|
|
Newsletter
Glucose Control and Mortality in Critically Ill PatientsFinney SJ, Zekveld C, Elia A, Evans TW. JAMA 2003; 290: 2041-2047Reviewer: Michael H. Wall, MD Dallas, TX This single-center prospective observational study investigated whether the mortality reduction seen in a previous trial of intensive insulin therapy1 was due to blood glucose level, quantity of insulin administration or a combination of both. Patients admitted during the first 6 months of 2002 were enrolled. All clinical observations, laboratory data, drug infusion rates, physiologic monitoring data, length of stay and mortality statistics were stored in a computerized information system. Blood glucose was maintained between 90-145 mg/dL, without the use of a fixed regimen, using human insulin infusions. Drugs were delivered in 5% dextrose and all patients received entral feeding unless extubation was planned within 12 hours. Six bands of glycemic control were prospectively defined: hypoglycemic (blood glucose level < 80 mg/dL), stringent (80-110 mg/dL), normal (111-144 mg/dL), intermediate (145-180 mg/dL), liberal (181-200 mg/dL) and hyperglycemic (> 201 mg/dL). Each patient's glucose values were time-weighted by assuming a linear trend between individual measurements and calculating the minutes spent in each band, expressing the result as a proportion of the whole admission. The primary outcome variable was ICU mortality. Secondary end points were hospital mortality, length of stay and predicted threshold glucose level associated with death. Five hundred, twenty-three patients (73% male, average age 64) were analyzed. 86% were admitted following cardiac surgery, 3% following thoracic surgery and 11% were medical admissions. Blood glucose levels did not statistically differ between survivors and non-survivors. Most patients spent a considerable amount of time outside the target range of 90-145 gm/dL, but few patients had glucose values < 79 gm/dL. Multi-variable logistic regression and modeling using confounding and non-confounding variables were done. The amount of exogenous insulin given was significantly greater in patients who died. At all glucose levels (except those with a blood glucose < 79 mg/dL) increased insulin administration was associated with a significantly increased risk of death [odds ratios 1.02 (95% confidence intervals 1.003 -1.04); p<0.001]. This implies that glucose control, not the amount of exogenous insulin, is responsible for the improvement in mortality shown in other studies of tight glucose control. This study is important because, even though it is an unblinded observational study, it represents what is done in clinical practice and shows an association between increased insulin dose and mortality at all glucose levels > 80 mg/dL. Studies need to be done to find better protocols for glucose control and to more clearly define what the target glucose should be. Reference
Table of Contents:
© Society of Cardiovascular Anesthesiologists Questions or comments? Please send email to webmaster@scahq.org |
