge. For cases of diarrhea, defined as 
3 loose or watery stools in 24 hours, a stool specimen was collected and aliquoted. Field workers inserted two swabs into stool samples, transferred them to a tube of Cary Blair transport media, then placed one portion of stool in a tube of formol saline, and another into a sterile container for storage. All samples were placed on ice packs in a cooler, transported to the project laboratory, and refrigerated. All Cary Blair specimens were cultured within 48 hours of collection or frozen for future testing. Persons unable to provide a sample immediately were asked to insert two rectal swabs that were immediately transferred into Cary Blair transport media, and were also given a specimen cup for a stool specimen that was retrieved later the same day. Diarrhea was treated with oral rehydration solution and, if indicated, antimicrobial and antimotility agents. Only one stool sample was collected for each episode of diarrhea.
During cotrimoxazole prophylaxis, questions on adherence, adverse effects, pill counts, and a re-supply of drugs were included. Immediately before prophylaxis and at the end of the study, HIV-positive participants provided additional blood samples for viral load and CD4 cell count testing. At the time of this study, antiretroviral therapy was not provided by TASO to its clients because of extremely limited availability in Uganda.
Laboratory testing. Stool samples in Cary Blair media were cultured for Salmonella, Shigella, Campylobacter, Vibrio cholerae, Pleisiomonas, and Aeromonas using standard procedures. To test for enterotoxigenic Escherichia coli and enteropathogenic E. coli, stool samples were cultured on Mac-Conkey agar. A sweep of colonies with appearance typical of E. coli was taken from each plate, stored in glycerol broth, and subsequently tested by polymerase chain reaction using DNA probes.15 Antimicrobial resistance was measured for bacterial stool pathogens using standard, disk-diffusion techniques.16 Evergreen Concentrate Kits (formalin-ethyl acetate, fecal parasite concentrator; Evergreen Scientific, Los Angeles, CA) were used to identify ova, cysts, and parasites from formalinized whole stool samples. Modified Zeihl Neelson stain was used to identify Cryptosporidium cysts. Frozen stool samples were thawed and tested for Rotavirus using the Murex Rotavirus enzyme-linked immunosorbent assay kit (Med-Ox Diagnostics, Inc., Ottawa, Ontario, Canada).
Plasma samples were screened for HIV infection using two enzyme-linked immunoassays (EIAs) in parallel (Recombigen HIV-1/HIV-2; Trinity Biotech, Dublin, Ireland and Murex HIV 120; Abbot Diagnostics, Chicago, IL). Specimens negative on both EIA screening tests were considered negative and specimens positive on both assays were considered positive. Specimens with discordant results were re-tested by Western Blot (LAV Blot, Bio-Rad Laboratories, Hercules, CA). HIV testing of dried blood spots consisted of a screening EIA (Vironostika HIV; BioMerieux, Durham, NC) and confirmation of reactive specimens by Western blot. The polymerase chain reaction was conducted on all HIV-reactive specimens from children less than 24 months of age. The CD4 cell counts and HIV viral loads were measured using standard procedures (FACScount; Becton-Dickinson, La Jolla, CA and Cobas Amplicor Monitor version 1.5; Roche, Nutley, NJ).
At baseline, a random sample of 20% of study households was selected for microbiologic testing of source and stored water. In October 2001 and October 2002, follow up microbiologic water quality testing was conducted on additional random samples of 20% of the households. Field workers collected water samples in sterile 500-mL plastic containers and transported them to the project laboratory in a cooler with ice packs. Samples were processed on the day of collection using the membrane filtration technique with E. coli as the indicator organism (Hach Co., Loveland, CO).17
Data analysis. Data were analyzed using SAS version 9 software (SAS Institute, Cary, NC). Multivariable Poisson regression models using a log link function were developed for assessing associations between SWS and diarrhea episodes, days with diarrhea, and days of work or school lost due to diarrhea, adjusting for age, sex, time of year in three-month intervals, cotrimoxazole use, water quality, presence of latrine in compound, soap at home, and household wealth, including World Health Organization clinical stage and CD4 cell count for persons with HIV. The CD4 count measured at the beginning of the study was used as a baseline for the time period before cotrimoxazole prophylaxis, and the CD4 count measured just before starting treatment with cotrimoxazole was used as a baseline for the cotrimoxazole period. The combined effect of SWS and cotrimoxazole prophylaxis was assessed by comparing the rates of diarrhea in the intervention arm during cotrimoxazole prophylaxis with the rates in the control arm before cotrimoxazole prophylaxis.
