Comparative Retention in Care for Key Populations: Differentiated Service Delivery Models in Kenyan Lake Victoria Fishing Communities, 2002

Fishing communities around Lake Victoria in Kenya experience a high HIV prevalence and population mobility, which challenges retention in care for key populations such as female sex workers and fisherfolk. Differentiated service delivery (DSD) models have been introduced to address this, but comparative evidence on their effectiveness is scarce. This study compared retention in HIV care among key populations enrolled in different DSD models within these fishing communities. A retrospective cohort analysis was performed using routine programme data. Participants were enrolled in one of three DSD models: facility-based adherence clubs, community-based peer-led groups, or fast-track drug refill. Retention was defined as being alive and in care 12 months after enrolment. Multivariable logistic regression identified factors associated with retention. Retention at 12 months was highest for participants in community-based peer-led groups (78%), compared to facility-based clubs (65%) and fast-track refill (71%). After adjusting for covariates, enrolment in a community-based model was associated with significantly increased odds of retention. Community-based, peer-led DSD models were associated with better retention in care for key populations in this high-mobility setting than more facility-centric approaches. HIV programme planners should prioritise and scale up community-led, peer-supported DSD models in fishing communities. Further operational research is needed to understand the specific components driving success and to adapt models for other mobile populations. HIV, retention in care, differentiated service delivery, key populations, fishing communities, Kenya This research provides comparative evidence from a high-priority setting to inform the scale-up of differentiated HIV services for mobile key populations in East Africa.