• View in gallery

    Emergency department visits classified as diarrhea and acute malnutrition at Nyangabgwe Hospital, Francistown, Botswana, November 15, 2005–April 7, 2006 in children less than five years of age (Creek T and others, unpublished data).

  • 1

    UNDP, 2005. Human Development Index. Geneva: World Health Organization.

  • 2

    Boonstra E, Lindbaek M, Fidzani B, Bruusgaard D, 2001. Cattle eradication and malnutrition in under five’s: a natural experiment in Botswana. Public Health Nutr 4 :877–882.

    • Search Google Scholar
    • Export Citation
  • 3

    Ismail S, 1991. Nutritional surveillance: experience from developing countries. Proc Nutr Soc 50 :673–679.

  • 4

    Grivetti LE, 1978. Nutritional success in a semi-arid land: examination of Tswana agro-pastoralists of the eastern Kalahari, Botswana. Am J Clin Nutr 31 :1204–1220.

    • Search Google Scholar
    • Export Citation
  • 5

    UNDP, 2003. Human Development Report 2003. New York: Oxford University Press.

  • 6

    Lemeshow S, Robinson D, 1985. Surveys to measure programme coverage and impact: a review of the methodology used by the expanded programme on immunization. World Health Stat Q 38 :65–75.

    • Search Google Scholar
    • Export Citation
  • 7

    Hoshaw-Woodard S, 2001. Description and Comparison of the Methods of Cluster Sampling and Lot Quality Assurance Sampling to Assess Immunization Coverage. Geneva: World Health Organization.

  • 8

    World Health Organization, 2005. The Treatment of Diarrhoea: A Manual for Physicians and Other Senior Health Workers. Fourth edition. Geneva: World Health Organization.

  • 9

    World Health Organization, 1995. Physical Status: The Use and Interpretation of Anthropometry. Geneva: World Health Organization.

  • 10

    American Public Health Association, American Water Works Association, Water Pollution Control Federation, 1980. Standard Methods for the Examination of Water and Wastewater. 15th edition. Washington, DC: American Public Health Association.

  • 11

    Olson B, Clark D, Milner B, Stewart M, Wolfe R, 1991. Total coliform detection in drinking water: comparison of membrane filtration with Colilert and Coliquick. Appl Environ Microbiol 57 :1535–1539.

    • Search Google Scholar
    • Export Citation
  • 12

    Alam NHF, Rahaman MM, 1989. Reporting errors in one-week diarrhoea recall surveys: experience from a prospective study in rural Bangladesh. Int J Epidemiol 18 :697–700.

    • Search Google Scholar
    • Export Citation
  • 13

    Ramakrishnan RVT, Koya PK, Kamaraj P, 1999. Influence of recall period on estimates of diarrhoea morbidity in infants in rural Tamilnadu. Indian J Public Health 43 :136–139.

    • Search Google Scholar
    • Export Citation
  • 14

    Genser B, Strina A, Teles CA, Prado MS, Barreto ML, 2006. Risk factors for childhood diarrhea incidence: dynamic analysis of a longitudinal study. Epidemiology 17 :658–667.

    • Search Google Scholar
    • Export Citation
  • 15

    Molbak K, Aaby P, Hojlyng N, da Silva AP, 1994. Risk factors for Cryptosporidium diarrhea in early childhood: a case-control study from Guinea-Bissau, West Africa. Am J Epidemiol 139 :734–740.

    • Search Google Scholar
    • Export Citation
  • 16

    Rose JB, 1997. Environmental ecology of Cryptosporidium and public health implications. Annu Rev Public Health 18 :135–161.

  • 17

    Thomson MC, Mason SJ, Phindela T, Connor SJ, 2005. Use of rainfall and sea surface temperature monitoring for malaria early warning in Botswana. Am J Trop Med Hyg 73 :214–221.

    • Search Google Scholar
    • Export Citation
  • 18

    Schmitt CV, 2006. Africa Weather Hazards Assessment February 2–8, 2006. Famine Early Warning System Network.

  • 19

    Leshomo K, 2006. Diarrhea Epidemic Leads to Threat of Severe Malnutrition in Botswana. New Yok: UNICEF.

  • 20

    Southern African Development Community Food Agricultural and Natural Resources Directorate, 2006. Agromet-Update: 2005/2006 Agricultural Season. Food Security Early Warning System.

