The role of community health workers and local leaders in reducing attrition among participant in the AIDS indicator survey and HIV incidence in a national cohort study in Rwanda

Mwumvaneza Mutagoma, Dieudonné Sebuhoro, Jean Pierre Nyemazi, Edward J. Mills, Jamie I. Forrest, Eric Remera, Augustin Murindabigwi, Mouhamed Semakula, Sabin Nsanzimana

Retention of participants in longitudinal prospective surveys can challenging for population health researchers. Community health workers (CHWs) may help reduce attrition. We used data came from a longitudinal prospective household-based survey targeting women and men in Rwanda, collected between June 2013 and December 2014. The sample was drawn from a population that included all residents of all 30 districts, 416 sectors, and 14,837 villages in Rwanda. The outcome measure was time to loss-to-follow-up. Follow up visits occurred at three, six and nine, and 12 months. A Cox proportional hazards model was constructed to identify factors independently associated with time to loss-to-follow-up. Overall, 14,222 respondents consented to be interviewed at baseline. At the end of 12 months of follow up, 13,728 were revisited and consented to participate at 12 months of follow up. The overall attrition rate was 8.0%. A majority of those lost (54.3%) were less than 25 years of age, male (55.1%), not living in union (67.3%), had no education level or had primary education level (71.4%), or were in the highest wealth index (54.2%). Compared to illiterate, secondary education was negatively associated with attrition. The Rwanda AIDS indicator and HIV incidence survey recorded a very high retention of participants after 12 months. CHWs and local leaders played a major role to reduce attrition rate and identifying factors associated with loss-to-follow-up can help CHWs strengthen the quality of longitudinal survey data.