Toward Understanding Tight Glycemic Control in the ICU : A Systematic Review and Metaanalysis

Paul E. Marik and Jean-Charles Preiser Chest 2010;137;544-551; Prepublished online December 16, 2009; DOI 10.1378/chest.09-1737 The online version of this article, along with updated information and services can be found online on the World Wide Web at: http://chestjournal.chestpubs.org/content/137/3/544.full.html

Chest is the official journal of the American College of Chest Physicians. It has been published monthly since 1935. Copyright2010by the American College of Chest Physicians, 3300 Dundee Road, Northbrook, IL 60062. All rights reserved. No part of this article or PDF may be reproduced or distributed without the prior written permission of the copyright holder. (http://chestjournal.chestpubs.org/site/misc/reprints.xhtml) ISSN:0012-3692

Downloaded from chestjournal.chestpubs.org by guest on September 12, 2011 2010 American College of Chest Physicians

CHEST
A Systematic Review and Metaanalysis
Paul E. Marik, MD, FCCP; and Jean-Charles Preiser, MD

Original Research
CRITICAL CARE MEDICINE

Toward Understanding Tight Glycemic Control in the ICU

Background: Following publication of the Leuven Intensive Insulin Therapy Trial in 2001, tight glycemic control became the standard of care in ICUs around the world. Recent studies suggest that this approach may be awed. The goal of this systematic review was to determine the benets and risks of tight glycemic control in ICU patients and to explain the differences in outcomes among reported trials. Methods: Prospective, randomized controlled clinical trials (RCTs) that studied the impact of tight glycemic control (blood glucose 80-110 mg/dL) on mortality in ICU patients were identied through a search of MEDLINE, EMBASE, and the Cochrane Database of Systematic Reviews, and a citation review of relevant primary and review articles. Data were abstracted on study design, study size, and patient characteristics, as well as on the mean (or median) and SD of the ICU blood glucose level, mean daily dose of insulin administered, average daily caloric intake, percentage of calories given intravenously (parenteral nutrition), incidence of hypoglycemia, need for dialysis, and 28-day/hospital mortality. Metaanalytic techniques were used to analyze the data; subgroup analysis and metaregression were used to explain differences in the treatment effect. Results: We identied seven RCT studies that included 11,425 patients. Overall, tight glycemic control did not reduce the 28-day mortality (odds ratio [OR] 0.95; 95% CI, 0.87-1.05), the incidence of blood stream infections (OR 1.04; 95% CI, 0.93-1.17), or the requirement for renal replacement therapy (OR 1.01; 95% CI, 0.89-1.13). The incidence of hypoglycemia was signicantly higher in patients randomized to tight glycemic control (OR 7.7; 95% CI, 6.0-9.9; P , .001). Metaregression demonstrated a signicant relationship between the treatment effect (28-day mortality) and the proportion of calories provided parenterally (P 5 .005). This suggests that the difference in outcome between the two Leuven Intensive Insulin Therapy Trials and the subsequent trials could be related to the use of parenteral nutrition. When the two Leuven Intensive Insulin Therapy Trials were excluded from the metaanalysis, mortality was lower in the control patients (OR 0.90; 95% CI, 0.81-0.99; P 5 .04; I2 5 0%). Conclusions: There is no evidence to support the use of intensive insulin therapy in general medical-surgical ICU patients who are fed according to current guidelines. Tight glycemic control is associated with a high incidence of hypoglycemia and an increased risk of death in patients not receiving parenteral nutrition. CHEST 2010; 137(3):544551
Abbreviations: APACHE 5 Acute Physiology And Chronic Health Evaluation; IIT 5 intensive insulin therapy; NICESUGAR 5 Normoglycemia in Intensive Care EvaluationSurvival Using Glucose Algorithm Regulation; OR 5 odds ratio; RCT 5 randomized controlled trial

is common in ill and Stress hyperglycemiaPrior to 2001,critically hyperstress injured patients.

glycemia was dened as a plasma glucose above 180 to 200 mg/dL.2,3 However, following the Leuven Intensive Insulin Therapy Trial, presented as a proof-of-concept study, the American Association of Clinical Endocrinologists, the American Diabetes

Association, and other authorities suggested that stress hyperglycemia be considered in any critically ill patient with a blood glucose in excess of 110 mg/dL.4-7 In the Leuven Intensive Insulin Therapy Trial, 12% of patients had a baseline blood glucose above 200 mg/dL on the day following ICU admission. However, 74.5% of patients had a baseline blood glucose above
Original Research

544

Downloaded from chestjournal.chestpubs.org by guest on September 12, 2011 2010 American College of Chest Physicians

