Introduction: Hyperglycemia in sepsis is managed by intensive insulin therapy, which can cause hypoglycemia.
Objectives: The aim of the study was to evaluate the glycemic profile as well as safety and effectiveness of a nurse-controlled insulin therapy protocol in patients with severe sepsis and septic shock.
Patients and methods: The study included 16 septic patients who died (nonsurvivors) and 61 septic patients who survived. Glycemia was measured every 4 h, and the dose of insulin infusion was adjusted to maintain glycemia of 4.4 mmol/l to 8.3 mmol/l. We analyzed glycemia levels and daily variations, insulin dose, episodes of hypo- and hyperglycemia.
Results: Nonsurvivors and survivors had similar mean glycemia levels (7.38 vs. 7.08 mmol/l; p = 0.20) and insulin requirements (median [Me] = 26.9 vs. 23.9 units/d; p = 0.22; Me = 1.7 vs. 1.4 units/h; p = 0.25). Daily glycemia variation (Me = 4.81 vs. 3.03 mmol/l; p <0.001), episodes of hypoglycemia (18.8% vs. 3.3%; p = 0.02), spontaneous severe hypoglycemia (12.5% vs. 0%; p = 0.006) and hyperglycemia (75.0% vs. 45.9%; p = 0.04) were higher and more frequent in nonsurvivors. Three of 5393 blood samples (0.05%) met severe insulin-induced hypoglycemia criteria, and 74.4% of samples met the recommended range of 4.4-8.3 mmol/l.
Conclusions: Patients who died experienced more episodes of hyperglycemia, spontaneous hypoglycemia and greater variation in the daily glycemia level. Daily glycemia variation is more reliable than a mean glycemic level in evaluating glucose homeostasis in septic patients. Few episodes of severe insulin-induced hypoglycemia occurred while using the nurse-controlled insulin therapy protocol.