Modelling the incidence of childhood diarrhea

Diarrheal disease is a primary cause of childhood mortality and morbidity in developing countries. Significant reduction depends on the identification of high-risk households which can be targetted for proven-effective interventions, such as oral rehydration therapy. This paper describes the development of a household risk assessment model in the context of an on-going diarrheal disease control in Grenada. Contingency and logit analyses of household survey data are used to identify easily observed markers of high-risk households. A three-variable logit model gives accurate predictions of self-reported diarrhea, showing that the approach can be used to develop an effective and efficient risk assessment tool.