110 mg/dL, with 97.5% having a recorded blood glucose level above 110 mg/dL sometime during their ICU stay.4 Excessive counter-regulatory hormones, such as glucagon, growth hormone, catecholamines, and glucocorticoids, as well as cytokines, such as interleukin-1, interleukin-6, and tumor necrosis factor-a, result in increased gluconeogenesis and insulin resistance, which are the major factors leading to stress hyperglycemia.2,3,8 In addition, exogenous catecholamines, intravenous dextrose, and nutritional support compound the problem. Until recently, stress hyperglycemia was considered an adaptive response, providing a ready source of fuel during a time of increased demand. However, both short-term and long-term hyperglycemia are now recognized as having signicant deleterious effects. Hyperglycemia increases oxidative injury, potentiates the proinammatory response, promotes clotting, causes abnormal vascular reactivity, and impairs leukocyte and mononuclear cell immune responsiveness.2,3 Stress hyperglycemia is believed to be associated with a worse outcome following acute myocardial infarction, stroke, and cardiac surgery, and in patients with congestive heart failure.9-13 In 2001, Van den Berghe and coworkers4 published a landmark study (the Leuven Intensive Insulin Therapy Trial), in which they demonstrated that tight glycemic control using intensive insulin therapy (IIT) improved the outcome of critically ill surgical patients. Following this study, tight glycemic control was rapidly adopted as the standard of care in ICUs throughout the world and it was endorsed by the Institute for Health Care Improvement and other national organizations in the United States.5-7 In 2006, Van den Berghe and colleagues14 repeated this study design in medical ICU patients. Although failing to reproduce the improvement in survival in the entire set of patients, this study demonstrated a reduction in morbidity in the patients randomized to the tight glycemic group, with a reduction in mortality in the subset of patients with an ICU stay of 3 days or more. Recently Normoglycemia in Intensive Care EvaluationSurvival Using Glucose Algorithm Regulation (NICE-SUGAR), a large, multicenter,
Manuscript received July 22, 2009; revision accepted November 13, 2009. Afliations: From the Division of Pulmonary and Critical Care Medicine (Dr Marik), Eastern Virginia Medical School, Norfolk, VA; and the Department of General Intensive Care (Dr Preiser), University of Liege, Liege, Belgium. Correspondence to: Paul Marik, MD, FCCP, Pulmonary and Critical Care Medicine, Eastern Virginia Medical School, Internal Medicine, HH 410, 825 Fairfax Ave, Norfolk, VA 23507; e-mail: [email protected] 2010 American College of Chest Physicians. Reproduction of this article is prohibited without written permission from the American College of Chest Physicians (www.chestpubs.org/site/ misc/reprints.xhtml). DOI: 10.1378/chest.09-1737
www.chestpubs.org

randomized controlled study, was published,15 which was unable to conrm the ndings of Van den Berghe et al.14 Indeed, this study demonstrated a 2.6% absolute increase in 90-day mortality in patients randomized to tight glucose control (P 5 .02). The aim of this systematic review was to determine the mortality benet of tight glycemic control in the ICU setting and to explain differences in the treatment effect among studies by metaregression. Unlike previous metaanalyses,16,17 this study only included randomized controlled trials (RCTs) that specically attempted to conrm the benets of IIT (blood glucose of 80-110 mg/dL) as compared with less strict glycemic control in ICU patients. Materials and Methods
Identication of Trials Our aim was to identify all relevant RCTs that compared the mortality of ICU patients randomized to an IIT protocol aimed to achieve tight glycemic control (glucose of 80-110 mg/dL) with those randomized to a control arm that received less strict glucose control. The primary outcome measure was 28-day (or hospital) mortality. Secondary outcome measures included the need for dialysis, acquired blood stream infections, and the incidence of hypoglycemia (dened as a blood glucose of , 40 mg/dL). We used a multimethod approach to identify relevant studies for this review. The National Library of Medicines MEDLINE database was searched for relevant studies published from 2001 to August 2009 using the following medical subject headings and keywords: intensive care unit, critical care, critical illness, blood glucose, and RCT (publication type). In addition, we searched EMBASE and the Cochrane Database of Systematic Reviews. Bibliographies of all selected articles and review articles that included information on tight glycemic control were reviewed. We excluded studies that used a glucose-insulin-potassium infarct-size-limiting protocol, studies that were performed outside the ICU, and pilot studies that randomized fewer than 100 patients. We performed this metaanalysis according to the guidelines proposed by the QUOROM group.18 Data Extraction and Quality Assessment Data from all eligible studies were analyzed using a standardized form. Both authors independently abstracted data on study design, study size, study setting, patient characteristics, mean Acute Physiology And Chronic Health Evaluation (APACHE) II score, mean/median daily glucose level, mean daily insulin dose administered, mean daily caloric intake, and percentage of calories given intravenously during the ICU stay, as well as the percentage of patients who were diabetic or septic (on admission). The SD of the mean glucose level during the ICU stay was used as an index of glucose variability.19-21 Missing data (unpublished) were obtained from the authors of the primary studies. Data were recorded by intention to treat. The two reviewers independently assessed allocation concealment and the likelihood of bias to determine the methodologic quality of the included trials. The allocation concealment was ranked as adequate, uncertain, or clearly inadequate, and the likelihood of bias was scored on the Jadad ve-point scale,22 which contained two questions each on randomization and masking and one question on the reporting of dropouts and withdrawals. Any disagreement between reviewers was resolved by consensus.
CHEST / 137 / 3 / MARCH, 2010

545

Downloaded from chestjournal.chestpubs.org by guest on September 12, 2011 2010 American College of Chest Physicians

