Effects of Computerized Order Sets on Inpatient Management of Hyperglycemia: A Randomized Controlled Trial

Background:

Inpatient hyperglycemia is common and associated with adverse patient outcomes. Current guidelines recommend treatment to achieve glucose levels below 180 mg/dL in the inpatient non‐ICU setting. Previous research has suggested various approaches to improving inpatient glucose management. However, few studies have rigorously evaluated the ability of specific, low‐cost interventions to increase adherence with guidelines and improve glycemic control.

Methods:

At an academic medical center, a multidisciplinary team developed, tested, and refined 3 order sets built into the hospital's computer provider order entry (CPOE) system: 1 each for patients eating discrete meals, receiving continuous enteral nutrition, or receiving nothing by mouth. Elements of the order sets included orders for basal, nutritional, and supplemental insulin, integrated dosing guidelines, and provisions for glucose monitoring, diet orders, and treatment of hypoglycemia. We then conducted a cluster‐randomized controlled trial on 4 hospita ist‐run general medicine service teams. All study teams received a detailed subcutaneous insulin protocol based on ADA guidelines, and case‐based education was provided to nurses, residents, and hospitalists. The 2 intervention teams also received access and orientation to the CPOE order sets. Study subjects were prospectively identified: consecutively enrolled patients with type 2 diabetes or at least 1 laboratory glucose value > 180 mg/dL. The primary outcome was the mean percentage of glucose readings per patient between 60 and 180 mg/dL. Multivariable binomial logistic regression using general estimating equations was conducted to adjust for potential confounders and clustering by intern.

Results:

Between April 5 and June 20, 2006, we identified 179 eligible study subjects: 90 patients admitted to intervention teams and 89 to usual‐care teams. There were no significant differences between study groups with respect to age, admission glucose, Hb A1C, prior insulin use, or prior diagnosis of diabetes. The mean percentage of readings per patient between 60 and 180 mg/dL was 75% in the intervention group and 71% in the usual‐care group (adjusted relative risk, 1.36, 95% Cl 1.03‐1.80). Compared with usual care, the intervention group also had a lower patient‐day weighted mean glucose (148 vs. 158 mg/dL, P = 0.04), less use of sliding‐scale insulin by itself (25% vs. 58%, P = 0.01), and no difference in the rate of severe hypoglycemia (glucose < 40 mg/dL; 0.5% vs. 0.3% of patient days, P=0.40).

Conclusions:

The use of order sets built into a hospital's CPOE system led to improvements in glycemic control and insulin ordering without causing a significant increase in hypoglycemia, compared with an insulin protocol and education alone. Other institutions with CPOE should consider adopting similar order sets for their own use as part of a comprehensive glycemic management program.

Author Disclosure:

J. Schnipper, SHM, Speaker; C. Liang, none; C, Ndumele, none; M. Pendergrass, Medco, employment.