The selection and validation of indicators for monitoring progress towards self-sustainment in community-directed, ivermectin-treatment programmes for onchocerciasis control in Uganda.

A retrospective analysis was made of quantitative data on coverage obtained over 4 years of annual ivermectin treatment of the eligible populations (approximately 56,000 individuals) of 71 communities with endemic onchocerciasis in the Kabale, Kisoro and Rukungiri districts of Uganda. The objective was to formulate methods for defining sustainability in community-directed, ivermectin-treatment programmes (CDITP). Three dependent-variable scales of programme sustainability (PS), PS1, PS2, and PS3, were tested for statistical significance by analysis of variance. The inhabitants of a random sample of 230 households drawn from 23 communities [each containing one community leader and one community-based distributor (CBD)] were then invited to answer a questionnaire covering seven independent variables. These variables were analysed in regression and correlation models, with the PS scales as dependent variables. In the regression model, only one variable, selection of CBD by community members (P = 0.038), which scored 100% on the scale of programme-indicator sensitivity, passed as a useful indicator for predicting the sustainability and monitoring the sustainment of CDITP at the community level. The same variable was also selected in the correlation model (P = 0.028). Although two other variables--involvement of CBD in other primary-health-care activities (P = 0.0594) and provision of incentives for the CBD (P = 0.0558)--showed weak negative associations with sustainability in the correlation model, they did not exhibit a linear relationship with it and cannot therefore be used as valid indicators for predicting sustainability or monitoring sustainment.