Statistical Analysis Statistical analysis was performed using Comprehensive Meta-analysis 2.0 (Biostat; Englewood, NJ). We assessed heterogeneity among studies using the Cochran Q statistic,23 with P .10 indicating signicant heterogeneity,24 and I2 with suggested thresholds for low (25%-49%), moderate (50%-74%), and high (. 75%) values.25,26 We used a random effects model if the Q statistic was signicant; otherwise, we used a xed effects model. Summary effects estimates were presented as odds ratio (OR) with 95% CI; we considered P .05 (two-sided) as signicant. We carried out subgroup analysis (based on the results of the metaregression) and assessed the presence of publication bias visually with a funnel plot. A priori, the relationship between the treatment effect and the following variables was assessed by metaregression: APACHE II score, mean daily glucose level, SD of the mean glucose level (as an index of glucose variability), mean daily insulin dose administered, mean daily caloric intake, and percentage of calories given intravenously, as well as the percentage of patients who were diabetic or septic.

(published in abstract form) was identied from a review of articles identied by the initial search strategy30; this study has subsequently been published in full.31 In all, seven studies met the inclusion criteria; these included the two Leuven Intensive Insulin Therapy Trials published by Van den Berghe et al,4,14 as well as the NICE-SUGAR trial.15,27-29,31 In all, 11,425 patients were included in the seven studies. The study size varied from 504 to 6,022 analyzable patients. Tables 1 and 2 show the characteristics of the included studies. All the studies were of high quality (Jadad score 3), used random allocation, and clearly reported allocation of concealment. In none of the studies were the treating clinicians blinded to treatment allocation; however, the study investigators/ assessors were blinded to treatment and outcome. Intention-to-treat analysis was reported in all studies with follow-up of all patients being reported. Outcomes of Metaanalysis The mean blood glucose level was 112 mg/dL in the IIT patients as compared with 151 mg/dL in the control group, the mean difference in glucose being 39 mg/dL. Similarly, the mean daily insulin dose was 54 units in the IIT patients as compared with 16 U in the control group, the mean difference in daily insulin dose being 37 units. Overall, IIT did not reduce the 28-day mortality (OR 0.95; 95% CI, 0.87-1.05) (Fig 2), the incidence of blood stream infections (OR 1.04; 95% CI, 0.93-1.17), or the requirement for renal replacement therapy (OR 1.01; 95% CI, 0.89-1.13). The incidence of hypoglycemia was signicantly higher in patients randomized to IIT (OR 7.7; 95% CI, 6.09.9; P , .001). Visual inspection of the funnel plot failed to reveal publication bias. As is evident from Figure 2, there was a difference in the treatment effect (28-day mortality) between the Leuven Intensive Insulin Therapy Trials and the subsequent trials. This difference was explored by metaregression using the predened variables. The use of intravenous calories (total parenteral nutrition) was the only variable that was signicant, with this variable largely explaining the difference in primary outcome among the trials (P 5 .005, Fig 3) Figure 3 demonstrates a positive treatment effect (reduced mortality) associated with IIT when a large percentage of the calories was provided intravenously, whereas it demonstrates a negative treatment effect (increased mortality) associated with IIT when a low percentage of calories was provided intravenously (ie, enteral nutrition). Neither the mean daily insulin dose administered nor the mean daily caloric intake was associated with 28-day/hospital mortality. When the two trials that provided a high percentage of calories
Original Research

Results Trials Included Figure 1 shows details of study identication, inclusion, and exclusion. Our search strategy initially yielded 62 citations. Of these, six unique studies met our inclusion criteria.4,14,15,27-29 One additional study

Figure 1. Flow of study identication, inclusion, and exclusion. Pt 5 patient.

Downloaded from chestjournal.chestpubs.org by guest on September 12, 2011 2010 American College of Chest Physicians

Table 1Quality Assessment of Included Randomized Trials on Intensive Insulin Therapy in the ICU
Study/Year Van den Berghe /2001 Van den Berghe14/2006 Glucontrol31/2007 VISEP27/2008 de la Rosa28/2008 Arabi29/2008 NICE-SUGAR15/2009
4

Clinician Blinding No No No No No No No

Intention to Treat Yes Yes Yes Yes Yes Yes Yes

Allocation Concealment Adequate Adequate Adequate Adequate Adequate Adequate Adequate

Jadad Score (0-5) 3 3 3 3 3 3 3

NICE-SUGAR 5 Normoglycemia in Intensive Care EvaluationSurvival Using Glucose Algorithm Regulation; VISEP 5 efcacy of Volume substitution and Insulin therapy in severe SEPsis.

intravenously (the Leuven Intensive Insulin Therapy Trials) were excluded from the metaanalysis, mortality was lower in the control patients (OR 0.90; 95% CI, 0.81-0.99; P 5 .04; I2 5 0%). These data suggest that IIT is harmful in patients receiving enteral nutrition.

