Aim: Diabetic ketoacidosis is a hyperglycaemic emergency that is often treated in intensive care units (ICUs) despite having a low mortality and good prognosis. Current risk stratification is based primarily on acidosis, but it has been suggested that hyperosmolarity may also be an important marker of increased severity. Our aim was to evaluate the relationship between raised serum osmolarity and adverse clinical outcomes in ICU admissions for ketoacidosis.
Methods: Retrospective review of prospectively collected data for adult admissions with ketoacidosis in the Australian and New Zealand Intensive Care Society Adult Patient Database over a 15-year period (2004-2018). Exclusions were readmissions and records with critical missing data. Serum hyperosmolarity was defined as > 320 mosm/l. The primary outcome was hospital mortality; secondary outcomes were ICU mortality and other adverse clinical events.
Results: Some 17 379 admissions were included in the study population. People with hyperosmolarity had fourfold increased mortality, a higher incidence of renal failure and need for mechanical ventilation, and prolonged ICU and hospital length of stay. The relationship with mortality remained highly significant even after adjusting for severity of acidosis, hospital type, year of admission, time to ICU, and a modified Australia and New Zealand Risk of Death propensity score.
Conclusions: Although adults with ketoacidosis have a good prognosis overall, hyperosmolarity was independently associated with a significantly higher incidence of multiple adverse outcomes including mortality. Whether or not this is directly causal, it may have practical applications to improve risk stratification and identify individuals at risk of adverse outcomes.
© 2020 Diabetes UK.