Identifying social effects in a community-based IMCI project in San Luis, Honduras

We evaluated a community-based Integrated Management of Childhood Illnesses (IMCI) project carried out in the Municipality of San Luis, Honduras during 2004-2006. The project relied heavily on participatory, community-based strategies with hygiene the main focus. The Municipal Government provided the twenty-eight target villages with cement, pipes and technical assistance for latrine construction and water system improvements. Local Red Cross fieldworkers periodically visited each village, educating parents and providing monitoring and feedback. Baseline and endline surveys were done in each village using standard WHO cluster sampling techniques. We found increases in the mean probabilities of several behaviors, particularly handwashing. We then fit two-level logit models with handwashing the outcome and found significant village-level random effects in both waves. We ranked the residuals, plotted the ranks and identified the four villages whose ranks had most increased and the four whose ranks had most declined. We returned to these eight villages and conducted 38 key informant and focus group interviews. We used structured questionnaires to probe both perceptions of both formal and informal organizations. Respondents were asked to identify successful community projects and recent collective action experiences. All interviews were recorded, transcribed and analyzed using Atlas-ti. Four researchers independently coded the transcripts and ranked the eight villages from most to least likely to have generated social effects. The subjective and quantitative residual ranks were in close agreement. Social effects, we conclude, played an important role in this phase of San Luis' health transition. But they did not materialize everywhere, mainly for idiosyncratic reasons.