  • 21

    Dop MC, 2002. Breastfeeding in Africa: will positive trends be challenged by the AIDS epidemic? [in French]. Sante 12 :64–72.

  • 22

    UNICEF, 2000. Botswana Multiple Indicator Survey. New York: The United Nations Children’s Fund.

  • 23

    Gracey M, 1996. Diarrhea and malnutrition: a challenge for pediatricians. J Pediatr Gastroenterol Nutr 22 :6–16.

  • 24

    The Sphere Project, 2004. Humanitarian Charter and Minimum Standards in Disaster Response. Geneva: Oxfam Publishing.

 

 

 

 

Population-Based Study of a Widespread Outbreak of Diarrhea Associated with Increased Mortality and Malnutrition in Botswana, January–March, 2006

View More View Less
  • 1 Centers for Disease Control and Prevention, Atlanta, Georgia; BOTUSA Project, Gaborone, Botswana; Botswana Ministry of Health, Gaborone, Botswana

In early 2006, coinciding with heavy rains, Botswana health facilities reported increases in diarrhea, mortality, and acute malnutrition among young children. Data on diarrhea, human immunodeficiency virus, feeding, mortality, and water/sanitation were collected by random cluster survey. Anthropometric data were measured and drinking water samples were tested. Of 537 surveyed children less than five years of age, one-third experienced ≥ 1 episode of diarrhea. Prevalence of acute malnutrition was 7.9%, and the mortality rate for children less than five years of age was 2.6/10,000/day during the outbreak. Significant risk factors for diarrhea included an age less than two years; breastfeeding was protective. Diarrhea lasting for more than 14 days and failure to thrive were risk factors for acute malnutrition. Prevalence of acute malnutrition was higher than previously documented and the mortality rate in children less than five years of age during the outbreak was above the international threshold for emergency action with an estimated 547 excess deaths. This survey highlights the need for safe infant feeding and effective treatment of malnutrition and diarrhea in young children.

BACKGROUND

In early 2006, a four-fold increase in the number of diarrhea cases among young children was reported to the Ministry of Health through routine surveillance in Botswana, a country in southern Africa with a population of approximately 1.76 million.1 At the same time, cases of severe acute malnutrition, a rare condition in Botswana,24 were reported by health personnel providing care to patients with diarrhea. The number of reported deaths caused by diarrhea among young children was more than 25 times higher than in the previous two years. On the basis of Ministry of Health surveillance data, the area most heavily impacted by the outbreak appeared to be in the eastern and most densely populated part of the country.

Consultations for diarrhea at the emergency department at Nyangabgwe Hospital in Francistown, the country’s second largest referral hospital at the center of this outbreak, began to increase early in January 2006, peaked in late February, and returned to normal in April (Figure 1) (Creek T and others, unpublished data). Additionally, cases of severe acute malnutrition were first seen in the hospital in late January, peaked in March, and were still occurring when our investigation ended in mid-April (Arvelo W and others, unpublished data). The total number of visits for diarrhea during this period was 1,276, and 83 cases of severe acute malnutrition were identified in the emergency department among children less than five years of age. Almost all the children admitted for acute malnutrition had experienced diarrhea earlier in 2006 (Creek T and others, unpublished data).

Laboratory testing of in-patient stool specimens from Nyangabgwe Hospital performed at The Centers for Disease Control and Prevention in Atlanta showed a multipathogenic etiology of diarrhea. Cryptosporidium parvum, Cryptosporidium hominis, and enteropathogenic Escherichia coli were the most common organisms identified. Norovirus and rotavirus were relatively uncommon (Arvelo W and others, unpublished data).

This outbreak occurred in a country with one of the world’s highest prevalence of human immunodeficiency virus (HIV): approximately 33% of pregnant women are infected with HIV.5 A national program to prevent mother-to-child transmission (PMTCT) of HIV recommends that HIV-positive mothers feed infants with formula instead of breast milk. The PMTCT program distributes free formula to HIV-exposed infants less than 12 months of age.

We conducted a population-based, cross-sectional survey to estimate the burden of illness at the community level and to quantify age-specific attack rates of diarrhea, mortality, and prevalence of acute malnutrition among children less than five years of age during this outbreak, and to better understand risk factors associated with these outcomes. This survey was conducted as an outbreak response measure and the findings were immediately used for action.

