Abstract
While community-based interventions are a proven high-impact strategy to increase contraceptive uptake in low-income countries, their capacity to support women's contraceptive choices (including continued use, switching and discontinuation) in the long run remains insufficiently discussed. This cohort study follows 883 women 3 and 6 months after they received a modern method during community campaigns organized in Kinshasa (D.R. Congo), to analyze their contraceptive trajectories and the factors associated with ever discontinuing contraceptive use in the first 6 months following a campaign. In the community-based distribution (CBD) model currently institutionalized in DRC, campaign clients are not provided with additional doses or support, besides baseline counseling, to (dis-)continue using the method they received, but must rely on Family Planning resources within the existing local health system. Almost a third (28.9%) of all women discontinued modern contraception during the study period, with much higher discontinuation rates for short-acting methods (38.7% for pills and up to 68.9% for DMPA-SC). Variables previously associated with high discontinuation (marital status, fertility intentions and side-effects) led to higher odds of "ever discontinuing". However, these variables became non-significant when controlling for resupply issues. Women's self-reported reasons for discontinuation confirmed the multivariate regression results. Detailed sub-analysis of resupply issues for pills, injectables and Cyclebeads pointed to the role of cost, unreliable campaign schedules and weak integration of community-based strategies into the formal health system. Extremely low rates of implants removal suggest similar access to FP services issues. The study highlights the need to identify CBD strategies best suited to support women's choices and preferences towards successful contraceptive trajectories in fragile health systems.