Abstract

OBJECTIVE—Hyperglycemia occurs in most critically ill patients. Using continuous glucose monitoring (CGM), we investigated whether intensive insulin therapy based on discontinuous glucose monitoring can achieve normoglycemia (80–110 mg/dl) in a medical intensive care unit (MICU).

RESEARCH DESIGN AND METHODS—Fifty adults (men/women 31/19, age 62 ± 16 years, nondiabetic/diabetic 30/20, intravenous/subcutaneous insulin 22/28, and Acute Physiology and Chronic Health Evaluation II score 22 ± 7) were prospectively recruited. Forty-eight–hour CGM was performed using a subcutaneous glucose sensor (GlucoDay) and compared with arterial glycemia. Main outcome measures were percent of time in normoglycemia and accuracy/applicability of CGM.

RESULTS—During 48-h CGM, glycemia reached target (80–110 mg/dl) in only 22 ± 18%, was >140 mg/dl in 39 ± 27%, and was <60 mg/dl in 5 ± 10% of the time. Patients on subcutaneous versus intravenous insulin had more glycemia readings >110 mg/dl (P = 0.016). Glycemia was higher in diabetic patients (170 ± 77 vs. 129 ± 35 mg/dl, P = 0.013). BMI was an independent determinant for bad glycemic control (β = 0.73, P < 0.0001). Diabetic state (β = 0.47, P < 0.0001), septic shock (β = 0.22, P = 0.045), sequential organ failure assessment score (β = 0.40, P = 0.001), and use of corticoids (β = 0.28, P = 0.014) and inotropics (β = −0.24, P = 0.035) were independent determinants of insulin dose. GlucoDay values and arterial glycemia correlated well (r = 0.85, P < 0.0001, n = 555 after six-point calibration), with 97% of data falling in regions A and B of error grid analysis. There were no adverse events using GlucoDay.

CONCLUSIONS—GlucoDay, a well-tolerated 48-h CGM system, revealed that normoglycemia was only achieved 22% of the time in MICU patients. Further studies should investigate whether application of CGM to titrate insulin therapy can improve patient outcome.