Discussion The NICE-SUGAR study, as well as four additional RCTs, were unable to replicate the ndings of the two Leuven Intensive Insulin Therapy Trials and, indeed, raised the possibility that tight glycemic control may increase organ failure and death in patients fed according to current guidelines. It seems unlikely that the difference in outcome between these trials was due to a failure to reach the target blood glucose levels, because the mean daily glucose and insulin use was similar among studies. Furthermore, it is unlikely that this outcome can be explained by chance alone.32 Rather, as demonstrated by our analysis, we believe that the explanation for these disparate ndings lies with the high rate of use of total parenteral nutrition in the Van den Berghe studies.4,14 Our data suggest that IIT may be harmful in patients receiving enteral nutrition. IIT, however, appears to improve the outcome of patients receiving the majority of their carbohydrate load parenterally. The benet of IIT in patients receiving a high parenteral glucose load is supported by both experimental and clinical studies, which have demonstrated that excessive parenteral glucose in the absence of IIT leads to hyperglycemia and increased cellular glucose uptake, which, in turn, is associated with organ failure and death.33-35 The increased mortality in enterally fed patients receiving IIT may be related to the adverse effects associated with hypoglycemia. Hyperglycemia and insulin resistance are common in critically ill patients receiving parenteral nutrition and this may be associated with increased mortality. In a metaanalysis that compared enteral with parenteral nutrition in postoperative patients, Moore and
www.chestpubs.org

colleagues36 reported the mean blood glucose to be 130 mg/dL with enteral as compared with 224 mg/dL with parenteral nutrition (a difference of 94 mg/dL). In the Veterans Affairs Total Parenteral Nutrition Cooperative Study, severe hyperglycemia (glucose . 300 mg/dL) occurred in 20% of the patients receiving parenteral nutrition as compared with 1.5% in the control group.37 van der Voort and colleagues38 have demonstrated that both the ICU and hospital mortality of critically ill patients were independently related to the mean amount of infused glucose. In a retrospective analysis of 111 hospitalized patients receiving parenteral nutrition, Cheung and coworkers33 reported that hyperglycemia was independently associated with an increased risk of cardiac complications, sepsis, acute renal failure, and death. In the study by Cheung and colleagues,33 the mortality of subjects with a blood glucose in the highest quartile was 10.9 times that of subjects in the lowest quartile. Ahrens and colleagues39 demonstrated that the administration of a low-calorie parenteral nutrition formula containing a reduced amount of glucose resulted in fewer, and less severe, hyperglycemic events and lower insulin requirements. It would therefore appear counterintuitive to administer large amounts of intravenous glucose to patients at risk of developing stress hyperglycemia. A number of mechanisms lead to hyperglycemia and insulin resistance in patients receiving parenteral nutrition.40 The most obvious is the higher total glucose (carbohydrate) load; a typical 1,800 nonprotein calorie parenteral formula contains about 350 g of glucose, as compared with 230 g in a standard enteral formula. It is well known that oral glucose administration is associated with a much greater increase in the secretion of insulin, with higher plasma insulin levels, when compared with the same amount of glucose given intravenously.41,42 The concept that factors produced by the intestinal mucosa in response to nutrient ingestion are capable of stimulating the release of substances from the endocrine pancreas and thereby reducing blood glucose levels was rst introduced in the early 1900s.43 This
CHEST / 137 / 3 / MARCH, 2010

547

Downloaded from chestjournal.chestpubs.org by guest on September 12, 2011 2010 American College of Chest Physicians

I-mortality

phenomenon has been dubbed the incretin effect, and it is estimated to account for approximately 50% to 70% of the total insulin secreted after oral glucose administration.44,45 Two incretins (ie, glucosedependent insulinotropic polypeptide and glucagonlike peptide-1), as well as their receptors, have been identied, puried, and sequenced and their gene structure determined.44,45 Activation of both incretin receptors on the b cell of the pancreas leads to a rapid increase in the levels of cyclic adenosine monophosphate and intracellular calcium, followed by insulin exocytosis, in a glucose-dependent manner. Glucagon-like peptide-1 also inhibits glucagon secretion and enhances glucose disposal through neural mechanisms.44,45 Independent of the incretin effect, hepatic glucose output is affected by the portal venous glucose concentration following enteral feeding. An elegant series of experiments46-52 demonstrated that the arterial-portal glucose gradient alters net hepatic glucose balance, such that intraportal glucose administration suppresses hepatic glucose output, whereas peripheral glucose administration increases hepatic glucose output. By these mechanisms, parenteral, as opposed to enteral, nutrition results in signicantly higher blood glucose levels with insulin resistance. This results in cellular glucose overload and oxidative injury in those cells that do not require insulin for glucose uptake. By increasing glucose-4-mediated glucose uptake (muscle and adipose tissue) and decreasing blood glucose levels, insulin decreases glucose uptake in non-glucose transporter-4 cells, thereby decreasing oxidative injury in these cells. This phenomenon has been conrmed by Vanhorebeek and colleagues.34 The results of our metaanalysis and metaregression have demonstrated that tight glycemic control may only be of benet in patients receiving parenteral nutrition, particularly those with a low severity of illness. Tight glycemic control is, however, associated with signicant hazards. Severe hypoglycemia (blood glucose , 40 mg/dL) occurs in up to 18% of treated patients, with these patients being at an increased risk of death.14 Using cerebral microdialysis in patients with traumatic brain injuries, Oddo and colleagues53 demonstrated that tight glycemic control was associated with a greater risk of brain energy crisis and death. These data suggest that tight glycemic control may result in hypoglycemia and neuroglycopenia at a time of increased cerebral metabolic demand. Most ICUs in the world use bedside capillary blood glucose measurements to monitor insulin infusion regimens. Multiple studies have shown that these devices are inaccurate and tend to overestimate true plasma glucose levels.54,55 This further increases the risk of hypoglycemia in ICU patients.
Original Research

