[The optimal blood glucose target in critically ill patient: comparison of two intensive insulin therapy protocols].

BACKGROUND AND OBJECTIVE: Recent studies in critically ill patients receiving insulin intravenous therapy (IIT) have shown an increased incidence of severe hypoglycemia, while intermittent subcutaneous insulin «sliding scales» (conventional insulin therapy [CIT]) is associated with hyperglycemia. The objective of this study is to assess whether glycemic control range IIT can affect glucose levels and their variability and to compare it with CIT.

PATIENTS AND METHOD: Prospective comparative cohort study in intensive care unit, with 2 study periods: Period 1, IIT with glycemic target range 110-140 mg/dL, and Period 2, IIT of 140-180 mg/dL. In both periods CIT glycemic target was 110-180 mg/dL. We assessed severe hypoglycemia (< 50 mg/dL), moderate hypoglycemia (51-79 mg/dL), hyperglycemia (> 216 mg/L) and the variability of blood glucose.

RESULTS: We studied 221 patients with 12.825 blood glucose determinations. Twenty-six and 17% of patients required IIT for glycemic control in Period 1 and 2, respectively. Hypoglycemia was associated with a discontinuous nutritional intake, glycemic target 110-140 mg/dL and low body mass index (BMI) (P = .002). Hyperglycemia was exclusively associated with a history of diabetes mellitus (OR 2.6 [95% CI 1.6 to 4.5]). Glycemic variability was associated with a discontinuous nutritional intake, low BMI, CIT insulinization, diabetes mellitus, elderly and high APACHE II (P < .001).

CONCLUSIONS: The use of IIT is useful to reduce the variability of blood glucose. Although the 140-180 mg/dL range would be more secure as to presenting greater variability and hyperglycemia, the 110-140 mg/dL range is most suitable.