Hypernatraemic Dehydration in Infantile Gastro-enteritis

Although it is possible to find occasional reports written before the last war of the occurrence of hyperelectrolytaemia in sick infants and children (for early references see Rapoport, 1947) the recognition of the condition as a fairly common one and the elucidation of the clinical features date from 1947 when Rapoport gave an account of 14 patients with hyperosmolarity of the plasma. Of the 14 patients, 10 were suffering from diarrhoea. Five of these were noted to have symptoms of central nervous system disturbance, from marked lethargy to convulsions. The first indication of the possible frequency of this state came in a report by Tarail, Bass and Runco (1953) who found that, of 191 infants with diarrhoea, 21 or 11% had serum sodium levels of 150 mEq/litre or higher. In 1955 Finberg and Harrison gave an account of seven years' experience of infantile diarrhoea involving 274 patients. They reported an incidence of hypernatraemia (serum sodium level over 150 mEq/l.) of 25% and pointed out the great frequency of 'manifestations referable to the nervous system', nearly two-thirds of the hypernatraemic cases showing such signs. They reported, for the first time, the occurrence of raised protein concentrations in the cerebrospinal fluid, and drew attention to the fact that nearly half the cases did not show the customary signs of dehydration. They indicated the danger that rapid rehydration with dilute fluids might give rise to symptoms similar to those seen in water intoxication, especially to convulsions, and they suggested that hypernatraemia could be a cause of permanent brain damage. This paper seemed to define a distinct clinical entity of considerable significance. It was followed by one by Weil and Wallace in 1956 which emphasized the points made by Finberg and Harrison, especially the neurological disturbances. Several other papers from America have reported similar findings and opinion in that country is probably represented by the account of 'Hypertonic dehydration' given by Cooke (1959). According to this writer hypertonic dehydration occurs in approximately 20% of cases of diarrhoea in infancy; 'of particular importance are the cerebral changes which result from severe hyperosmolarity'; the development of oedema during treatment is referred to as 'characteristic'. From consideration of the available reports the 'clinical picture' of hypertonic dehydration emerges as follows: the patients are predominantly infants in the first year of life; as the result of fluid loss in the stools and by vomiting there is often loss of 10-15% of the body weight without the classical signs of such a severe degree of dehydration, e.g. the skin turgor may be deceptively normal; symptoms and signs of cerebral damage are common and include disturbances of consciousness ranging from lethargy to coma; hyperirritability to stimuli in spite of extreme lethargy; increased muscle tone amounting in many cases to generalized rigidity with very brisk tendon reflexes; convulsions, especially during rehydration; raised concentration of protein in the cerebrospinal fluid and rather exceptionally, the development of sub-dural effusions. Oedema is liable to appear during the administration of rehydrating fluids, sometimes before full rehydration has been effected. It has been suggested (Franz and Segar, 1959) that this state is becoming commoner and also (Finberg, 1958) that it may be due in many instances to the unwise administration of fluids with 'high-solute' content. We wish to describe the two cases which first drew our attention to this condition and then to give an account of an analysis of the records of 100 infants with diarrhoea admitted to the Duchess of York Hospital for Babies on whom serumelectrolyte estimations were performed.