METHODS

A two-stage, 30-cluster survey was conducted in three health districts of Botswana: Francistown, Tutume (also known as Central), and North-East during the first two weeks of April 2006. The districts were selected based on Botswana surveillance data and on logistical feasibility. The combined population of these districts was estimated to be 256,000, with approximately 32,000 children less than five years of age (2001 census data). We defined a cluster as a village or section of town listed in the 2001 census and a household as a group of persons sharing food from the same cooking pot. In the first stage, 30 clusters were randomly selected from census lists. The selection was proportional to population size, ensuring that all households in the sampling universe had an equal chance of being selected. For the purposes of this survey, we treated the population of these three districts as one entity. Of the 30 clusters randomly selected for the survey, 20 were rural villages and 10 were urban or periurban communities. The population of the clusters ranged from 300 to 18,000 persons.

In the second stage, households with children less than five years of age were randomly selected in each cluster using the Expanded Program on Immunization method 6,7 until 23 children were identified in each cluster. All children less than five years of age in a selected household were included in the survey. Households without a child less than five years of age were not included unless household members reported a death of a child less than five years of age during the previous three months. If household members in a selected household were absent on two separate visits, the household was skipped and not replaced.

On the basis of 2001 census data, we estimated an average of one child less than five years of age per household in Botswana. Assuming an overestimated prevalence of acute malnutrition at 20%, 367 children less than five years of age would be needed to estimate prevalence with a precision of ± 5%, taking into account an anticipated design effect of 1.5. To accommodate for possible loss of data caused by refusal to participate and absent families, and to use available resources adequately, we aimed to survey 23 children in each of 30 clusters, resulting in a projected sample size of 690 children less than five years of age.

Each survey team consisted of one interviewer, one person tasked with taking anthropometric measurements and collecting water samples, and a driver. Once a household was selected, the survey team identified the female head of household, obtained informed consent, and administered a questionnaire in the local language. We collected information about the history of diarrhea among children less than five years of age between January 1, 2006 and the interview date in April 2006, nutrition practices, drinking water sources and storage practices, and HIV status of child and mother. The HIV status was self-reported or based on data from the child health card. The interviewee was also asked to report any deaths among children less than five years of age occurring between January 1, 2006 and the date of interview. Diarrhea was defined as three or more loose or watery stools in 24 hours. Data on number of episodes was not collected, only presence or absence of diarrhea during the recall period was collected. Long-lasting diarrhea was defined as diarrhea lasting for more than 14 days.8

Available child health cards were reviewed to assess child growth history. In Botswana, children less than five years of age are weighed monthly in health centers, and their weight is plotted against their age in a health card. We defined failure to thrive as having no reported weight gain during the three most recent health center visits.

Children between 6 months and 5 years of age, whose parents consented, were weighed with digital Seca (Hanover, MD) scales (with precision of ± 0.1 kg) and height was measured using Shorr (Olney, MD) height boards (with precision of ± 0.5 cm). The child was classified as having nutritional edema (kwashiorkor) if pitting edema was present on both feet at the time of anthropometric measurements.

Acute malnutrition was categorized according to World Health Organization (WHO) guidelines.9 Moderate acute malnutrition was defined as a weight-for-height Z-score more than 2 SD below the median (1978 WHO reference standards). Severe acute malnutrition (SAM) was defined as a weight-for-height Z-score more than 3 SD below the median, or the presence of nutritional edema (kwashiorkor). Global acute malnutrition (GAM), also referred to as acute malnutrition, included children who were either severely or moderately malnourished.

The mortality rate of children less than five years of age was calculated as the number of deaths among children less than five years of age per 10,000 person-days. Cause of death was reported by the primary caregiver and divided into basic categories: diarrhea and/or vomiting, breathing problems or cough, HIV, insufficient food, injury, other, or unknown.

Drinking water samples were collected from approximately every fourth surveyed household. Samples were collected from the water source and from the vessel used for storage of drinking water at the household. 10 Tap fixtures were sterilized and flushed to clear the service line before sampling. Water samples were collected in Whirl-paks (Nasco International, Inc., Fort Atkinson, WI) or in sterilized bottles containing sodium thiosulfate. Water samples were transported to the laboratory in a cooler with ice for duplicate testing within six hours of collection. Water was tested for total coliform bacteria and for E. coli by using either membrane filtration (Millipore Corporation, Billerica, MA) or Colilert ® (IDEXX, Westbrooke, MA) systems. 11

Water contamination was categorized by concentration of E. coli per 100-mL sample. Categories included no contamination (0 E. coli per 100 mL); low (0–9 E. coli per 100 mL); moderate (10–99 E. coli per 100 mL); and high (> 100 E. coli per 100 mL). Arithmetic mean was used for duplicate testing.

