With respect to van Dommelen et al’s1 interesting paper we would like to make the following contribution.

Screening for neonatal hypernatraemic dehydration is complicated by an unclear, but possibly rising, incidence2 and the need to avoid undermining the “breast is bests” message while recognising the severity of the condition. The notoriously variable clinical presentation adds to this difficulty.

In common with many centres, we have seen a number of dehydrated babies following a period of sub-optimal feeding. A number of these babies have had severe neurological impairment following thrombotic events secondary to dehydration. We report four such cases, three breast fed and one mixed breast/formula fed, who presented to our paediatric unit in the past 5 years.

Case 1

A healthy second child, born at term by lower segment caesarian section for failure to progress, discharged on day 4 having established breast feeding for 40 minutes every 3–4 hours. Admitted at 11 days old with a four-day history of being floppy, quiet and one day of recognised poor feeding. Weight had not …