7.2 29.9 24.7 18.7 36.6 16.9 22.3 Pts 5 patients; APII 5 APACHE II score; C 5 control; cal 5 calories; DM 5 diabetic; glu 5 glucose; I 5 intervention (IIT); NA 5 not available. See Table 1 for expansion of other abbreviations. Percentage of calories given intravenously. bMean morning glucose. cMean daily insulin. d28-d/hospital mortality. e63% cardiac surgery. fMedian morning glucose. gMean daily glucose.
a

Table 2Summary of Included Randomized Controlled Trials Testing the Benet of Intensive Insulin Therapy in ICU Patients

C-mortalityd I-insulin C-insulinc I-glu IV-cala DM (%) Septic (%) APII No. Pts ICU Study/Year C-glub

548

Van den Berghe4/2001 Van den Berghe14/2006 Glucontrol31/2007 VISEP27/2008 de la Rosa28/2008 Arabi29/2008 NICE-SUGAR15/2009

Surgicale MICU Mixed Mixed Mixed Mixed Mixed

1,548 1,200 1,078 537 504 523 6,022

9 23 15 20 16 23 21

6 50 NA 100 32 23 22

13 17 19 30.0 12 40 20

87.0 87.0 27.0 66.0 7.0 7.9 29.5

153 6 33 153 6 31 139f 6 34 151 6 33g 155 6 49 171 6 34g 145 6 26

103 6 19 111 6 29 110f 6 36 112 6 18g 124 6 47 115 6 18g 118 6 25

33 10 10 5 12 31 17

71 59 43 32 52 71 50

10.9 30.1 25.9 15.3 32.4 13.6 20.8

Downloaded from chestjournal.chestpubs.org by guest on September 12, 2011 2010 American College of Chest Physicians

Figure 2. Effect of intensive insulin therapy (IIT) on 28-day mortality. a-TPN 5 parenteral nutrition; b-ENT 5 enteral nutrition.

It should be appreciated that the results of our metaanalysis are greatly inuenced by the NICESUGAR trial, which enrolled more patients than all the other studies combined. Nevertheless, the treatment effect was remarkably similar among NICE-SUGAR and the four non-Van den Berghe studies. In addition, the Efcacy of Volume Substitution and Insulin Therapy in Severe Sepsis (VISEP) study tested ITT and pentastarch in a 2 3 2 factorial design.27 However, there were no signicant interactions between the two study interventions with respect to the rate of death at 28 days. It is likely that additional factors may explain the differing outcomes between the Leuven Intensive Insulin studies and the subsequent studies. Most notably, there were differences among trials in achieving the tight control target in the intervention group, with some studies not achieving their goal.31 Furthermore, there were variations in the difference between the mean blood glucose between the control and intervention groups. For example, the difference between groups was 50 mg/dL in the rst Van Den Berghe study4 compared with 27 mg/dL in NICE-SUGAR.15 Most of the trials used the mean morning blood glucose to assess glucose control; however, this variable may be a poor representation of the actual blood glucose control achieved.56 Blood glucose levels in critically ill patients vary markedly, even when using continuous feeding and insulin infusions. Variability of glycemia is expressed as the magnitude of the uctuation in glycemic control.57 Even in the presence of the same mean value, the nature of glycemic control can be quite different in terms of variability. It has been well demonstrated in both healthy controls and in patients with diabetes that oscillating blood glucose is associated with greater oxidative injury than sustained hyperglycemia.58,59 A number of reports indicate that glucose variability
www.chestpubs.org

may be an independent predictor of outcome in critically ill patients.19-21 Furthermore, it has been suggested that the difference in glucose variability (as measured by the SD) between the two arms of the rst Leuven IIT Study may have accounted for its positive outcome.57 In our metaregression, there was no statistically signicant relationship between the difference in the SDs of the mean blood glucose values (glucose variability) between the treatment arms and study outcome (mortality). It is important to emphasize that the studies included in our metaanalysis compared tight (blood glucose of 80-110 mg/dL) to less stringent glycemic control, and that therefore the results do not necessarily apply to more moderate approaches to glycemic control (blood glucose targets . 110 mg/dL and , 180 mg/dL). A joint statement recently released by the American Association of Clinical Endocrinologists and the American Diabetes Association suggests that

Figure 3. Regression of the percentage of calories given parenterally while in the ICU vs the log odds ratio of the mortality benet of intensive insulin therapy. A log odds ratio above zero indicates a mortality benet from intensive insulin therapy, whereas a log odds ratio below zero indicates harm. TPN 5 total parenteral nutrition.
CHEST / 137 / 3 / MARCH, 2010