Survey data were analyzed with Epi Info version 3.3.2 Centers for Disease Control and Prevention), SAS version 9.1 (SAS Inc., Cary, NC), and Sudaan version 9 (RTI International, Research Triangle Park, NC), accounting for cluster design of the study. Univariate associations are reported by use of crude risk ratios (RRs) with 95% confidence intervals (CIs) or by use of adjusted risk ratios when significant confounders were found. Multivariate associations based on logistic regression models are reported by use of adjusted ORs with 95% CIs. All risk factors included in the univariate analysis were included in the initial multivariate regression model. The final model included only factors that were found to be significant at P < 0.05.

The survey protocol was reviewed and approved by the Centers for Disease Control and Prevention and determined to be non-research; it was approved by the Botswana Health Research Development Committee.

RESULTS

The study included 537 children less than five years of age in 339 households, of which 271 (50.5%) were male and 259 (48.2%) were less than two years of age. The household response rate was 79%. Non-responders included those households that did not consent to participate in the study (< 10% of non-responders) and those households where residents were absent on two attempts during the data collection phase (> 90% of non-responders).

During the first three months of 2006, 170 (32.0%) children were reported to have had at least one episode of diarrhea (Table 1). Children 1–2 years of age had the highest attack rate, and children in the oldest age group (4–5 years) had the lowest attack rate. The median duration of diarrhea was 7 days, and long-lasting diarrhea was reported among 23.2% of diarrhea cases. A high proportion of children with diarrhea received treatment (152 of 168, 90.5%). Of these children, 144 (94.8%) consulted a physician or a nurse and 4 (2.6%) visited a traditional healer. Most caregivers reported offering less food and drink to their children during diarrhea episodes.

The prevalence of GAM among 407 children with complete anthropometric data between six months and five years of age was 7.9%, the prevalence of SAM was 2.7%; and the prevalence of kwashiorkor was 0.2% (Table 1). There were no significant differences in prevalence between sexes or between age groups.

On the basis of data from 414 (81.7%) children with available health cards, an estimated 16.3% (95% CI = 7.9–24.8%) were failing to thrive at the time of the survey. The proportion of children less than two years of age with failure to thrive was 18.1%; among children 2–5 years of age it was 15.0%. This difference was not statistically significant. A total of 13 children less than five years of age was reported to have died during the first three months of 2006, providing an estimated mortality rate of children less than five years of age of 2.6/10,000/day (95% CI = 0.6–4.5). Parents of 11 (84.6%) of the deceased children reported diarrhea as the cause of death.

Among the households sampled, 205 (65.7%) mothers had been tested and knew their HIV status. Of those mothers with known HIV status, 70 (34.1%) were HIV positive and 135 (65.9%) were HIV negative. Among children of HIV positive mothers, 48 (41.9%) had been tested for HIV. Among these children, 4 (8.3%) were infected with HIV. One additional child was found to be infected with HIV according to the child’s health card. However, the child’s mother was not aware of her own HIV status.

At the time of the survey, 73 (30.8%) children less than two years of age and 41 (67.2%) of children less than one year of age were being breastfed. Breastfeeding varied significantly by mother’s HIV status. Breastfeeding at the time of survey was reported among 51 (47.4%) children less than two years of age born to HIV-negative mothers compared with 1 (1.7%) born to HIV-positive mothers. There were 384 (74.0%) children who consumed any type of porridge in the 24 hours preceding the survey, which reflected porridge consumption by 8.2% of children less than 6 months of age and 82.3% of children 6 months to 5 years of age.

Approximately one-third of 339 surveyed households had a water tap in the house, one-third had a standpipe in their compound, and one-third used water from a shared standpipe. A small proportion of households (< 1%) drank water from a river or an open well. Samples of source water tested from approximately every fourth household (84) showed contamination with E. coli in 22 (26.2%). In 58 (69%) households, the drinking water was stored in various containers, mostly in buckets without lids. Stored water was contaminated in 22 (38%) of households. However, levels of contamination in stored water were not found to be significantly different from those of source water.

