Childhood diarrhoea in Sri Lanka: a case-control study of the impact of improved water sources

Between January 1987 and March 1988 a case-control study of the impact of improved water sources on childhood diarrhoea was conducted in Kurunegala District, Sri Lanka. Two thousand four hundred and fifty eight cases of diarrhoea were recruited at five of the hospitals in the district. Another 4140 children presenting at the same hospitals with complaints other than diarrhoea were recruited as controls. Data from the five hospitals suggest that children in households drawing their drinking water from handpumps suffer 46% fewer episodes of diarrhoea than children in families using unprotected traditional sources (95% c. i. 29-59%), while children in families using protected traditional wells suffer 35% fewer episodes than children in families using unprotected traditional sources (95% c. i. 27-41%). There were, however, substantial differences between the different hospitals. Among children recruited at one of the hospitals, the reduction in diarrhoea rates associated with the use of improved sources was estimated to be 93% compared with an average of 18% for the other four hospitals. In common with other case-control studies conducted in Malawi and the Philippines, little evidence of confounding of the association between diarrhoea and water supply was observed. Our results suggest that, in Sri Lanka, the use of improved water supplies, including protected traditional wells, rather than unprotected traditional sources may lead to a substantial reduction in diarrhoea morbidity among children under five years of age.