Postpartum Contraception Usage Among Somali Women in Olmsted County, Minnesota

OBJECTIVE: To compare postpartum contraception use between Somali and non-Somali women. METHODS: A retrospective cohort study was performed using the Rochester Epidemiology Project. All Somali women aged 18 and older with live singleton births in Olmsted County, Minnesota, in 2009-2015 (n=317) were included, and a group of age-matched non-Somali women (n=317) were identified. Postpartum contraception was defined as the use of any method within 12 months after the first delivery within the study period. Rates of contraception use and types of contraception used were compared between groups using χ analysis. Among Somali women, an a priori list of factors was evaluated for associations with postpartum contraception use by including these factors in a multivariable logistic regression model. RESULTS: After the index birth, 33 Somali women did not present for follow-up, compared with 12 non-Somali women (10.4% vs 3.8%, difference 6.6%, 95% CI 2.3-10.9%). Somali women were less likely to use postpartum contraception than non-Somali women (69.4% vs 92.8%, odds ratio [OR] 0.18, 95% CI 0.11-0.29). Among those using postpartum contraception, both groups had comparable use of long-acting reversible contraception (LARC) (19.9% vs 23.7%, difference -3.8%, 95% CI -11.8% to 4.0%) and non-LARC hormonal contraception (39.3% vs 42.4%, difference -3.1%, 95% CI -12.7% to 6.0%). However, Somali women were more likely to use less reliable methods (40.3% vs 20.8%, difference 19.5%, 95% CI 11.0-28.5%). Among Somali women, contraception use was less likely with older maternal age (OR 0.12 for oldest vs youngest quartile, 95% CI 0.04-0.37) and more likely with prepregnancy contraception use (OR 15.46, 95% CI 5.18-46.18). CONCLUSION: Somali women were less likely to use postpartum contraception. Similar practices and beliefs may also be present in other immigrant populations. Recognition of sociocultural differences in immigrant populations and potential health care provider biases may improve patient-provider relationships and counseling practices.