Hyperglycemia in hospitalized, non‐critically ill patients is associated with poor outcomes including increased length of hospital stay, infection, disability after discharge from hospital, and death. Current guidelines for the management of inpatient hyperglycemia suggest the use of basal‐bolus insulin regimens over sliding‐scale insulin regimens. In recent clinical trials, basal‐bolus insulin regimens were superior to sliding scale insulin regimens in achieving glycemic goals, and were associated with a lower risk of hypoglycemia. Dosing of basal‐bolus regimens is weight‐based, and can be modified based on patient characteristics (e.g. elderly, underweight, poor diet). A suggested dosing range for most patients with diabetes is between 0.4 to 0.5 units of insulin/kg/day divided evenly between basal and nutritional insulin. Anecdotal observations suggest frequent under‐dosing of basal‐bolus insulin regimens and low achievement of glycemic control in hospitalized medicine patients. The objective of this study is to characterize the dosing patterns of basal‐bolus regimens in hospitalized, non‐critically ill patients with diabetes and to measure the proportion of patients who achieve target glycemic goals.
This is a retrospective, cohort study including Veterans Affairs (VA) patients with diabetes hospitalized between January 1, 2010 and December 31, 2011. After inclusion and exclusion criteria were applied, 82 patients were included in the final analysis. In addition to descriptive analyses, we performed simple linear regression to determine if insulin dosing was weight based. The proportion of patients achieving glycemic goals and hypoglycemic events was measured. Lastly, we performed multivariate logistic and linear regression to identify factors associated with achieving glycemic goals.
In patients receiving basal‐bolus insulin regimens, 32 out of 82 (39%) achieved glycemic goals. In patients achieving glycemic goals, the initial mean weight‐based insulin dose was 0.50 units/kg compared to 0.34 units/kg in patients not achieving goal (p = 0.11). Of 24 patients not receiving outpatient insulin, 13 (54%) were initiated on less than 0.2 units/kg/day of insulin. Factors associated with achieving target glycemic goals were outpatient insulin use (OR 9.3; 95% CI 1.7,51.8; p=0.01), initial basal‐bolus dosing > 0.2 units/kg (OR 4; 95% CI 0.83, 19.7; p=0.08), and A1c at goal on admission (OR 6.3; 95%CI 1.6,24.7; p<0.01). Patients achieving glycemic goals compared to those who did not were more likely to have an episode of hypoglycemia (< 70 mg/dL), 47% vs. 15%, p=0.01, but not severe hypoglycemia (< 40 mg/dL).
In hospitalized, non‐critically ill patients with a history of diabetes, achievement of target glycemic goals is low. A pattern of under‐dosing and non‐weight based basal‐bolus insulin regimens was observed. The implementation of weight‐based basal‐bolus insulin protocols is needed to improve appropriate dosing and achievement of glycemic goals. Furthermore, the use of approved protocols is recommended to reduce the incidence of hypoglycemic events in patients receiving weight‐based insulin regimens.