Perioperative blood glucose management in patients undergoing tumor hepatectomy

Study Objective: To determine whether our institutional insulin management (modified Atlanta) protocol is efficient and safe in controlling blood glucose levels in the perioperative period in surgical patients undergoing tumor hepatectomy. Design: Retrospective study. Setting: Large community hospital. Patients: 20 consecutive patients undergoing liver resection for hepatocellular carcinoma, liver metastasis, or other hepatobiliary tumors. Interventions and Measurements: All patients continuously received intravenous glucose (5% dextrose in water, one mL/kg/hr); insulin was administered according to a strict algorithm, and dose adjustments were based on measurements of whole-blood glucose intraoperatively at one-hour intervals, and in the intensive care unit (ICU). Lower and upper blood glucose limits were set at 85 mg/dL and 110 mg/dL, respectively, in the operating room (OR). In the ICU, lower and upper limits were 90 mg/dL and 140 mg/dL, respectively. Main Results: Intraoperatively, 51.3% of measurements were within the target range. In the ICU, 75.2% of rneasurements showed a blood glucose level of 90 - 140 mg/dL. Two of 78 (2.6%) and two of 363 (0.5%) measurements had a blood glucose level < 70 mg/dL in the OR and ICU, respectively. The lowest blood glucose levels were 65 mg/dL (OR) and 66 mg/dL (ICU). Conclusions: The modified Atlanta protocol is efficient and safe in controlling blood glucose levels in the perioperative period of hepatic tumor resection. Because of decreased insulin needs in the ICU, the use of a more liberal algorithm successfully reduced the risk of hypoglycemia. (C) 2009 Elsevier Inc. All rights reserved.