Performance of the APACHE II and SOFA Scores in Diabetic Ketoacidosis

Abstract

Accurate prediction of severity and mortality in diabetic ketoacidosis (DKA) is important for allocation of resources. The APACHE II and SOFA scores are used to predict mortality in critically ill patients, however neither has been tested exclusively in DKA. We sought to determine if these scoring systems can accurately predict mortality in patients with DKA. This was an observational study of patients presenting to an urban tertiary care center with a diagnosis of DKA. Adult patients (age ≥ 18 years) with glucose > 250 mg/dL, bicarbonate ≤ 20 mEg/L, an anion gap ≥ 16 mEg/L, pH ≤ 7.30, and urine ketones were included. Predicted mortality based upon APACHE II and SOFA scores were compared to observed mortality. A total of 50 patients were included. There was no observed mortality in our population. The median APACHE II score was 10 (IQR: 6, 15) which predicted a mortality of 15% and the median SOFA score was 1 (IQR: 0, 2) which predicted a mortality of 0%. In summary, we found the APACHE II illness severity score does not accurately predict mortality in a population of patients with DKA, while the SOFA score appears to predict mortality in the same population.