Generalized estimating equation methods with an exchangeable correlation structure and an offset for the number of days at risk were used to control for intra-household and intra-individual disease clustering. A Wald test was used to compare differences between groups. Multivariable results are presented unless otherwise specified. An intention-to-treat approach was used for analyses. Analysis of the effect of SWS on changes in CD4 cell count and viral load were calculated comparing the rate of change of CD4 cell count or viral load using analysis of covariance, adjusting for initial CD4 cell count or viral load for each period.
|
RESULTS |
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
). Seventy-four percent of the persons with HIV were female and median age was 34 years (interquartile range [IQR] = 28–40 years); 27 (5%) were less than five years old. At baseline, 27% of the persons with HIV had CD4 cell counts < 200 cells/mm3, 37% had CD4 cell counts of 200–500 cells/mm3, and 36% had CD4 cell counts > 500 cells/mm3. Of the 1,521 HIV-negative family members, 730 (49.6%) were female and median age was 10 years, (IQR = 6–15 years); 255 (17%) were less than five years old. The median follow-up time for persons with HIV was 547 days (IQR = 313–563 days) in the intervention group and 556 days (IQR = 469–567 days) in the comparison group (P = 0.059). There were no significant differences between intervention and comparison groups in baseline demographic characteristics, HIV status, CD4 cell count, or socioeconomic status.
Water, hygiene, and sanitation.
Most households used springs or boreholes as water sources (Table 1). More than 90% of households’ stored water at home using wide-mouth containers. Of all water samples collected at baseline, more stored household samples had detectable E. coli than source samples (64 [90%] versus 40 [56%]; P < 0.0001), and stored household-water samples had higher levels of E. coli than source water samples (median = 158 [range = 0–40,000] versus 14 [range = 0–19,800] colony-forming units per 100 mL; P < 0.0001). Most households did not treat water. The only effective water treatment practices reported were boiling for 21% of intervention and 24% of comparison households, and bleach for one household (0.5%) in each group. Sanitary conditions, water sources, water handling practices, hygienic practices, and baseline E. coli colony counts were similar in both groups, except the intervention group had slightly worse E. coli contamination of source and stored water at baseline (P = 0.08 and 0.09, respectively), more often had water available for hand washing (P = 0.06), and more frequently reported hand washing after defecation (P = 0.07), although none of these findings were statistically significant (Table 1).
|
200 cells/mm3 (IRR = 2.50, 95% CI = 1.96–3.21, P < 0.0001) and more days of work or school lost due to diarrhea (IRR = 3.16, 95% CI = 1.95–5.12, P < 0.0001).
Overall, SWS was associated with a 20% reduction in diarrhea episodes (P = 0.0469) and 26% fewer days with diarrhea (P = 0.055) (Table 2). Among persons with HIV, SWS was associated with a 25% reduction in diarrhea episodes (P = 0.015) and 33% fewer days with diarrhea (P = 0.021). The SWS was not associated with significant reductions in days of work or school lost due to diarrhea among all participants (IRR = 0.79, 95% CI = 0.54–1.15, P 0.223) or persons with HIV (IRR = 0.76, 95% CI = 0.50–1.15, P = 0.197).
|
Among HIV-negative household members, interactions were found between age group and the association of SWS with diarrhea. The SWS was associated with reductions in diarrhea episodes only among persons > 59 years old (P = 0.056) and those 3–12 years old (P = 0.025) and reductions in days with diarrhea only among persons > 59 years old (P = 0.028) (Table 2). There were no reductions in the days of school or work lost due to diarrhea among family members (IRR = 0.89, 95% CI = 0.51–1.52, P = 0.662). The SWS had no effect on hospitalizations and clinic visits among persons with HIV or HIV-negative family members.
Stool pathogens. During the study, 936 (80%) episodes of diarrhea had associated stool specimens collected. Participants in the intervention group reported fewer stools with visible blood or mucous than the comparison group (28% versus 39%; P < 0.0001). The results were similar for persons with HIV (32% versus 42%; P = 0.0065) and HIV-negative family members (22% versus 36%; P = 0.0022). The proportion of different pathogens recovered from stool specimens collected from persons with HIV varied: hookworms (33%), Strongyloides stercoralis (15%), enterotoxigenic E. coli (14%), Aeromonas species (8%), enteropathogenic E. coli (7%), Shigella species (6%), Cryptosporidia parvum (6%), Salmonella speicies (3%), and Campylobacter species (4%). For persons with HIV, there was no difference in pathogens between the intervention group and comparison groups. Among HIV-negative family members, stool samples from participants in the intervention group had lower rates than the comparison group of hookworms (27% versus 40%; P = 0.0138) and Shigella species (1% versus 5%; P = 0.0292).