In our sample of 339 households, 240 (71.4%) used a pit latrine as their main toilet, 70 (20.8%) had flush toilets, and 26 (7.7%) did not have any toilet facilities and used open space and bushes to urinate and defecate. In 67 (21.7%) households, the interviewers observed that soap was available at the place where household members washed their hands.

Significant risk factors for diarrhea included child’s age less than 24 months and a child being HIV positive. Current breastfeeding was found to be significantly protective against diarrhea (Table 2).

Consumption of any type of porridge, the primary weaning food of Botswana, during the 24 hours preceding the survey was significantly associated with history of diarrhea among children less than two years of age (Table 2) No other factors, including other foods consumed, were associated with diarrheal disease.

Having a flush toilet was considered to be a possible proxy for household socioeconomic status. However, we did not observe any association either between owning a flush toilet (or a pit latrine) or having soap available for hand-washing and self-report of any diarrhea during a three-month recall period. We did not observe a significant association between contamination of drinking water with E. coli, measured at the time of survey, and history of diarrhea. Prevalence of diarrhea was not associated with type of water source (in house versus yard or public tap) or with water storage practice.

When analyzing risk factors for development of acute malnutrition, long-lasting diarrhea and failure to thrive during the prior three months were significant risk factors. HIV status, feeding practices, owning a flush toilet, or formal education of the caretaker were not significantly associated with acute malnutrition (Table 2).

DISCUSSION

The results of this cross-sectional survey in three heavily affected areas of Botswana suggest that the outbreak of diarrhea affected a large proportion of the children less than five years of age and led to increased levels of acute malnutrition and mortality among children.

Approximately one-third of children less than five years of age were reported to have experienced at least one episode of diarrheal illness between January and March 2006. Although research suggests that 24–72-hour recall periods for pediatric diarrhea are optimal and that longer recall periods are likely to be unreliable, we were unable to capture diarrhea incidence in this manner and instead used a three-month recall period. 12,13 The overall attack rate is not unusual for this recall period. 14 However, the dramatic increase in hospitalizations for diarrhea (Figure 1), the high proportion of long-lasting diarrheas, and increased levels of acute malnutrition and pediatric deaths provide evidence that this was an outbreak with unusually severe consequences.

Relevant results from related studies conducted prior to this survey contribute to a picture of the possible etiology behind the outbreak. Laboratory findings from hospitalized children suggested probable sewage contamination of the environment during heavy rains in late 2005 and early 2006. Cryptosporidium species are typically indicative of exposure to sewage contamination of water or the environment. 15,16 The rain was unusually heavy when compared with available meteorologic data from previous years, and it caused flooding and overflowing of latrines (Arvelo W and others, unpublished data). 1720 It is likely that the environmental contamination occurred early in the outbreak and that poor personal hygiene and sanitary conditions placed food and drinking water stored within homes at risk of contamination and permitted person-to-person transmission. A high proportion of non-breastfed children created a pool of persons who were more likely to be severely ill or to die because they lacked protective maternal antibodies. 21 Additionally, most HIV-positive women receiving free infant formula from the PMTCT program reported having been short of formula at some point because of supply or distribution problems.

In light of our finding that porridge is an important weaning food and that it was associated with diarrhea, additional investigation is needed to describe whether the risk is caused by the properties of the different types of porridge, their preparation, or whether this finding is an indicator of introduction of solid foods in general.

Although none of the environmental factors were associated with diarrhea in this analysis, they are well-known risk factors for diarrhea. Efforts to improve water quality, sanitation, and personal hygiene would likely reduce diarrhea incidence in these communities. Most parents reported offering less food and drink than usual to their children while they had diarrhea, which suggests poor awareness of WHO protocols for oral rehydration and nutrition during episodes of diarrhea.

The prevalence of global acute malnutrition in this survey was higher than that reported in the 2000 Botswana National Multiple Indicator Cluster Survey (4.9%). 22 However, because of the limited size of the survey, this difference did not reach statistical significance. Our finding of a significant association between long-lasting diarrhea and acute malnutrition supported a hypothesis, generated while observing hospitalized patients, that repeated and prolonged diarrhea led to acute malnutrition. 23 The observed association between failure to thrive, documented at clinic visits, and acute malnutrition, measured during the survey, points to a missed opportunity to intervene when growth faltering is observed by health workers.