549

Downloaded from chestjournal.chestpubs.org by guest on September 12, 2011 2010 American College of Chest Physicians

it would be a serious error to conclude that judicious control of glycemia in critically ill patients, and in non-ICU patients in general, is not warranted.60 Furthermore, they suggest that perhaps major benecial effects on outcomes can be derived from a higher target range of glucose than 80-110 mg/dL in comparison with uncontrolled hyperglycemia, recommending that the blood glucose level be maintained between 140 and 180 mg/dL. However, it is noteworthy that the critical inuence of parenteral nutrition on the outcome of IIT was not addressed in this statement. Conclusions In conclusion, we believe that the variation in intravenous glucose load may explain the conicting outcomes between the Leuven Intensive Insulin Therapy studies and the subsequent conrmatory studies. Furthermore, our metaanalysis suggests that there are no data to support the use of IIT in patients who are fed enterally. Acknowledgments
Author contributions: Dr Marik: was responsible for performing the metaanalysis, interpreting the data, and writing the manuscript. Dr Preiser: was responsible for performing the metaanalysis, interpreting the data, and writing the manuscript. Financial/nonnancial disclosures: The authors have reported to CHEST that no potential conicts of interest exist with any companies/organizations whose products or services may be discussed in this article.

References
1. Dungan KM, Braithwaite SS, Preiser JC. Stress hyperglycaemia. Lancet. 2009;373(9677):1798-1807. 2. McCowen KC, Malhotra A, Bistrian BR. Stress-induced hyperglycemia. Crit Care Clin. 2001;17(1):107-124. 3. Marik PE, Raghavan M. Stress-hyperglycemia, insulin and immunomodulation in sepsis. Intensive Care Med. 2004; 30(5):748-756. 4. Van den Berghe G, Wouters P, Weekers F, et al. Intensive insulin therapy in the critically ill patients. N Engl J Med. 2001;345(19):1359-1367. 5. Garber AJ, Moghissi ES, Bransome ED Jr, et al; American College of Endocrinology Task Force on Inpatient Diabetes Metabolic Control. American College of Endocrinology position statement on inpatient diabetes and metabolic control. Endocr Pract. 2004;10(1):77-82. 6. American Diabetes Association. Standards of medical care in diabetes2008. Diabetes Care. 2008;31(Suppl 1):S12-S54. 7. Institute for Health Care Improvement. Establish a glycemic control policy in your ICU. http://www.ihi.org/IHI/Topics/ CriticalCare/IntensiveCare/Changes/IndividualChanges/ EstablishaGlycemicControlPolicyinYourICU.htm. Accessed February 20, 2009. 8. Mesotten D, Delhanty PJ, Vanderhoydonc F, et al. Regulation of insulin-like growth factor binding protein-1 during protracted critical illness. J Clin Endocrinol Metab. 2002; 87(12):5516-5523.
550

9. Norhammar AM, Rydn L, Malmberg K. Admission plasma glucose. Independent risk factor for long-term prognosis after myocardial infarction even in nondiabetic patients. Diabetes Care. 1999;22(11):1827-1831. 10. Malmberg K, Norhammar A, Wedel H, Rydn L. Glycometabolic state at admission: important risk marker of mortality in conventionally treated patients with diabetes mellitus and acute myocardial infarction: long-term results from the Diabetes and Insulin-Glucose Infusion in Acute Myocardial Infarction (DIGAMI) study. Circulation. 1999;99(20):2626-2632. 11. Capes SE, Hunt D, Malmberg K, Pathak P, Gerstein HC. Stress hyperglycemia and prognosis of stroke in nondiabetic and diabetic patients: a systematic overview. Stroke. 2001;32(10):2426-2432. 12. Barsheshet A, Garty M, Grossman E, et al. Admission blood glucose level and mortality among hospitalized nondiabetic patients with heart failure. Arch Intern Med. 2006;166(15): 1613-1619. 13. Furnary AP, Gao G, Grunkemeier GL, et al. Continuous insulin infusion reduces mortality in patients with diabetes undergoing coronary artery bypass grafting. J Thorac Cardiovasc Surg. 2003;125(5):1007-1021. 14. Van den Berghe G, Wilmer A, Hermans G, et al. Intensive insulin therapy in the medical ICU. N Engl J Med. 2006; 354(5):449-461. 15. Finfer S, Chittock DR, Su SY, et al; NICE-SUGAR Study Investigators. Intensive versus conventional glucose control in critically ill patients. N Engl J Med. 2009;360(13):1283-1297. 16. Wiener RS, Wiener DC, Larson RJ. Benets and risks of tight glucose control in critically ill adults: a meta-analysis. JAMA. 2008;300(8):933-944. 17. Griesdale DE, de Souza RJ, van Dam RM, et al. Intensive insulin therapy and mortality among critically ill patients: a meta-analysis including NICE-SUGAR study data. CMAJ. 2009;180(8):821-827. 18. Moher D, Cook DJ, Eastwood S, Olkin I, Rennie D, Stroup DF. Improving the quality of reports of meta-analyses of randomised controlled trials: the QUOROM statement. Quality of reporting of meta-analyses. Lancet. 1999;354(9193): 1896-1900. 19. Egi M, Bellomo R, Stachowski E, French CJ, Hart G. Variability of blood glucose concentration and short-term mortality in critically ill patients. Anesthesiology. 2006;105(2): 244-252. 20. Ali NA, OBrien JM Jr, Dungan K, et al. Glucose variability and mortality in patients with sepsis. Crit Care Med. 2008;36(8):2316-2321. 21. Dossett LA, Cao H, Mowery NT, Dortch MJ, Morris JM Jr, May AK. Blood glucose variability is associated with mortality in the surgical intensive care unit. Am Surg. 2008;74(8): 679-685, discussion 685. 22. Jadad AR, Moore RA, Carroll D, et al. Assessing the quality of reports of randomized clinical trials: is blinding necessary? Control Clin Trials. 1996;17(1):1-12. 23. Cochran W. The combination of estimates from different experiments. Biometrics. 1954;10:101-129. 24. Berlin JA, Laird NM, Sacks HS, Chalmers TC. A comparison of statistical methods for combining event rates from clinical trials. Stat Med. 1989;8(2):141-151. 25. Higgins JP, Thompson SG. Quantifying heterogeneity in a meta-analysis. Stat Med. 2002;21(11):1539-1558. 26. Higgins JP, Thompson SG, Deeks JJ, Altman DG. Measuring inconsistency in meta-analyses. BMJ. 2003;327(7414):557-560. 27. Brunkhorst FM, Engel C, Bloos F, et al; German Competence Network Sepsis (SepNet). Intensive insulin therapy and pentastarch resuscitation in severe sepsis. N Engl J Med. 2008;358(2):125-139.
Original Research

