NEJM publishes arresting new data on NICE-SUGAR and hypoglycemia – September 24, 2012

Executive Highlights

  • New data in the NEJM shows association between tight glycemic control, hypoglycemia, and death.
  • New tools that would enable safer dosing of insulin should be sought and used.

The NEJM just published arresting new data on NICE SUGAR, the tight-glycemic-control in the ICU study that left so many so stunned in 2009 (see our summary from ICISEM in Belgium at, along with a cogent editorial by Dr. Irl Hirsch (University of Washington) titled “Understanding Low Sugar from NICE-SUGAR”. This new work shows an association between moderate and severe hypoglycemia and mortality. As the article notes, while these data don't prove a causal relationship, the association is nonetheless apparent. Of the patients who experienced moderate and severe hypoglycemia, 82% and 93%, respectively, were in the intensive-control group (this was based on follow-up data for 6,026 patients). Nearly half of the patients in the entire study experienced hypoglycemia, reinforcing the ongoing challenges of dosing insulin. The adjusted hazard ratios for death among these patients were 1.41 (moderate hypo) and 2.10 (severe hypo). Our takeaway thoughts:

First, better attention to safety in insulin dosing is critical. We hope that as the population with diabetes grows, reasonable glycemic targets should be used in the hospital, and tools that help with this should be sought, e.g., continuous glucose monitoring (CGM). Hirsch highlights in the editorial that although NICE SUGAR protocol called for patients to have their blood glucose checked approximately every hour, the reality was that the time between was much more – over two hours. I recall after NICE SUGAR hearing Dr. Hirsch ask this question. We’re very glad it has been answered, though the answer is quite dispiriting, since some of the hypoglycemia could have been prevented.

Second, and on a related note, patients should advocate for a safer environment in the hospital. Dr. Hirsch points out that unacceptable “nonchalant” attitudes still exist in some hospitals – these should be abandoned, as Dr. Hirsch stresses. We assume that many families would be surprised to see so many patients experiencing hypoglycemia and that many would say this is unacceptable even in the absence of CGM – more frequent and careful BGM would also help. We hope to see more frequent and systematic BGM and ultimately use of CGM (continuous glucose monitoring) in the hospital to improve overall patient safety and we plan to encourage patients’ families to make sure they are asking for this.

Third, long-term, we would like the impact of better glycemic control uncomplicated by hypoglycemia to be assessed – this would be of great interest to patients and providers and payers.

Let's hope the word on this study in NEJM is broadly communicated so that the prevalence of hypoglycemia is communicated and reduced. As we understand it, after NICE-SUGAR, many hospitals threw not only tight control out the window, but even “reasonable” glycemic controls. If hospitals could adhere to standards of the control group of NICE-SUGAR (140-180 mg/dL), they would be doing an excellent job, according to Dr. Hirsch. We still don't know the impact of good glycemic control without hypoglycemia on patients is but we would advocate for studying this with CGM (a NICE SUGAR with CGM) so that the confusion can be addressed. Ultimately, this NEJM piece shows that we still don't know the causes (versus associations) with mortality and that is something we would like to see studied again with better tools. In the meantime, we wouldn’t wish on any patient to be in the hospital without very close attention being paid to their glycemic health. The extent of the association between hypoglycemia and mortality in NICE SUGAR is truly troubling, especially given how preventable hypoglycemia is.

-- by Kelly Close