Quantifying the association between doula care and maternal and neonatal outcomes

The United States suffers from an increasing rate of severe maternal morbidity, paired with a wide disparity in maternal health by race. Doulas are posited to be a useful resource to increase positive outcomes and to decrease this disparity. This study aimed to evaluate the association between doula care and a broad range of maternal and neonatal outcomes in various subpopulations. This was a retrospective cohort study of deliveries that were recorded from January 2021 to December 2022 at a single institution where they received prenatal care. The exposure was receipt of doula care prenatally and at delivery. We evaluated both the maternal (cesarean delivery, cesarean delivery of nulliparous, term, singleton, vertex infant, vaginal birth after cesarean, gestational hypertension, preeclampsia, postpartum emergency department visit, readmission, and attendance of postpartum office visit) and neonatal (neonatal intensive care unit admission, unexpected complications in term newborns, breastfeeding, preterm delivery, and intrauterine growth restriction) outcomes. Because our institution previously employed targeted outreach by offering doula services to patients at highest risk, we used multiple methods to generate an appropriate comparison population. We conducted a multivariate logistic regression and conditional regressions using propensity scores to model the likelihood of doula care to generate adjusted risk differences associated with doula care. Analyses were repeated in populations stratified by race (White vs Black) and then by payor status (public vs commercial). Our cohort included 17,831 deliveries; 486 of those received doula care and 17,345 did not. Patients who received doula care were more likely to self-report Black race, be publicly insured, and to live in a more disadvantaged neighborhood. Regardless of the analytical approach, for every 100 patients who received doula care, there were 15 to 34 more vaginal births after cesarean (adjusted risk difference, 15.6; 95% confidence interval, 3.8–27.4; adjusted risk difference, 34.2; 95% confidence interval, 0.046–68.0) and 5 to 6 more patients who attended a postpartum office visit (adjusted risk difference, 5.4; 95% confidence interval, 1.4–9.5; adjusted risk difference, 6.8; 95% confidence interval, 3.7–9.9) when compared with those who did not receive doula services. Infants born to these patients were 20% more like likely to be exclusively breastfed (adjusted risk ratio, 1.22; 95% confidence i