Contraceptive initiation after delivery among women with and without opioid use disorders: A retrospective cohort study in a statewide Medicaid population, 2005–2016

Abstract

BackgroundThis study examined contraceptive initiation patterns in the 12 months following childbirth among women with opioid use disorder (OUD), women with non-opioid substance use disorders (SUDs), and women without SUDs. MethodsWe conducted a retrospective cohort study using claims data from South Carolina Medicaid-enrolled women aged 15–44 who had singleton live birth between January 2005 and December 2016. Study outcomes were initiation of most or moderately effective (MME) contraceptive methods. Using multivariable and propensity score-weighted logistic regression, we analyzed the relationship between OUD and contraceptive initiation within 12 months after delivery. ResultsWe identified 71,283 live birth deliveries during the study period. In multivariable analysis, women with non-opioid SUDs and women without SUDs compared to women with OUD were more likely to initiate a MME method vs a least effective method or no method by 3 months (non-opioid SUDs: odds ratio [OR] = 1.32, 95 % confidence interval [CI] = 1.14−1.52; no SUDs: OR = 1.55, 95 % CI = 1.36−1.77) and 12 months (non-opioid SUD: OR = 1.23, 95 % CI = 1.06−1.42; no SUD: OR = 1.46, 95 % CI = 1.27−1.66) after delivery. With regards to the timing of initiation, women with non-opioid SUDs and women without SUDs were more likely than women with OUD to initiate a MME method vs a least effective method or no method after the immediate postpartum period through 3 months following delivery (non-opioid SUDs: OR = 1.41, 95 % CI = 1.18−1.68; no SUDs: OR = 1.87, 95 % CI = 1.59−2.21). We detected the similar patterns in analyses that used propensity score weighting. ConclusionOUD was associated with decreased likelihood of initiating a MME contraceptive method within 12 months after delivery.