110 likes | 251 Views
Household treatment of drinking water and child diarrhea: Estimating the effect of each one on the other in the Philippines. Joseph J. Capuno & Carlos Antonio R. Tan, Jr. University of the Philippines. Diarrhea remains a child health problem.
E N D
Household treatment of drinking water and child diarrhea: Estimating the effect of each one on the other in the Philippines Joseph J. Capuno & Carlos Antonio R. Tan, Jr. University of the Philippines
Diarrhea remains a child health problem • In 2008, diarrheal diseases accounted for 16% of the 3.7 million deaths among under-5 children in African and Southeast Asian countries alone (WHO 2011). • In the Philippines, under-5 children accounted for 61% of the 7,505 and 50% of 5,435 cases of acute bloody diarrhea in 2005 and 2007, respectively. In 2003 and 2004, 0.5-0.6 infants out of every 1,000 live births died due to diarrheal diseases (DOH 2000, 2005 and 2007).
Water and sanitation interventions to prevent diarrhea • Proposed interventions to break the fecal-oral route of diarrhea-causing microorganisms • Widen access to improved water sources and sanitation facilities • Improve the quality of drinking water at the point of use • Promote desired hygiene practices • The most cost-effective could be country-specific (e.g., Gunther and Fink 2010, Fewtrell et al. 2005, Waddington et al. 2009, Kremer and Zwame 2007).
Sample of households with children below 5 years old, 1998, 2003 and 2008 Note: Sample of under-5 children limited to dejure members of households. Source: National Demographic and Health Survey (various rounds). Authors’ calculations.
Policy research question: Do households treat water before or after the onset of diarrhea? Diarrhea in under-5 children Household water treatment
Framework Z Diarrhea in under-5 children Proportion of under-5 children with diarrhea Household water treatment H
Estimation models • Treatment effects model • Instrumental variable probit model
Naïve model: OLS estimates Other controls: urban, wealth quintiles, region dummies, year dummies. ***Significant at the 1% level. **Significant at the 5% level.
Treatment effects model a plus other control variables. bTreated water= f(improved water source, mother’age, time to water source, if mother is employed, if mother finished high school, if father finished high school, if household head is male, if urban, share of under-5 children in the household, dummies for wealth quintiles, dummies for regions, dummies for years). ***Significant at the 1% level. **Significant at the 5% level.
IV-probit model Aplus other controls. bInstrumented variable = proportion of under-5 children with diarrhea; instruments = improved water source, mother’age , time to water source, if mother is employed, if mother finished high school, if father finished high school, if household head is male, if urban, share of under-5 children in the household, dummies for wealth quintiles, dummies for regions, dummies for years. ***Significant at the 1% level.
Summary • The observed household-level incidence diarrhea in under-5 children is both a reason for and the result of a decision to treat drinking water in Philippine households. • Ignoring this two-way relationship could lead to an onerous conclusion that treated water aggravates the incidence of child diarrhea. • The quality of drinking water from all sources must be regularly monitored at the point of use.