Josephine Ho and associates1 report an unfortunate case of a 6-year-old girl with diabetic ketoacidosis (DKA) and thromboembolic stroke. Although the authors do a credible job of describing the diverse causes of pediatric stroke and the controversies surrounding treatment of children, there was little emphasis on the danger of extreme hyperosmolar states and risks of thrombosis. More information about the initial presentation of the patient, with specific reference to the concentration of serum sodium and serum osmolarity, would have been helpful in determining her risks of thrombosis.
Diabetes is associated with a prothrombotic state through a number of mechanisms.2 The mostly adult entity of hyperosmolar nonketotic coma has had various degrees of association with thrombosis,2,3 as has extreme hypernatremia in breast-feeding neonates.4 Recent evidence has also demonstrated that among children with DKA, there is a higher incidence of deep venous thrombosis with femoral central venous lines.5,6 Serum glucose and sodium concentrations and hence effective plasma osmolarity were significantly higher in those patients with blood clots.5
Although there is no direct evidence for its efficacy, our practice has been to use prophylactic anticoagulation in patients with DKA who are in a significant hyperosmolar state, as well as to eliminate the use of femoral catheters in patients with these risk factors. There is significant controversy surrounding the dose of anticoagulant therapy, specifically whether the efficacy of dosages for prophylaxis of deep venous thrombosis outweighs the risks associated with full systemic anticoagulation.7 As with most clinical issues, particularly in pediatric critical illness, this controversy lends itself well to a clinical trial in patients with extreme hyperosmolar states, including those with DKA.
Footnotes
-
Competing interests: None declared.