Melson, Valerie

ABSTRACT Objective To compare maternal and neonatal adverse outcomes between women who are English proficient (EP) and those who have limited English proficiency (LEP). Design Retrospective cohort study. Setting Single US academic medical centre with interpreter services. Population All pregnant women with deliveries during the study period. Methods We retrospectively identified EP and LEP individuals (categorised by self‐report) who delivered at Mayo Clinic, Rochester, Minnesota, USA, from 2015 to 2019. Overall composite adverse outcomes, as well as maternal and neonatal composite outcomes, were measured with generalised linear models and logistic regression, reporting estimates and 95% CIs. Multivariable logistic regression analysis was used to assess associations between language proficiency and outcome measures while controlling for other clinical factors. Main Outcome Measures Composite overall, maternal and neonatal adverse outcomes. Results Among 10 427 EP and 900 LEP participants, the LEP group had a higher incidence of gestational diabetes, unplanned intensive care unit admission, uterine rupture, postpartum haemorrhage with transfusion, neonatal death and 5‐min Apgar score less than 7. The composite overall, maternal and neonatal adverse outcomes were more prevalent in the LEP group than the EP group in unadjusted analysis. Of these, the composite adverse outcome remained significant when the model was adjusted (odds ratio, 1.40; 95% CI, 1.03–1.90). Conclusions Pregnant women with LEP had increased incidence of adverse delivery outcomes, even in a setting with universal availability of medical interpreters.