We estimated the excess mortality rate in children less than five years of age during January–March 2006 to be 1.9/10,000/day by subtracting a baseline mortality rate of 0.7/10,000/day (converted child mortality rate from the 2005 annual report of The United Nations Children’s Fund) from our estimate of 2.6/10,000/day. By extrapolating to the total population of children less than five years of age in the three health districts, we estimated that approximately 547 (95% CI = 115–950) children less than five years of age died as a result of this outbreak, a figure greater than the number of deaths reported to the Ministry of Health for the same period of time for the entire country. Furthermore, the mortality rate in children less than five years of age likely exceeded the Sphere Project threshold for emergency action in sub-Saharan Africa (2.3/10,000/day). 24

This study had several limitations. First, according to hospital data, the outbreak of diarrhea began in mid-December 2005, peaked in February 2006, and had subsided by the end of March. Thus, our survey conducted in April had the potential to be influenced by a recall bias. This relatively long recall period, coupled with the publicity surrounding the outbreak, may also have influenced parents to incorrectly identify the occurrence, duration, and date of diarrhea in their children. The study design could not account for repeated episodes of diarrhea in children and parents may not have paid sufficient attention to loose stools in their children. However, the fact that more than 90% of children with reported diarrhea received treatment supports the validity of the parental report. It is also possible that parents did not correctly remember what they fed their children. Additionally, 21% of households in our sampling frame were considered non-respondents because of households being found empty on two or more occasions or consent for interview being denied. Finally, water contamination was measured at the time of survey while environmental contamination during the diarrhea outbreak would have occurred several months prior. This fact may explain the lack of observed association between contamination of drinking water and recalled diarrhea morbidity.

Having a high proportion of non-breastfed young children in a setting where poor environmental and hygiene conditions exist is a potential hazard that can contribute to and exacerbate diarrhea outbreaks. The role of infant feeding recommendations for HIV-positive mothers regarding cessation and non-initiation of breastfeeding should be explored. Furthermore, suboptimal feeding during diarrheal disease may have resulted in increased rates of acute malnutrition and high mortality among young children. Failure to respond to growth faltering and to implement proper treatment measures for long-lasting and severe diarrhea and acute malnutrition likely contributed to the high levels of mortality after this outbreak.

Table 1

Diarrhea and acute malnutrition among children less than five years of age in three health districts in Botswana, 2006*

Table 1
Table 2

Risk factors for diarrhea and global acute malnutrition, Botswana, 2006*

Table 2
Figure 1.
Figure 1.

Emergency department visits classified as diarrhea and acute malnutrition at Nyangabgwe Hospital, Francistown, Botswana, November 15, 2005–April 7, 2006 in children less than five years of age (Creek T and others, unpublished data).

Citation: The American Journal of Tropical Medicine and Hygiene Am J Trop Med Hyg 80, 5; 10.4269/ajtmh.2009.80.812

*

Address correspondence to Ondrej Mach, Global Immunization Division, Centers for Disease Control and Prevention, Mailstop E05, 1600 Clifton Road, Atlanta, GA 30333. E-mail: omach@cdc.gov

Authors’ addresses: Ondrej Mach, Lydia Lu, Tracy Creek, Anna Bowen, Wences Arvelo, Muireann Brennan, and Thomas Handzel, Global Immunization Division, Centers for Disease Control and Prevention, Mailstop E05, 1600 Clifton Road, Atlanta, GA 30333, E-mail: omach@cdc.gov. Molly Smit, BOTUSA Project, Gaborone, Botswana. Japhter Masunge, Botswana Ministry of Health, Gaborone, Botswana.

Acknowledgments: We thank L. Zaks, A. Kim, T. Finkbeiner, B. Woodruff, E. Mintz, N. Shaffer, C. Bern, M. Davis and others in Botswana and Atlanta for their participation in this study.

Financial support: This study was supported by the Centers for Disease Control and Prevention. However, there was no special grant or fund made available specifically for this investigation; operational funds for outbreak response were used.

REFERENCES

  • 1

    UNDP, 2005. Human Development Index. Geneva: World Health Organization.

  • 2

    Boonstra E, Lindbaek M, Fidzani B, Bruusgaard D, 2001. Cattle eradication and malnutrition in under five’s: a natural experiment in Botswana. Public Health Nutr 4 :877–882.

