. . . Including Sugar and Stress

Everything is more intense in the ICU. Critically ill patients demand more frequent and detailed monitoring as well as timely and precise interventions to address their fragile state. These patients health can be impacted by a wide range of factors, dreamstime_xs_46074433including blood sugar.

Statistics show that approximately 30 to 50% of ICU patients experience stress-induced hyperglycemia, which is typically unrelated to underlying diabetes. Multiple studies have associated this hyperglycemia with increased risk of infection, organ dysfunction, and mortality. Like others, ICU patients treated for high blood sugar frequently fall victim to a hyperglycemia – hypoglycemia swing. In this setting, where life often hangs in the balance, rapid and precise glycemic control is vital.

A recent study at the Oregon Health & Science University (OHSU) provided unique data from critically ill patients. The study compared glycemic control achieved using a computerized glucose management software against the paper insulin protocols. The software proved to be superior in decreasing mean glucose levels and avoiding hyperglycemia. For most patients, the advanced computer-based protocol was also associated with enhanced blood glucose stability. However, it had a higher incidence of moderate but not severe hypoglycemia.

Large Study, Multiple ICU Settings

The study examined 1,896 patients retrospectively in four ICU settings–surgical, medical/cardiac, trauma, and neuroscience–at an academic tertiary care hospital. All cardiothoracic surgery (CTS) patients were started on a protocol upon hospital admission. Other patients received treatment only after developing hyperglycemia.

For the cardiothoracic group, the mean glucose level for the software was 130.9 mg/dL, compared to a paper protocol result of 138.8 mg/dL. The proportion of patients with severe hyperglycemia of 10% or more were also lower in the software group (6.2%) than in paper protocol patients (15.5%).

For those placed on an insulin protocol after developing hyperglycemia, the mean software group glucose level was 141.5 mg/dL compared to 159.9 mg/dL. Likewise, the proportion of patients with severe hyperglycemia in the software group (35.2%) was lower than the paper protocol group (64.1%).