321: OUTPATIENT DIABETES MANAGEMENT INFLUENCES GLYCEMIC CONTROL IN ICU PATIENTS DURING NUTRITION SUPPORT

Abstract

Introduction: Appropriate glycemic control improves clinical outcomes during critical illness. We proposed that outpatient diabetes management would influence glycemic control in critically ill patients who require nutrition support (NS). Methods: Adult patients admitted to the trauma intensive care unit (ICU), who required NS, and exhibited hyperglycemia (blood glucose concentration (BG) > 150 mg/dL) or had diabetes mellitus (DM), were retrospectively evaluated. Outpatient DM pharmacotherapy was discontinued upon ICU admission. NS was initiated on day 3 + 3 following ICU admission. A serum hemoglobin (Hgb) A1c was obtained on day 3 + 3. The daily amount/mode of insulin therapy, carbohydrate intake, and point of care BGs for the first 7 days of NS were recorded. Patients were classified as controlled DM (Hgb A1c < 7%), uncontrolled DM (Hgb A1c >/= 7%), or as hyperglycemic patients without DM (Hgb A1c < 6.5%). Target BG was 70 to 150 mg/dL. Hypoglycemia was defined as a BG < 70 mg/dL. Regular human insulin sliding scale (SSI), subcutaneous NPH insulin with concurrent SSI, or continuous intravenous insulin infusion (CIVI) were used for glycemic control at the discretion of the NS service or primary service. Continuous data were given as mean + S.D. Chi-square analysis or ANOVA were used to compare groups. Results: Seventy-eight (14 controlled DM, 21 uncontrolled DM, and 43 non-DM) patients were studied. Maximum BG prior to initiation of NS was significantly different between groups (201 + 54 vs. 334 + 192 vs. 191 + 57 mg/dL, respectively, P < 0.001). The proportion of patients who required higher intensity insulin therapy (NPH + SSI or CIVI) was different among groups (50% vs. 86% vs. 15%, respectively, P < 0.001). The amount of insulin given per gram of carbohydrate intake was 0.45 + 0.50, 1.24 + 1.21, and 0.15 + 0.21 units/g, respectively, P < 0.001. Time within the target BG range was 11 + 6, 12 + 5, and 17 + 6 hrs/day, respectively, P < 0.001. Hypoglycemia occurred in 1 (7%), 5 (24%), and 3 (7%) patients, respectively (P = 0.120). Conclusions: Glycemic control during NS in the ICU is more challenging for those patients whose outpatient DM was uncontrolled compared to those with controlled DM and those with stress-induced hyperglycemia without DM.