    • Search Google Scholar
    • Export Citation
  • 3

    Ismail S, 1991. Nutritional surveillance: experience from developing countries. Proc Nutr Soc 50 :673–679.

  • 4

    Grivetti LE, 1978. Nutritional success in a semi-arid land: examination of Tswana agro-pastoralists of the eastern Kalahari, Botswana. Am J Clin Nutr 31 :1204–1220.

    • Search Google Scholar
    • Export Citation
  • 5

    UNDP, 2003. Human Development Report 2003. New York: Oxford University Press.

  • 6

    Lemeshow S, Robinson D, 1985. Surveys to measure programme coverage and impact: a review of the methodology used by the expanded programme on immunization. World Health Stat Q 38 :65–75.

    • Search Google Scholar
    • Export Citation
  • 7

    Hoshaw-Woodard S, 2001. Description and Comparison of the Methods of Cluster Sampling and Lot Quality Assurance Sampling to Assess Immunization Coverage. Geneva: World Health Organization.

  • 8

    World Health Organization, 2005. The Treatment of Diarrhoea: A Manual for Physicians and Other Senior Health Workers. Fourth edition. Geneva: World Health Organization.

  • 9

    World Health Organization, 1995. Physical Status: The Use and Interpretation of Anthropometry. Geneva: World Health Organization.

  • 10

    American Public Health Association, American Water Works Association, Water Pollution Control Federation, 1980. Standard Methods for the Examination of Water and Wastewater. 15th edition. Washington, DC: American Public Health Association.

  • 11

    Olson B, Clark D, Milner B, Stewart M, Wolfe R, 1991. Total coliform detection in drinking water: comparison of membrane filtration with Colilert and Coliquick. Appl Environ Microbiol 57 :1535–1539.

    • Search Google Scholar
    • Export Citation
  • 12

    Alam NHF, Rahaman MM, 1989. Reporting errors in one-week diarrhoea recall surveys: experience from a prospective study in rural Bangladesh. Int J Epidemiol 18 :697–700.

    • Search Google Scholar
    • Export Citation
  • 13

    Ramakrishnan RVT, Koya PK, Kamaraj P, 1999. Influence of recall period on estimates of diarrhoea morbidity in infants in rural Tamilnadu. Indian J Public Health 43 :136–139.

    • Search Google Scholar
    • Export Citation
  • 14

    Genser B, Strina A, Teles CA, Prado MS, Barreto ML, 2006. Risk factors for childhood diarrhea incidence: dynamic analysis of a longitudinal study. Epidemiology 17 :658–667.

    • Search Google Scholar
    • Export Citation
  • 15

    Molbak K, Aaby P, Hojlyng N, da Silva AP, 1994. Risk factors for Cryptosporidium diarrhea in early childhood: a case-control study from Guinea-Bissau, West Africa. Am J Epidemiol 139 :734–740.

    • Search Google Scholar
    • Export Citation
  • 16

    Rose JB, 1997. Environmental ecology of Cryptosporidium and public health implications. Annu Rev Public Health 18 :135–161.

  • 17

    Thomson MC, Mason SJ, Phindela T, Connor SJ, 2005. Use of rainfall and sea surface temperature monitoring for malaria early warning in Botswana. Am J Trop Med Hyg 73 :214–221.

    • Search Google Scholar
    • Export Citation
  • 18

    Schmitt CV, 2006. Africa Weather Hazards Assessment February 2–8, 2006. Famine Early Warning System Network.

  • 19

    Leshomo K, 2006. Diarrhea Epidemic Leads to Threat of Severe Malnutrition in Botswana. New Yok: UNICEF.

  • 20

    Southern African Development Community Food Agricultural and Natural Resources Directorate, 2006. Agromet-Update: 2005/2006 Agricultural Season. Food Security Early Warning System.

  • 21

    Dop MC, 2002. Breastfeeding in Africa: will positive trends be challenged by the AIDS epidemic? [in French]. Sante 12 :64–72.

  • 22

    UNICEF, 2000. Botswana Multiple Indicator Survey. New York: The United Nations Children’s Fund.

  • 23

    Gracey M, 1996. Diarrhea and malnutrition: a challenge for pediatricians. J Pediatr Gastroenterol Nutr 22 :6–16.

  • 24

    The Sphere Project, 2004. Humanitarian Charter and Minimum Standards in Disaster Response. Geneva: Oxfam Publishing.

Save