Hospital-wide breastfeeding rates vs. breastmilk provision for very-low-birth-weight infants

Aim To investigate the relationship between breastmilk feeding in very-low-birth-weight infants in the neonatal intensive care unit and breastmilk feeding rates for all newborns by hospital. Methods This was a cross-sectional study of 111 California hospitals in 2007 and 2008. Correlation coefficients were calculated between overall hospital breastfeeding rates and breastmilk feeding rates of very-low-birth-weight infants. Hospitals were categorized in quartiles by crude and adjusted very-low-birth-weight infant rates to compare rankings between measures. Results Correlation between breastmilk feeding rates of very-low-birth-weight infants and overall breastfeeding rates varied by neonatal intensive care unit level of care from 0.13 for intermediate hospitals to 0.48 for regional hospitals. For hospitals categorized in the top quartile according to overall breastfeeding rate, only 46% were in the top quartile for both crude and adjusted very-low-birth-weight infant rates. On the other hand, when considering the lowest quartile for overall breastfeeding hospitals, three of 27 (11%) actually were performing in the top quartile of performance for very-low-birth-weight infant rates. Conclusions Reporting hospital overall breastfeeding rates and neonatal intensive care unit breastmilk provision rates separately may give an incomplete picture of quality of care.