Evaluation of Admission Blood Glucose Levels in The Intensive Care Unit

Abstract

To assess the availability and clinical value of blood glucose (BG) testing at the time of admission to the intensive care unit (ICU) after such testing was implemented as routine care in the ICU. We studied ICU admission BG testing rates and the prevalence of hyperglycemia. In this effort, we assessed the frequency of baseline BG testing in 330 consecutive patients during a period of 3 months and then implemented routine BG monitoring in 1,147 consecutive ICU patients during a 7-month period. Of the total study population, 25% had previously diagnosed diabetes (PDD). At baseline, 70% had BG measured within 4 hours before or after ICU admission (99% of patients with and 60% of patients without PDD). After implementation of routine BG monitoring, there was a significant increase in testing (70% before versus 87% after, P<0.001; 70% during the baseline 3-month period versus 93% in the final 3 months of the study, P<0.001). In patients without PDD, 41% had BG levels < or =140 mg/dL, and 8% had BG concentrations < or =200 mg/dL. Overall in the ICU setting, 57% of BG values < or =140 mg/dL and 33% of BG levels < or =200 mg/dL were in patients without PDD. Frequencies of BG testing by admission diagnosis included the following (at baseline and during the final 3 months after implementation of routine BG tests): postsurgical status (46%, 85%), peripheral vascular disease (51%, 90%), neurologic disease (52%, 83%), gastrointestinal disease (58%, 91%), infection (69%, 100%), and diabetes (100%, 100%). Rates of routine BG testing are low in ICU patients without PDD. Elevations in BG levels were detected in 41% of our study patients without PDD, suggesting that routine implementation of BG monitoring in an ICU will identify patients at increased risk for hyperglycemia-associated higher morbidity and mortality.