Environmental and hygienic factors associated with diarrhea. The first two quarters of the year, when compared with the last two quarters, were associated with more diarrhea episodes (January through March; relative risk [RR] = 1.41, 95% CI = 1.13–1.76, P = 0.002 and April through June; RR = 1.24, 95% CI = 1.02–1.51, P = 0.032). Among persons with HIV, the presence of a latrine in a compound, compared with those without a latrine, was associated with fewer episodes of diarrhea, (IRR = 0.69, 95% CI = 0.53–0.91, P = 0.009), fewer days with diarrhea, (IRR = 0.63, 95% CI = 0.40–1.00, P = 0.048), and fewer days of work or school lost due to diarrhea (IRR = 0.63, 95% CI = 0.41–0.97, P = 0.038). Among persons with HIV, the presence of soap in the house was also associated with fewer days with diarrhea, (IRR = 0.58, 95% CI = 0.35–0.97, P = 0.038), and fewer days of work or school lost due to diarrhea (IRR = 0.56, 95% CI = 0.34–0.93, P = 0.024), but not episodes of diarrhea (IRR = 0.79, 95% CI = 0.60–1.07, P = 0.134). The presence of a latrine or soap were not associated with significant reductions in diarrhea episodes, days with diarrhea, or days of work or school lost due diarrhea among HIV-negative family members.
Association between diarrhea, HIV viral load, and SWS. Each diarrhea episode was associated with a 0.12 log10 copies/mL per year increase in viral load (95% CI = 0.01–0.23, P = 0.037). The HIV viral load increased by 0.40 log10 per person-year for persons with HIV using SWS compared with 0.71 log10 per person-year for those not using SWS (adjusted mean pairwise difference = –0.14 log10 per person-year, 95% CI, = –0.55 to 0.27, P = 0.510).
Water testing.
During repeat water testing in October 2001, only stored household water samples from the comparison arm households had E. coli contamination. In October 2002, more samples from households of the comparison arm had E. coli contamination than the intervention arm. However, the median E. coli counts of the contaminated water samples evaluated in October 2002 did not differ by arm (P = 0.308) (Table 3).
|
|
DISCUSSION |
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Among HIV-negative family members, the SWS was effective only among certain age groups, i.e., children 3–12 years old and adults > 59 years old, but not in other age groups. These findings are consistent with previous studies that have shown similar reductions in these age groups.7–9 It may be that the effect of SWS was restricted to those persons most likely to be at home during the day and able to access the intervention (children and adults > 59 years old). The SWS may have been effective among adults with HIV between 13 and 59 years of age because they were more likely to be at home due to illness and a decreased ability to work. This added benefit for some family members improves the cost-effectiveness of the intervention. Among children with HIV less than five years of age, the SWS was associated with a non-significant 30% reduction in diarrhea episodes. This was likely due to limited power because only 27 children with HIV participated in the study.
In this cohort, source water quality was relatively good and improved over the course of the study. The beneficial effect of the SWS may be greater in populations with more fecal contamination of drinking water. In addition, although water stored in the home is at risk of contamination,18 this risk may have been reduced in the study by good hygienic practices and a high degree of access to soap and toilet facilities, all of which lower the burden of pathogens in the household environment. Hand washing and access to latrines have been associated with reduced risk of diarrhea.19,20 The association of soap and latrines with reduced risk and severity of diarrhea in persons with HIV support these findings. Weekly home visits may have influenced adherence to the SWS and improved hygiene practices. Although the effect of health education on sanitation and hygiene practices would likely be similar across both study arms, the effectiveness of the SWS during large-scale programmatic implementation might be less than what was seen in this study because of lower adherence rates in a less intensive intervention.
Diarrhea is common in persons with HIV and in this study was associated with increased HIV viral load. Although this has not been previously examined, other studies have shown associations between opportunistic infections, viral load, and HIV disease progression.21–25 Interventions such as SWS that prevent diarrhea could potentially slow down the progression of HIV disease, although this study did not have sufficient power to address this question.
Point-of-use water quality interventions, such as the SWS, are the most cost-effective water interventions currently available.26 The cost per family for locally produced sodium hypochlorite solution is less than $0.01 per day, and the initial cost of buying the water vessel is approximately $3. This excludes costs to families of transport to obtain a vessel and chlorine or to the healthcare system for distribution, which were provided by the study infrastructure.
The results of this study support the statement by Lee Jongwook, the Director-General of the World Health Organization, that "successful treatment [of HIV] depends on safe water . . . . Antiretroviral treatment prolongs lives, but winning the war against the pandemic demands a combination of medicine, food, and clean water . . . ."27 Simple, inexpensive interventions such as the SWS and cotrimoxazole prophylaxis should be considered as components of a basic care package provided to all persons with HIV in resource-poor settings. The SWS might be especially useful in settings where diarrhea is common and drinking water quality is poor.