Downloaded from chestjournal.chestpubs.org by guest on September 12, 2011 2010 American College of Chest Physicians

28. De La Rosa Gdel C, Donado JH, Restrepo AH, et al; Grupo de Investigacion en Cuidado intensivo: GICI-HPTU. Strict glycaemic control in patients hospitalised in a mixed medical and surgical intensive care unit: a randomised clinical trial. Crit Care. 2008;12(5):R120. 29. Arabi YM, Dabbagh OC, Tamim HM, et al. Intensive versus conventional insulin therapy: a randomized controlled trial in medical and surgical critically ill patients. Crit Care Med. 2008;36(12):3190-3197. 30. Devos P, Preiser JC, Melot C. Impact of tight glucose control by intensive insulin therapy on ICU mortality and the rate of hypoglycaemmia: nal results of Glucontrol [European Society of Intensive Care Medicine 20th Annual Congress abstract 0735]. Intensive Care Med. 2007;33(suppl 2):S189. 31. Preiser JC, Devos P, Ruiz-Santana S, et al. A prospective randomised multi-centre controlled trial on tight glucose control by intensive insulin therapy in adult intensive care units: the Glucontrol study. Intensive Care Med. 2009;35(10):1738-1748. 32. Finfer S, Delaney A. Tight glycemic control in critically ill adults. JAMA. 2008;300(8):963-965. 33. Cheung NW, Napier B, Zaccaria C, Fletcher JP. Hyperglycemia is associated with adverse outcomes in patients receiving total parenteral nutrition. Diabetes Care. 2005;28(10): 2367-2371. 34. Vanhorebeek I, De Vos R, Mesotten D, Wouters PJ, De Wolf-Peeters C, Van den Berghe G. Protection of hepatocyte mitochondrial ultrastructure and function by strict blood glucose control with insulin in critically ill patients. Lancet. 2005;365(9453):53-59. 35. Marik PE, Pinsky M. Death by parenteral nutrition. Intensive Care Med. 2003;29(6):867-869. 36. Moore FA, Feliciano DV, Andrassy RJ, et al. Early enteral feeding, compared with parenteral, reduces postoperative septic complications. The results of a meta-analysis. Ann Surg. 1992;216(2):172-183. 37. The Veterans Affairs Total Parenteral Nutrition Cooperative Study Group. Perioperative total parenteral nutrition in surgical patients. N Engl J Med. 1991;325(8):525-532. 38. van der Voort PH, Feenstra RA, Bakker AJ, Heide L, Boerma EC, van der Horst IC. Intravenous glucose intake independently related to intensive care unit and hospital mortality: an argument for glucose toxicity in critically ill patients. Clin Endocrinol (Oxf). 2006;64(2):141-145. 39. Ahrens CL, Barletta JF, Kanji S, et al. Effect of low-calorie parenteral nutrition on the incidence and severity of hyperglycemia in surgical patients: a randomized, controlled trial. Crit Care Med. 2005;33(11):2507-2512. 40. OKeefe SJ, Lee RB, Anderson FP, et al. Physiological effects of enteral and parenteral feeding on pancreaticobiliary secretion in humans. Am J Physiol Gastrointest Liver Physiol. 2003;284(1):G27-G36. 41. McIntyre N, Holdsworth CD, Turner DS. New interpretation of oral glucose tolerance. Lancet. 1964;284(7349):20-21. 42. Elrick HL, Stimmler L, Hlad CJ Jr, Arai Y. Plasma insulin response to oral and intravenous glucose administration. J Clin Endocrinol Metab. 1964;24:1076-1082. 43. Bayliss WM, Starling EH. On the causation of the so-called peripheral reex secretion of the panceas. Proc Soc Exp Biol Med. 1902;69:352-353.

