Background: The optimal level of glycemic control in ICU patients has been the subject of intense investigation over the last 20 years. A pivotal study (the NICE-SUGAR study) involving more than 6,000 patients has established a target between 8 and 10 mmol/l (144 to 180 mg/dl) as the current standard of care. However, this study did not address whether patients with diabetes should be treated differently and, in particular, whether in such patients a higher glucose target should be used. Main concepts: The last decade has seen multiple studies aiming to describe the association between glycemia in mortality according to whether patients have or do not have diabetes and whether, if they have diabetes, pre-ICU admission glucose control (assessed by glycated hemoglobin A1c (HbA1c) levels) affects the relationship between acute glycemia and outcome. All such studies (now involving thousands and thousands of patients) have consistently shown that diabetic patients have a different relationship between acute glycemia and mortality. In particular, in diabetic patients, increasing glucose levels up to 15 mmol/l (270 mg/dl) or more are not associated with increased risk of death. In patients with a high HbA1c (> 7%) prior to ICU admission, targeting a glucose level below 10 mmol/l (180 mg/dl) is associated with increased risk compared with permissive hyperglycemia. Finally, a recent controlled study comparing a glucose target between 10 and 14 mmol/l (180 to 252 mg/dl) to a glucose target between 6 and 10 mmol/l (180 mg/dl) in diabetic patients found no advantage from tighter glycemia control. A randomized controlled study called LUCID is now underway to test the hypothesis that permissive hyperglycemia might be safer in diabetic patients admitted to the ICU. Conclusions: Until the results of the LUCID trial are available, the burden of evidence is in favour with targeting a more relaxed level of glycemia in diabetic patients (10-14 mmol/l; 180-252 mg/dl), especially in those with poor pre-admission glycemic control.
- Randomized controlled trial