Received April 27, 2005. Accepted for publication July 11, 2005.
Acknowledgments: We thank the Tororo Hospital administrative and clinical staff; the volunteers, staff, and clients of TASO; the U.S. Embassy in Kampala; Global AIDS Program headquarters, Atlanta, GA; the staff of CDC-Uganda, including the informatics, clinical, laboratory, and administrative units of CDC-Tororo for their participation in this study. The staff of CDC was involved in the study design, data collection, analysis, and writing of the report. John R. Lule was the main author of the paper. John R. Lule, Jonathan Mermin, and Robert Quick were the principal investigators who wrote the protocol, supervised the study, guided data analysis, had full access to all the data, and take responsibility for the integrity of the data and accuracy of data analysis. Raymond Ransom supervised data entry and cleaning. John Paul Ekwaru and Samuel Malamba analyzed data and conducted statistical analyses. Robert Downing, Peter Hughes, and Aminah Kigozi supervised and conducted laboratory testing. Damalie Nakanjako and Winnie Wafula provided clinical oversight for the study. Rebecca Bunnell assisted in the design of the study, and she and Frank Kaharuza provided guidance in conducting the study and interpreting results. Alex Coutinho helped in designing and conducting the study, and ensured the protocol was applicable to TASO.
Financial support: This study was supported by CDC.
Disclosure: None of the authors had any conflicts of interest.
* Address correspondence to John R. Lule, Centers for Disease Control and Prevention-Uganda, Entebbe, Uganda. E-mail: nzl4{at}cdcuganda.org 
Authors’ addresses: John R. Lule, Jonathan Mermin, John Paul Ekwaru, Samuel Malamba, Robert Downing, Raymond Ransom, Damalie Nakanjako, Winnie Wafula, Rebecca Bunnell, Frank Kaharuza, and Aminah Kigozi, Centers for Disease Control and Prevention-Uganda, Entebbe, Uganda and Global AIDS Program, National Center for HIV, STD and TB Prevention, Centers for Disease Control and Prevention, Atlanta, GA 30333. Peter Hughes, Uganda Virus Research Institute, Entebbe, Uganda and Medical Research Council Programme on AIDS in Uganda, Entebbe, Uganda. Alex Coutinho, The AIDS Support Organization, Entebbe, Uganda. Robert Quick, Foodborne and Diarrheal Diseases Branch, Division of Bacterial and Mycotic Diseases, National Center for Infectious Diseases, Centers for Disease Control and Prevention, Atlanta, GA 30333.
|
REFERENCES |
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  M. L. BECKER, C. R. COHEN, M. CHEANG, R. G. WASHINGTON, J. F. BLANCHARD, and S. MOSES DIARRHEAL DISEASE AMONG HIV-INFECTED ADULTS IN KARNATAKA, INDIA: EVALUATION OF RISK FACTORS AND ETIOLOGY Am J Trop Med Hyg, April 1, 2007; 76(4): 718 - 722. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. MIGELE, S. OMBEKI, M. AYALO, M. BIGGERSTAFF, and R. QUICK DIARRHEA PREVENTION IN A KENYAN SCHOOL THROUGH THE USE OF A SIMPLE SAFE WATER AND HYGIENE INTERVENTION Am J Trop Med Hyg, February 1, 2007; 76(2): 351 - 353. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. F. ARNOLD and J. M. COLFORD JR TREATING WATER WITH CHLORINE AT POINT-OF-USE TO IMPROVE WATER QUALITY AND REDUCE CHILD DIARRHEA IN DEVELOPING COUNTRIES: A SYSTEMATIC REVIEW AND META-ANALYSIS Am J Trop Med Hyg, February 1, 2007; 76(2): 354 - 364. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. K. LAUFER and C. V. PLOWE Cotrimoxazole prophylaxis and malaria in Africa: have the important questions been answered? Am J Trop Med Hyg, September 1, 2006; 75(3): 373 - 374. [Full Text] [PDF]
|
|
|
|
|
|
R. K. SHRESTHA, E. MARSEILLE, J. G. KAHN, J. R. LULE, C. PITTER, J. M. BLANDFORD, R. BUNNELL, A. COUTINHO, F. KIZITO, R. QUICK, et al. COST-EFFECTIVENESS OF HOME-BASED CHLORINATION AND SAFE WATER STORAGE IN REDUCING DIARRHEA AMONG HIV-AFFECTED HOUSEHOLDS IN RURAL UGANDA. Am J Trop Med Hyg, May 1, 2006; 74(5): 884 - 890. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