44. Baggio LL, Drucker DJ. Biology of incretins: GLP-1 and GIP. Gastroenterology. 2007;132(6):2131-2157. 45. Drucker DJ, Nauck MA. The incretin system: glucagon-like peptide-1 receptor agonists and dipeptidyl peptidase-4 inhibitors in type 2 diabetes. Lancet. 2006;368(9548):1696-1705. 46. McGuinness OP, Fugiwara T, Murrell S, et al. Impact of chronic stress hormone infusion on hepatic carbohydrate metabolism in the conscious dog. Am J Physiol. 1993;265 (2 Pt 1):E314-E322. 47. Pagliassotti MJ, Myers SR, Moore MC, Neal DW, Cherrington AD. Magnitude of negative arterial-portal glucose gradient alters net hepatic glucose balance in conscious dogs. Diabetes. 1991;40(12):1659-1668. 48. Cherrington AD, Pagliassotti MJ, Myers SR, Adkins-Marshall B, McGuinness OP. Factors which regulate net hepatic glucose uptake in vivo. JPEN J Parenter Enteral Nutr. 1991;15(3): 71S-73S. 49. Myers SR, McGuinness OP, Neal DW, Cherrington AD. Intraportal glucose delivery alters the relationship between net hepatic glucose uptake and the insulin concentration. J Clin Invest. 1991;87(3):930-939. 50. Adkins-Marshall B, Pagliassotti MJ, Asher JR, et al. Role of hepatic nerves in response of liver to intraportal glucose delivery in dogs. Am J Physiol. 1992;262(5 Pt 1):E679-E686. 51. Adkins BA, Myers SR, Hendrick GK, Stevenson RW, Williams PE, Cherrington AD. Importance of the route of intravenous glucose delivery to hepatic glucose balance in the conscious dog. J Clin Invest. 1987;79(2):557-565. 52. Pagliassotti MJ, Tarumi C, Neal DW, Cherrington AD. Hepatic denervation alters the disposition of an enteral glucose load in conscious dogs. J Nutr. 1991;121(8):1255-1261. 53. Oddo M, Schmidt JM, Carrera E, et al. Impact of tight glycemic control on cerebral glucose metabolism after severe brain injury: a microdialysis study. Crit Care Med. 2008;36(12):3233-3238. 54. Critchell CD, Savarese V, Callahan A, Aboud C, Jabbour S, Marik P. Accuracy of bedside capillary blood glucose measurements in critically ill patients. Intensive Care Med. 2007;33(12):2079-2084. 55. Kanji S, Bufe J, Hutton B, et al. Reliability of point-of-care testing for glucose measurement in critically ill adults. Crit Care Med. 2005;33(12):2778-2785. 56. Braithwaite SS. Patient-level glucose reporting: averages, episodes, or something in between? Crit Care. 2008;12(2):133. 57. Egi M, Bellomo R, Reade MC. Is reducing variability of blood glucose the real but hidden target of intensive insulin therapy? Crit Care. 2009;13(2):302. 58. Ceriello A, Esposito K, Piconi L, et al. Oscillating glucose is more deleterious to endothelial function and oxidative stress than mean glucose in normal and type 2 diabetic patients. Diabetes. 2008;57(5):1349-1354. 59. Colette C, Monnier L. Acute glucose uctuations and chronic sustained hyperglycemia as risk factors for cardiovascular diseases in patients with type 2 diabetes. Horm Metab Res. 2007;39(9):683-686. 60. Moghissi ES, Korytkowski MT, DiNardo M, et al; American Association of Clinical Endocrinologists; American Diabetes Association. American Association of Clinical Endocrinologists and American Diabetes Association consensus statement on inpatient glycemic control. Diabetes Care. 2009;32(6):1119-1131.

www.chestpubs.org

CHEST / 137 / 3 / MARCH, 2010

551

Downloaded from chestjournal.chestpubs.org by guest on September 12, 2011 2010 American College of Chest Physicians

Toward Understanding Tight Glycemic Control in the ICU : A Systematic Review and Metaanalysis Paul E. Marik and Jean-Charles Preiser Chest 2010;137; 544-551; Prepublished online December 16, 2009; DOI 10.1378/chest.09-1737 This information is current as of September 12, 2011
Updated Information & Services Updated Information and services can be found at: http://chestjournal.chestpubs.org/content/137/3/544.full.html References This article cites 59 articles, 19 of which can be accessed free at: http://chestjournal.chestpubs.org/content/137/3/544.full.html#ref-list-1 Cited Bys This article has been cited by 6 HighWire-hosted articles: http://chestjournal.chestpubs.org/content/137/3/544.full.html#related-urls Permissions & Licensing Information about reproducing this article in parts (figures, tables) or in its entirety can be found online at: http://www.chestpubs.org/site/misc/reprints.xhtml Reprints Information about ordering reprints can be found online: http://www.chestpubs.org/site/misc/reprints.xhtml Citation Alerts Receive free e-mail alerts when new articles cite this article. To sign up, select the "Services" link to the right of the online article. Images in PowerPoint format Figures that appear in CHEST articles can be downloaded for teaching purposes in PowerPoint slide format. See any online figure for directions.

Downloaded from chestjournal.chestpubs.org by guest on September 12, 2011 2010 American College of Chest Physicians