Editor,—I read with interest Booth’s comments1 on my article2 and I take this opportunity to clarify our findings. Mackenzie et al’s study3 was fundamentally different from ours. The degree of dehydration was assessed by junior doctors and not specialists; all children were inpatients “thought to be” 5% to 10% dehydrated (only 25% were) and no cases of severe dehydration were included. Most were “social’ admissions.

During rehydration, children were not fed, and weight loss from catabolism may have led to underestimation of the degree of dehydration. Children’s weight may vary tremendously with voiding or defecation and those factors were not included; therefore, the weight increase after admission may not be the “gold standard”. Nine years later, successive editions of major paediatric textbooks from both sides of the Atlantic still rely on clinical parameters to estimate the severity of dehydration, although they are not perfect.

At Booth’s suggestion, we looked at the percentage of weight gain 24 hours after admission for the 64 children who were admitted to hospital, and compared it with their serum CO₂ on admission: those with a weight increase of ⩽ 5% had a CO₂ on admission of 17 mmol/l (range 11–22); those with a weight increase of 5–10% had a CO₂ on admission of 16 mmol/l (11 to 19); and those who gained > 10% had a mean CO₂ of 19 mmol/l (range 12–20). There was also a poor correlation between weight gain and the degree of dehydration as assessed clinically.

The ideal parameter to estimate the severity of dehydration would be the difference between admission and pre-admission weight. As the latter is often lacking, physicians can rely only on clinical parameters. I would be interested to learn how Booth immediately assesses the degree of fluid loss in a dehydrated child. Perhaps all textbooks will need to be modified if his methods prove to be scientifically valid.