Objective This study aimed to determine if formula selection, low lactose versus standard term formula, has an effect on outcomes with a comparison to breastfed infants. Study Design Retrospective cohort study of neonates ≥35 weeks gestation born with Neonatal Abstinence Syndrome (NAS) was conducted from July 2014 to November 2016. Primary outcomes included length of pharmacologic treatment (LOT), and length of stay (LOS), and weight change per day comparing term standard and low lactose formula majority feeding infants with secondary outcomes comparing breast fed majority feeding infants. Results After investigating feeding methods for 249 NAS infants, a direct comparison of formula groups showed no differences in LOS (3, 95% confidence interval [CI]: −1.1 to 7 days), LOT (3.9, 95% CI: −0.4 to 8.1 days), or weight change per day (−2.4, 95% CI: −11.7 to 6.9 g/day). Breastfeeding improved LOT by 6.9 (95% CI: 3.4–10.5) and 10.8 days (95% CI: 5.9–15.6) and LOS by 7.4 (95% CI: 4.1–10.7) and 10.3 (95% CI: 5.8–14.9) days all reaching significance, in comparison to term and low lactose formula groups, respectively. Weight change per day was greater in the breast versus formula feeding groups when compared individually. Conclusion We detected no benefit to low lactose formula in NAS infants. Breastfeeding is associated with clinical reduction in LOS and LOT but is associated with increased weight loss. Key Points

The aim of the study is to determine length of stay and length of treatment in infants with neonatal abstinence syndrome (NAS) in the neonatal intensive care unit (NICU) compared to those in the pediatric floor. Methods: Retrospective cohort of infants !34 weeks gestation admitted with diagnosis of NAS at a single regional perinatal referral center from July 2014 to October 2015. A standardized NAS protocol for both the NICU and pediatric floor, which included guidelines for the initiation of oral morphine, escalation, and weaning, was followed. Initial location of treatment, NICU or pediatric floor, was determined by physiological stability following birth. Statistical analysis included 1-way analysis of variance and chi-square. Multivariable analysis was performed using generalized linear models to account for confounding. Results: The study included 235 infants, 80 (34%) were cared for in the NICU. Infants in the NICU had a longer length of stay (27.1 AE 19.1 vs 14.2 AE 10.2 days, P < 0.01), and length of pharmacological treatment (18.0 AE 19.9 vs 9.0 AE 10.2 days, P < 0.01) compared to those on the pediatric floor, respectively. Forty-seven infants were transferred from the NICU to the pediatric floor for the remainder of their hospital stay with a mean time on the pediatric floor of 17.4 AE 14.5 days. After controlling for confounding, admission to the NICU was associated with an increased length of treatment of 12.6 days (95% confidence interval 8.3-16.8) and length of stay of 12.3 days (95% confidence interval 7.9-16.6). Conclusions: In our population, admission to the pediatric floor compared to the NICU was associated with a shorter length of stay, and a shorter length of pharmacological treatment. Our data suggest that caring for infants with NAS outside of the NICU setting has the potential to improve short-term outcomes and reduce associated costs.