Insulin, Quality Improvement

How Grady Improved Patient Safety in the ICU with Computer-Guided IV Insulin Therapy

Hyperglycemia in the ICU is associated with higher infections and mortality rates. When I manage critically ill patients with hyperglycemia, insulin infusions are what I reach for first. When used safely, insulin infusions can quickly lower glucose into target range, improving outcomes. Guidelines recommend beginning treatment for BGs >180 mg/dl, with goals around 140-180 mg/dl, based on the results of the NICE-SUGAR study.1 Overuse or improper use of insulin, however, can result in hypoglycemia and adverse drug effects, which leads many practitioners to avoid using insulin infusions to treat hyperglycemia.

Guidelines, such as the AACE and 2020 ADA Standards of Care recommendations support the use of insulin infusions for critically ill patients. Despite the recommendations, subcutaneous basal insulin is often used as initial treatment for management of critically ill patients in the ICU, excluding DKA or HHS patients. I frequently have colleagues simply monitor glucose, or start either sliding-scale insulin only or basal insulin, only to see hyperglycemia persist or being forced to switch to an insulin infusion eventually due to severe hyperglycemia. 

The common refrain I hear in pushback for beginning IV insulin infusions is the fear of hypoglycemia or trying to ease the nursing burden of managing insulin infusions. There is little evidence comparing IV insulin to subcutaneous insulin in the ICU, but one VA study from 2019 compared these 2 groups and showed a lower rate of hypoglycemia (<70 mg/dl) in the IV group compared to subcutaneous (1.2% vs 2.2%).2

This past year, Grady Memorial Hospital, in Atlanta, Georgia, looked to clarify the differences between IV insulin compared to SubQ insulin to manage critically ill patients in the ICU.3 They all were on treatment for at least 48 hours.

They examined 100 patients on IV insulin, using Glytec's Glucommander® insulin dosing decision support, compared to 100 patients on subcutaneous (SC) insulin with basal insulin. They excluded patients with DKA, as these patients would generally be managed with IV insulin only. The subcutaneous group was on a basal insulin and management was per provider preference. 

They had patients on a variety of different medical and surgical units in the hospital. Patients on IV insulin were managed with Glucommander, a computer-guided eGlycemic Management System® (eGMS®). 

Results:

  • Severe hypoglycemia rates (<40 mg/dl) were zero in the IV Glucommander group and 0.4% in the SC group. 
  • Mild hypoglycemia (<70 mg/dl) were similarly disparate, 0.1% in the IV group and 1.9% in the SC group. 
  • Time to Target was 7.1 hours on average for patients on IV insulin compared to 18.6 hours in the SC group. 
  • The IV group had significantly less readings >180 mg (IV 28% vs SC 49%).

This study disproves the common fear of hypoglycemia when insulin infusions are run safely and effectively. Severe hypoglycemia was nonexistent, even with 20% of IV insulin patients with chronic renal disease or liver disease, known risk factors for hypoglycemia.

This data supports guidelines continuing to support insulin infusions for management of hyperglycemia in critically ill patients and discouraging subcutaneous insulin for the initial management of these patients for glycemic control.

Watch the Grady Study Video


1 NICE-SUGAR Study Investigators, Finfer S, Chittock DR, et al. Intensive versus conventional glucose control in critically ill patients. N Engl J Med. 2009;360:1283-1297.

2 Tran KK, Kibert JL 2nd, Telford ED, Franck AJ. Intravenous Insulin Infusion Protocol Compared With Subcutaneous Insulin for the Management of Hyperglycemia in Critically Ill Adults. Ann Pharmacother. 2019;53(9):894-898. doi:10.1177/1060028019841363

3 Rabinovich M, Hall A, Gayed R, Chester K, Crowe J, Messler J. Improved Patient Safety and Glycemic Outcomes in the ICU Using Computer-Guided Intravenous Insulin Therapy Versus Subcutaneous Insulin Therapy. DTS Virtual Poster Meeting. June 18, 2020.

 #0842-A

Subscribe to our blog!