Women's family planning status in Ethiopia: A geo-additive modeling approach

Abstract

Background A complex interplay of socioeconomic, cultural, and geographic factors influenced women's family planning status in Ethiopia, resulting in significant zonal disparities. Traditional statistical models often fail to capture adequately the nonlinear and spatial components underlying these choices. Objective The study aimed to predict and assess the spatial, linear, and nonlinear effects of factors associated with women's family planning status in Ethiopia zones. Methods The study implemented a multi-stage cluster design and a geo-additive model. We drew secondary data from 7484 women of reproductive age, collected by the Performance Monitoring and Accountability (PMA2020) project. Results The geo-additive approach performed better than linear and additive models. Contraceptive use and unmet needs varied across zones, with spatial structure explaining 93% and 68% of the variation, respectively, confirming substantial spatial autocorrelation. Contraceptive use was more likely among those with higher education, higher socioeconomic status, fewer children, more health center visits, and a lower desire for more children. Unmet needs were associated with desire for children, sex avoidance, and the use of private healthcare facilities. Continuous variables have substantial and nonlinear effect. Conclusions The study's findings suggest that spatial and nonlinear variables affect women's family planning status. The spatial patterns closely aligned with the distribution of key predictors, such as education, wealth, parity, health center access, and fertility preferences. Consequently, reproductive health policies and programs in Ethiopia should be designed with attention to individual-level determinants and spatial disparities to ensure more equitable and effective outcomes on family planning.