Pregnancy coercion, correlates, and associated modern contraceptive use within a nationally representative sample of Ethiopian women.

Partner-perpetrated pregnancy coercion inhibits women's reproductive autonomy. However, few studies have quantified pregnancy coercion and its effects on women's health within low- and middle-income countries. Among a national sample of Ethiopian women, this study aimed to: (1) assess the prevalence of past-year pregnancy coercion and explore regional differences; (2) identify correlates; (3) examine the relationship between pregnancy coercion and modern contraceptive use. Analyses utilise cross-sectional data from Performance Monitoring for Action (PMA)-Ethiopia, a nationally representative sample of females aged 15-49 conducted from October to November 2019. Past-year pregnancy coercion was assessed via five items and analysed dichotomously and categorically for severity. Among women in need of contraception, bivariate and multivariable logistic regression examined associations between variables of interest, per aim, accounting for sampling weights and clustering by enumeration area. Approximately 20% of Ethiopian women reported past-year pregnancy coercion (11.4% less severe; 8.6% more severe), ranging from 16% in Benishangul-Gumuz to 35% in Dire Dawa. Increasing parity was associated with decreased odds of pregnancy coercion. Among women in need of contraception, experience of pregnancy coercion was associated with a 32% decrease in odds of modern contraceptive use (aOR = 0.68; 95% CI: 0.53-0.89); when disaggregated by severity, odds decreased for most severe pregnancy coercion (aOR = 0.59; 95% CI = 0.41-0.83). Results indicate that partner-perpetrated pregnancy coercion is prevalent across diverse regions of Ethiopia, and most severe forms could interrupt recent gains in contraceptive coverage and progress to sexual and reproductive health and rights. Providers must be aware of potential contraceptive interference and address coercive influences during contraceptive counselling.