Abstract
ObjectiveThis study aims to determine the effect of infant–mother separation following a short-stay (≤72 hours) admission to a Level 5 neonatal unit versus no admission on infant-feeding outcomes at hospital discharge.DesignRetrospective cohort study.SettingAn Australian Level 5 neonatal unit within a tertiary referral hospital.ParticipantsMothers and their infants born between 1 January 2018 and 31 December 2020 had a short-stay admission to the neonatal unit or no admission. All participants met admission criteria to the postnatal ward and were discharged home at ≤72 hours (n=12 540). Postnatal ward admission criteria included ≥36 weeks’ gestation and birth weight ≥2.2 kg.Main outcome measuresInfant feeding at discharge from hospital. Multivariate logistic regression analysis was conducted, adjusting for confounders associated with known breastfeeding issues. These included age, ethnicity, parity, obesity, socioeconomic score, hypertensive disorders of pregnancy, diabetes, infant gestation and birthweight centile, caesarean section birth, postpartum haemorrhage and skin-to-skin contact.ResultsOf the 12 540 live births meeting inclusion criteria, 1000 (8%) infants were admitted to the neonatal unit. The primary reasons for admission were suspicion of sepsis (24%), maternal diabetes (19%) and jaundice (16%). We found a reduction in full breast feeding at hospital discharge in cases of a short admission to the neonatal unit compared with no admission (aOR 0.40; 95% CI 0.34 to 0.47; p<0.001). We identified that women of different ethnicities had differing levels of risk for formula supplementation at hospital discharge. The ethnic grouping least likely to be fully breast feeding at discharge was Southeast Asian women (aOR 0.47; 95% CI 0.39 to 0.57; p<0.001).ConclusionsIdentifying mother–infant dyads at risk of non-exclusive breast feeding at hospital discharge will help target resources for practice improvement.