Does the Risk Associated With Hyperglycemia Occur Across All Patient Groups?
Abstract & Commentary
By Andrew M. Luks, MD, Pulmonary and Critical Care Medicine, University of Washington, Seattle. Dr. Luks reports no financial relationship to this field of study. This article originally appeared in the November 2009 issue of Critical Care Alert. It was edited by David J. Pierson, MD and peer reviewed by William Thompson, MD. Dr. Pierson is Professor, Pulmonary and Critical Care Medicine, Harborview Medical Center, University of Washington, Seattle, and Dr. Thompson is Staff Pulmonologist, VA Medical Center; Associate Professor of Medicine, University of Washington; they report no financial relationships relevant to this field of study.
Synopsis: This retrospective cohort study reaffirms the link between hyperglycemia and mortality in the ICU, but demonstrates that the risk does not apply equally to all patient groups and, instead, varies based on admission diagnosis.
Source: Falciglia M, et al. Hyperglycemia-related mortality in critically ill patients varies by admission diagnosis. Crit Care Med 2009;37:1-9.
After initial trials showed a mortality benefit from the use of insulin drips and tight glucose control in the ICU,1 subsequent studies have shown mixed results and, in some cases, increased mortality and episodes of hypoglycemia.2,3 Falciglia et al sought to determine whether the inconsistency in results across studies might be due to variability in the effects of hyperglycemia across different patient populations.
To answer this question, they conducted a retrospective cohort study using data from 259,040 patients admitted to 173 medical, cardiac, surgical, and mixed ICUs in more than 100 Veterans Health Administration hospitals over a three-year period from 2002 to 2005. The data used in their analysis were obtained from the VA Inpatient Evaluation Center, a national VA program that measures risk-adjusted outcomes in VA hospital ICUs and uses that information to guide evidence-based improvements in patient safety and quality of care. As part of this system, customized programs identify patients with ICU stays at each VA hospital and extract relevant data that can be used in subsequent analyses. Mean glucose was calculated for each patient from all values measured on chemistry panels during their ICU stay but did not include values obtained during point-of-care testing for monitoring insulin administration, as these values may be subject to greater measurement error. Patients were then stratified into five groups based on their mean glucose levels: 70-110, 111-145, 146-199, 200-300, and > 300 mg/dL. Two-level logistic regression analysis was then used to determine the relationship between hyperglycemia and mortality. In the first level of analysis, age, diagnosis, comorbidity, and laboratory variables were used to calculate a predicted mortality rate while, in the second level of analysis, the predicted mortality risk was combined with the mean glucose value for each patient to determine the association of hyperglycemia with hospital mortality.
The analysis, which included patients who were generally older (66% > 60 years of age) and overwhelmingly male (98%), demonstrated that hyperglycemia was associated with increased mortality regardless of illness severity, the type of ICU, length of time in the ICU, or whether the patient had diabetes. The odds ratio for mortality also increased as mean glucose values increased, varying from 1.31 in patients with mean glucose values of 111-145 mg/dL to 2.85 in patients whose mean values were > 300 mg/dL. The relationship between hyperglycemia and mortality varied based on the patients' admission diagnoses, with a strong relationship between hyperglycemia and mortality in some disorders (e.g., unstable angina, acute myocardial infarction, congestive heart failure, gastrointestinal bleeding, pneumonia, pulmonary embolism, and sepsis) and no relationship in other cases (e.g., COPD, hepatic failure, gastrointestinal malignancy, and patients admitted following surgery for coronary artery bypass grafting, peripheral vascular disease, and hip fracture). Of note, even within the group of disorders for which there was a relationship between hyperglycemia and mortality, the risk still varied between diagnoses.
Commentary
There are several interesting aspects of this study. The first is that, as the authors hypothesized, the relationship between hyperglycemia and mortality does, in fact, vary across patient populations. This may help explain why the benefits of insulin drips and tight glucose control, widely adopted in ICUs following a single-center study in a surgical patient population,1 have not been borne out in subsequent studies looking at broader or different patient populations. This suggests we may need to tailor management of hyperglycemia, and perhaps other problems in the ICU, a bit more and not apply a "one-size-fits-all" approach. Unfortunately, we lack the data at this point to help us adopt a more nuanced approach to the hyperglycemia issue and the likelihood of an adequately powered, prospective trial to help address these issues is low.
The second interesting, and perhaps more distressing, result of this study was that even moderately elevated mean glucose values over the course of an admission (111-145 or 146-149 mg/dL, for example) were associated with an increased risk of mortality across all of the patient populations, a result that puts ICU clinicians in a significant bind. The recent studies examining insulin drips and tight glucose control have largely shown that tight control (80-110 mg/dL) is associated with an increased risk of hypoglycemia and, in a recent, noteworthy trial, an increased risk of mortality.2 Given this, it would seem reasonable to liberalize protocols a bit and target more moderate levels of control such as 120-150 mg/dL and ensure patients are not markedly hyperglycemic. The study by Falciglia and colleagues, however, suggests that even this modest shift in target glucose levels may be bad for our patients.
How do we resolve this issue? In the long term, we need additional prospective studies comparing outcomes from insulin protocols targeting several ranges of glucose control. The study above was a retrospective analysis, and only prospective data will help clarify this issue. Until such a trial is completed, we should accept the risks of the mildly elevated levels and avoid overly tight targets. Hypoglycemia is a clear problem with a well-identified cause over which we have control in patients on insulin drips. Mortality is a more amorphous concept and many other factors which are often difficult to control may be contributing to this outcome in patients than just hyperglycemia.
References
1. Van den Berghe G, et al. Intensive insulin therapy in the critically ill patient. N Engl J Med 2001;345: 1359-1367.
2. NICE-SUGAR Study Investigators; et al. Intensive versus conventional glucose control in critically ill patients. N Engl J Med 2009;360:1283-1297.
3. Arabi YM, et al. Intensive versus conventional insulin therapy: A randomized controlled trial in medical and surgical critically ill patients. Crit Care Med 2008; 36:3190-3197.
This retrospective cohort study reaffirms the link between hyperglycemia and mortality in the ICU, but demonstrates that the risk does not apply equally to all patient groups and, instead, varies based on admission diagnosis.You have reached your article limit for the month. Subscribe now to access this article plus other member-only content.
- Award-winning Medical Content
- Latest Advances & Development in Medicine
- Unbiased Content