Meeting Contraceptive Preferences Among Low-Income Postpartum Texans: A Counterfactual Analysis of Pregnancy Trajectories

Abstract

Many people at risk of an undesired pregnancy are not using their preferred contraceptive method. On its own, discordant use is an important indicator of reproductive autonomy. It may also affect reproductive outcomes, although little research has explored the consequences of unsatisfied contraceptive preferences. Using prospective data for the two years following delivery for a cohort of low-income postpartum women in Texas, many of whom would have preferred to be using a more effective method than the one they were using, we ask: How would the pregnancy trajectories of those not using their preferred contraceptive have differed had they been able to access their preferred method? Taking an inverse probability of treatment weighting approach, we show that using a preferred method was associated with half the likelihood of conceiving a pregnancy likely to result in a birth within 21 months postpartum (adjusted hazard ratio 0.43; 95 percent confidence interval 0.32, 0.57). Our findings highlight the consequences of failing to provide people with their preferred method and reveal how, in a context where people face substantial barriers to reproductive healthcare, discordant use increases the risk that people will have a baby that they did not desire at that time or, in some cases, ever.