The automatic sampling shield: a device for sampling suckled breast milk.

The automatic sampling shield uses the negative pressure exerted by the breast-feeding baby during sucking to withdraw a series of small samples from the milk stream. Shield performance was studied at four feeds for each of 13 babies. Acceptance of the shield was good (96% of feeds) and samples were obtained in 88% of feeds. Sample weight was large enough for micro-biochemical analysis of composition (median weight 1.3 g), without depriving the baby of important amounts of milk (median percentage of weight of milk intake was 4.3% and sample weight was less than 5 g in all cases). The shield should sample a constant proportion throughout the feed, in order to obtain a sample that is representative of the infant's nutrient intake. Fractional test-weighing of feeds showed a tendency for a larger proportion of the feed to be collected at slower flow rates. Modelling on the basis of the observed flow profiles in conjunction with a theoretical fat concentration profile showed that the observed 'disproportionality' would have over-estimated fat intake by a mean of 2.6% per feed (S.D.9.1 range -18.9% to + 16.9%). Using the shield in random sequence at two out of the first four feeds after 0800 h, showed that an infant's mean milk intake with the shield was on average 32% less than on control feeds. Allowing for possible compensation by the infant when control feeds followed low shield intakes, mean milk intakes with the shield were 17% smaller than without the shield.