Independent predictors of human milk receipt at neonatal intensive care unit discharge.

Objective: The aim of this study is to construct a predictive model for very low birth weight (VLBW) infants' receipt of mother's own milk within 24 hours before neonatal intensive care unit (NICU) discharge.

Study Design: Vermont Oxford Network (VON) clinical data were analyzed retrospectively for VLBW infants admitted between 2002 and 2012 at an inner city, level IV NICU with a well-established lactation program. Bivariate analyses compared infant characteristics between recipients and nonrecipients of human milk before 24 hours of NICU discharge. Independent predictors identified in the bivariate analyses (p ≤ 0.05), were eligible for inclusion into a multivariable logistic regression model.

Results: We observed a 60.4% human milk feeding rate at NICU discharge. Multiple independent maternal (black race and Hispanic ethnicity) and clinical factors (need for mechanical ventilation, patent ductus arteriosus, late-onset sepsis, or discharge to home on a cardiorespiratory monitor) positively predicted the outcome.

Conclusions: Our results were inconsistent with previous studies and suggest that a strong NICU lactation program in combination with a community-based peer counselor program may increase rates of human milk receipt among VLBW infants born to black/Hispanic mothers and those with more complicated neonatal courses.