Send to

Choose Destination
Paediatr Perinat Epidemiol. 2019 Jan;33(1):59-69. doi: 10.1111/ppe.12536.

Unit policies and breast milk feeding at discharge of very preterm infants: The EPIPAGE-2 cohort study.

Author information

French National Institute of Health and Medical Research Unit 1153, Obstetrical, Perinatal, and Pediatric Epidemiology Team, Epidemiology and Biostatistics Sorbonne Paris Cité Research Center, Paris, France.
Paris Descartes University, Paris, France.
CHU Lille, Department of Neonatal Medicine, Jeanne de Flandre Hospital, Lille, France.
Department of Neonatology, Toulouse University Hospital, Toulouse, France.
Department of Neonatal Pediatrics and Intensive Care, Brest University Hospital, Brest, France.
Department of Neonatal Pediatrics, Dijon University Hospital Dijon, France.
Clinical Research Unit, Center for Clinical Investigation P1419, Cochin Broca Hôtel-Dieu Hospital, Paris, France.



Facilitating factors and barriers to breast milk feeding (BMF) very preterm (VP) infants have been widely studied at the individual level. We aimed to describe and analyse factors associated with BMF at discharge for VP infants, with a special focus on unit policies aiming to support BMF.


We described BMF at discharge in 3108 VP infants enrolled in EPIPAGE-2, a French national cohort. Variables of interest were kangaroo care during the 1st week of life (KC); unit's policies supporting BMF initiation (BMF information systematically given to mothers hospitalised for threatened preterm delivery and breast milk expression proposed within 6 hours after birth) and BMF maintenance (availability of protocols for BMF and a special room for mothers to pump milk); the presence in units of a professional trained in human lactation and regional BMF initiation rates in the general population. Associations were investigated by multilevel logistic regression analysis, with adjustment on individual factors.


In total, 47.2% of VP infants received BMF at discharge (range among units 21.1%-84.0%). Unit policies partly explained this variation, regardless of individual factors. BMF at discharge was associated with KC (adjusted odds ratio (aOR) 2.26 (95% confidence interval (CI) 1.40, 3.65)), with policies supporting BMF initiation (aOR 2.19 (95% CI 1.27, 3.77)) and maintenance (aOR 2.03 (95% CI 1.17, 3.55)), but not with BMF initiation rates in the general population.


Adopting policies of higher performing units could be an effective strategy for increasing BMF rates at discharge among VP infants.

[Indexed for MEDLINE]

Supplemental Content

Full text links

Icon for Wiley
Loading ...
Support Center