Objectives

To describe the use of pasteurised human donor milk (pHDM) in England and the influence of a human milk bank in the network.
Design

Prospective observational study
Setting

All 163 neonatal units (23 networks) in England 2012–2013.
Patients

Preterm infants born at <32 weeks gestational age (GA).
Main outcome measures

Proportion of infants and care-days fed pHDM during the first 30 postnatal days by network
Methods

We extracted daily patient-level data from the National Neonatal Research Database (NNRD). We fitted a logistic regression of pHDM exposure on the presence of a pHDM bank within the network, with GA, BW z score and network as covariates. Significance was assessed by the likelihood ratio (chi-squared) test.
Results

Data for 13,463 infants were included in the study. Across the networks, the proportion (95%CI) of infants ranged from 2.0% (1.0, 3.0) to 61.0% (57.4%, 64.6%), and the proportion of care-days in which pHDM was fed from 0.08% (0.04%, 0.10%) to 21.9% (19.9%, 24.0%). In three networks <5%, and in seven networks >30% of infants received any pHDM. Variation in the use of pHDM across networks remained significant after adjustment for presence of a human milk bank within the network and all covariates (p < 0.001).
Conclusions

Wide variation of pHDM use in England is not fully explained by presence of a pHDM bank or patient characteristics. This suggests clinical uncertainty about the use of pHDM.