Recent literature regarding tight glycemic control: pitfalls in the sweet debate

Articles concluding that tight glycemic control (TGC) in the intensive care unit (ICU) has no mortality benefit and an unacceptably high rate of hypoglycemia have been published recently in several journals. The Diabetes Special Interest Group (DSIG) [1] believes that the data from some of these recent papers have been interpreted incorrectly, misconstrued, or misunderstood. The DSIG agrees with the scientists whose editorial comments were published with these articles [2, 3] that the studies were underpowered to show a lack of benefit and agrees that hypoglycemia below 40 mg/dL is an undesirable complication. The incidence of hypoglycemia in these studies compares unfavorably with data from results with the Glucommander, which in published data has an overall hypoglycemia rate (below 40 mg/dL) of only 2.6% [4], and more recently, no blood sugar below 40 mg/dL was seen in patients on the Glucommander in the cardiovascular ICU [5]. Algorithms for achieving TGC are being continually refined. The target ranges for ICU patients are firmly established in only the post-cardiac surgical population. The DSIG joins others in the hope that the NICE-SUGAR (Normoglycemia in Intensive Care Evaluation – Survival Using Glucose Algorithm Regulation) trial (currently in the analysis phase, having enrolled over 6,000 subjects) will add to the knowledge base for these issues and also notes that the principal investigator for this study has commented that even a negative finding for benefit will not provide evidence in favor of abandoning glucose control entirely [6].

The DSIG has learned during its six-year effort that instituting TGC is an individual institutional undertaking that first requires broad commitment from, among others, both the leadership and the implementing staff. Policies and protocols specific to TGC are essential. Standardization is a must. Chosen targets should be evidence-based and realistic for the individual institution. Ongoing monitoring of outcomes, including both the success rate for achieving the glycemic target and the frequency of hypoglycemia, should guide continuing education and protocol adjustments. Some published protocols are more successful than others, although there are no published randomized clinical trials to clearly establish the best. Computerization of protocols with alarms and reminders drastically reduces protocol violations and calculation error and facilitates documentation. Achieving TGC requires good protocols and reasonable targets, but effective implementation at the institutional level (reflected by consistent improvement in glycemic control) may be more important than having the best protocol in safely achieving the desired target range.