LETTERS

Reducing the incidence of neonatal hypernatraemic dehydration Iyer et al1 report the effects of a neonatal weight-monitoring policy on the incidence of hypernatraemic dehydration. This policy led to an increased recognition of affected infants, and identified cases at an earlier and milder degree of dehydration. We examined the effect of introducing a similar policy on cases of more severe dehydration. We wished to test whether the background rate of hyperna-traemic dehydration was rising and to test whether a weight-monitoring policy had any effect on the background incidence. We studied cases of moderate and severe hypernatraemic dehydration in infants born in a single maternity unit attached to the sole in-patient paediatric unit for the city. The maternity unit delivers approximately 3500 births per annum. Infants requiring re-admission in the neonatal period may be admitted to the paediatric in-patient unit or to the maternity hospital of birth. We reviewed the biochemical records for 9 years (January 1998 – December 2006) from the single laboratory that serves both the paediatric in-patient unit and the attached maternity unit. We identified 198 cases of infants under the age of 4 weeks with a plasma sodium of.150 mmols/l. We were able to obtain and review 178 (90%) of these case-notes. Many cases of hypernatraemia were due to underlying medical or surgical problems including prematurity. We identi-fied 67 cases of hypernatraemic dehydration in otherwise healthy breast-fed neonates during this period. A weight-monitoring policy was introduced at the end of 2003. Babies were weighed around days 2, 5 and 10 and those with>10 % weight loss were seen by specialist breast-feeding support sisters for supervised feeding, advice on positioning and milk expression. Initially, we compare