In the NICE-SUGAR trial, intensive insulin therapy raised risks for severe hypoglycemia and 90-day mortality in critically ill patients, compared with conventional treatment (JW Gen Med Mar 24 2009). In this retrospective study that involved two Australian intensive care units and 4946 patients, investigators assessed the association between hypoglycemia and mortality in critically ill patients.
Overall, 1109 patients (22%) experienced at least one episode of hypoglycemia (blood glucose level, <81 mg/dL). In-hospital mortality was significantly greater among patients who experienced hypoglycemia than among normoglycemic patients (37% vs. 20%). Multivariate analysis revealed that illness severity and severity of hypoglycemia, but not receipt of insulin therapy, were associated independently with elevated mortality.
Comment: Hypoglycemia (either spontaneous or insulin-associated) is associated with excess mortality in critically ill patients. The reasons for this association are unclear. The authors speculate, however, that severity of hypoglycemia is associated with severity of illness, that hypoglycemia is a marker of imminent death, and that hypoglycemia has a true deleterious effect on critically ill patients. This study and prior research clearly indicate that hypoglycemia should be avoided in critically ill patients.
Published in Journal Watch General Medicine March 25, 2010