Estimating milk intake of hospitalized preterm infants who breastfeed.

This study describes the accuracy of clinical indices to estimate the milk intake of breastfed preterm infants. Twenty-nine mother-infant pairs were studied for a total of 39 breastfeeding sessions. Two NICU nurses (RN1 and RN2) and one certified lactation educator (CLE) collected data, each blind to the others' measures. RN1 and RN2 performed test-weights using the Smart Model 20 electronic scale. The CLE observed the breastfeeding and estimated milk intake, using clinical indices of milk transfer. Clinical indices did not provide an accurate estimate of milk intake (r = .48). The mean absolute difference between the test-weights and clinical estimates was 5.79 ml, with a maximal difference of 20 ml. These differences were random in that clinical indices did not consistently over- or underestimate milk intake. None of 17 clinical indices of milk intake significantly lowered the magnitude of error in the clinical estimate. These data suggest that clinical indices cannot serve as a replacement for test-weighing of preterm infants when an accurate estimate of milk intake is necessary.