We initiated a study at Baylor Jack and Jane Hamilton Heart and Vascular Hospital to compare the sliding scale insulin (SSI) protocol used in 2006 with the SSI protocol currently used to treat diabetic patients admitted for procedures or surgery. An audit of patients' records revealed greater variation in staff compliance with the current protocol than with the previous one. In addition, it seemed that more patients were refusing insulin coverage under the current protocol than under the prior version. Although the study was aborted, the initial findings motivated us to identify obstacles to glucose control and to launch a health care improvement initiative to increase compliance with the SSI protocol. As a result of this process, the hospital has made several changes, including re-educating staff nurses, initiating competency checks of protocol interpretation, promoting patient education, and implementing early identification of inconsistent glucose control.