Translating evidence into practice in managing inpatient hyperglycemia

James Krinsley, MD, has been director of critical care at Stamford Hospital in Stamford, CT since 1998. This 305-bed hospital, a major teaching affiliate of the Columbia University College of Physicians and Surgeons, has a 14-bed adult ICU that treats a heterogeneous mix of medical, surgical, and cardiac patients; cardiovascular surgery is not yet done at the institution. The data and protocoldriven model of care developed in the ICU resulted in the unit’s recognition in 2002 by the National Coalition on Healthcare and the Institute for Healthcare Improvement as 1 of the 11 “best practices” ICUs in the nation. In 2004 Stamford Hospital won the Codman Award from the Joint Commission on Accreditation of Hospitals and Organizations based on the ICU’s development of an intensive glycemic management protocol that resulted in a 29% reduction in mortality of patients admitted to the unit. The last 15 years have brought reports in the medical literature of exciting advances in describing the relationship between hyperglycemia and adverse outcomes in a variety of clinical contexts involving acutely ill patients. Hyperglycemia in hospitalized patients was long thought to be an adaptive mechanism and, at least in the intensive care setting, was rarely treated below threshold values of 225-250 mg/dL. The pioneering work of Furnary et al. and the Portland Diabetic Project was the first to demonstrate that close monitoring and treatment of hyperglycemia in diabetic patients undergoing cardiovascular surgery decreased the occurrence of deep sternal wound infections, a dreaded postoperative complication. A second publication documented the steady decrease in mortality among these patients over the years as the group’s glycemic target was steadily lowered. In the last several years the mortality rate of diabetic patients undergoing cardiovascular surgery has decreased so that it now approximates that of nondiabetics, eliminating the “diabetic disadvantage.” This work set the stage for the landmark Leuven study, performed at Catholic University in Belgium and published by Van den Berghe’s group in 2001. This prospective, randomized, controlled study involving 1548 mechanically ventilated patients in a surgical intensive care unit, 63% of whom had undergone cardiovascular surgery, compared the outcomes of patients treated with continuous intravenous insulin to achieve euglycemia (80-110 mg/dL) to those of a control group that received treatment only when glucose level exceeded 210 mg/dL. The outcomes including a 37% reduction in hospital mortality in the treated group and a 40%-50% reduction in numerous morbid conditions, including the need for renal replacement therapy, prolonged mechanical ventilation, prolonged antibiotic use, and critical illness polyneuropathy, that spawned a paradigm shift in ICU medicine. A large before-and-after study performed in a mixed medical-surgical ICU of a university-affiliated community hospital confirmed the mortality benefits of glycemic management, using a more modest target of 80-140 mg/dL. Finally, a prospective, randomized, controlled trial in a medical ICU population by the Leuven investigators reported improvement in several morbidities and a mortality advantage from intensive glycemic control, targeting 80-100 mg/dL, among patients with ICU stays longer than 3 days. Consequently, intensive glycemic management of critically ill patients is rapidly becoming a worldwide standard of care, presenting an array of challenges to clinicians involved in the care of these patients. This article presents an B R I E F R E P O R T