367 - Documented Breastfeeding Support, Not Breastfeeding Intention, is a Primary Driver of Breastfeeding Exclusivity at Hospital Discharge

Publication Number: 3354.367

Tiffany Gallagher, University of Minnesota, New Ulm, MN, United States; Ellen W. Demerath, University of Minnesota School of Public Health, Minneapolis, MN, United States; Wendy S. Looman, University of Minnesota School of Nursing, Minneapolis, MN, United States; Carrie Neerland, University of Minnesota School of Nursing, Minneapolis, MN, United States; Robin R. Austin, University of Minnesota School of Nursing, Minneapolis, MN, United States; Michelle A. Mathiason, University of Minnesota, Minneapolis, MN, United States; Katy B. Kozhimannil, University of Minnesota School of Public Health, MINNEAPOLIS, MN, United States; Anne Chevalier. McKechnie, University of Minnesota, Minneapolis, MN, United States

Background: Guidelines published by the American Academy of Pediatrics advise at least one documented breastfeeding assessment for postpartum patients every 8-12 hours during the hospital stay. There are currently no systematic methods to determine whether this guideline is met and sufficient to achieve exclusive breastfeeding (EBF) for all lactating patients.

Objective: The purpose of this study was to systematically analyze electronic health records (EHR) to quantify documented hospital-based breastfeeding care and relate it to the outcome of EBF at hospital discharge within a socioeconomically and geographically diverse sample.

Design/Methods: This secondary analysis included EHR data from nine Minnesota hospitals in a single hospital system, obtained from a clinical data repository. After cleaning the data and applying exclusion criteria, our sample included 22,857 parents intending to provide breastmilk to their healthy, full-term infants born between 2018 and 2022. We created variables for documented support (number of documented support interactions / expected number of support interactions = support ratio) and breastfeeding intensity (number of breastmilk feedings / total number of feedings), as well as patient demographic and clinical factors. Feeding intention was defined as "breastmilk" for patients intending to provide breastmilk in any form and "both" for those intending to provide breastmilk and formula. The expected amount of care was defined as three documented breastfeeding support interactions in 24 hours. Our analysis included descriptive statistics and multivariable logistic regression with restricted cubic splines to model the relationship between documented breastfeeding support and EBF.

Results: The overall breastfeeding initiation rate was 98.3%, with 71.0% of dyads EBF. The mean breastfeeding intensity was 0.84 (SD 0.24) and mean support ratio was 2.32 (SD 0.98). Patients intending to provide breastmilk had a higher mean breastfeeding intensity and more documented support; patients who EBF had the most support documented. A positive dose-response relationship was identified between the amount of documented support and probability of EBF, independent of patient demographic and clinical characteristics and feeding intention.

Conclusion(s): Having more documented breastfeeding support is associated with increased probability of EBF at discharge for patients of all demographics and clinical characteristics, and is a stronger predictor of EBF than feeding intention. Differences in early postpartum lactation support provided to patients may contribute to observed disparities in breastfeeding outcomes.

Overall Feeding, Demographic, and Birth Characteristics (n = 22,857)


Overall Feeding, Demographic, and Birth Characteristics (n = 22,857) (continued)


Probability of Exclusively Breastfeeding by Support Ratio
Probability of exclusive breastfeeding by support ratio (calculated as the number of documented support interactions divided by the expected number of support interactions) using multivariable logistic regression with restricted cubic splines (5). Models: Unadjusted-no adjustment; intention-adjusted for feeding intention (ref = breastmilk); demographic-adjusted for urban/rural hospital (ref = urban), race/ethnicity (ref = White), age (ref = 30 to 34 years old), interpreter needed (ref = no), marital status (ref = married or partnered), coverage type (ref = private); clinical-adjusted for body mass index (ref = <30 kg/m2), current smoking (ref = no), parity (ref = not first baby), delivery method (ref = vaginal), gestational age (ref = 39-40 weeks), time to first breastfeeding (ref = Less than 1 hour); demographic and clinical-adjusted for demographic and clinical characteristics above; demographic, clinical, and intention-adjusted for all variables above