To evaluate the benefits and risks of tight glucose control vs. usual care in critically ill adult patients.

Randomized controlled trials that met each of the following criteria:

• The setting was an adult ICU
• The intervention group received tight glucose control (glucose goal less than 150mg per dL) obtained using an insulin infusion during part of all of the ICU stay
• The comparison group received usual care (glucose goal and method of insulin administration could vary between studies). The primary or secondary end points included hospital or short-term mortality (up to 30 days), septicemia, new need for dialysis or hypoglycemia.

Studies were excluded:

• If the intervention was conducted primarily during the intra-operative period rather than during the ICU stay
• Unable to obtain adequate details of study methodology or results from the article or study investigators.

• Data Sources: MEDLINE (1950 to 2008), the Cochrane Library, clinical trial registries, reference lists, and abstracts from conferences from both the American Thoracic Society (2001 to 2008) and the Society of Critical Care Medicine (2004-2008)
• Study Selection: Studies in any language in which adult intensive care patients were randomly assigned to tight vs. usual glucose control. Of 1,358 identified studies, 34 randomized trials (23 full publications, nine abstracts, two unpublished studies) met inclusion criteria.
• Statistical Analysis: Two reviewers independently extracted information using a pre-specified protocol and evaluated methodological quality with a standardized scale. Both random- and fixed-effects models to estimate relatives risks (RRs) were used.

Type of Information Abstracted from Articles

• Primary outcome measure: Hospital mortality. Hospital mortality was defined as death occurring during the hospital stay or within 30 days following admission.
• Secondary outcome measures: Septicemia, new need for dialysis and hypoglycemia.

 How Information Was Combined

All outcome measures were calculated on a per-patient basis.

Analytic Methods Used

• Subgroup analyses: Glucose control and ICU setting
• Sensitivity analysis: Proportion of diabetics, use of insulin-only infusions and achieved mean glucose control
• Quantitative Data Synthesis: Use of software provided by Cochrane Collaboration for all analyses version 4.2. This software calculates relative risks (RRs) for studies with at least one occurrence in either study group for each outcome. Trials with missing outcome data or zero occurrences in both groups were excluded from meta-analysis of the outcome. Calculation was made of a pooled RR and 95% confidence interval (CI) for each outcome and considered statistically significant if the test for overall effect had a P value of less than 0.05.

• Initial N: Twenty-nine randomized controlled trials totaling 8,432 patients contributed to this meta-analysis
• Age: Adult, no mean age given.

NS difference in total hospital, ICU mortality or new need for dialysis. Rate of septicemia was significantly decreased with tight glucose control; however, incidence of hypoglycemia was much higher.

Other Findings

NS difference in death rate when stratified by intensive care unit setting.

• Surgical units had 8.8% (tight) vs. 10.8% (conventional); RR, 0.88; 95% Cl, 0.63 to 1.22
• Medical unit: 26.9% (tight) vs. 29.7% (conventional); RR,0,92; 95% Cl, 0.82 to 1.04
• Mixed medical-surgical units, mortality for tight glucose control 26.1 % vs. 27.0% for conventional control; RR, 0.95; 95% Cl, 0.80 to 1.13). 

• In critically ill adult patients, tight glucose control is not associated with significantly reduced mortality but is associated with an increased risk of hypoglycemia
• Given the overall findings of this meta-analysis, it seems appropriate that the guidelines recommending tight glucose control in all critically ill patients should be re-evaluated until the results of larger, more definitive clinical trials are available.

Sixteen international professional societies endorse tight glucose control in all critically ill patients. This extensive meta-analysis highlights the controversy over whether this practice is warranted.

Studies Included in the Meta-analysis

• Very tight control (less than 110mg per dL):
  • Surgical: van den Berghe et al, 2001; Hoedemaekers et al, 2005; van Wezel et al, 2006; He et al, 2007; Stecher et al, 2006
  • Medical: Bland et al, 2005; van den Berghe et al, 2006; Oksanen et al, 2007; Benito et al, 2008; Fernandez et al, 2005
  • Med-Surg: Yu et al, 2005; Mitchell et al, 2006; Brunkhorst et al, 2008; Iapichino et al, 2008; Mackenzie et al, 2005; Arabi et al, 2006; DelaRosa et al, 2006; Devos, 2007.
• Moderately tight control (lLess than 150mg per dL)
  • Surgical: Grey and Perdrizet, 2004; Bilotta et al, 2007; Bilotta et al, 2008; Kia et al, 2005; Chan, unpublished, 2007
  • Medical: Davies et al, 1991; Stefanidis et al, 2002; Walters et al, 2006; Gray et al, 2007; Bruno et al, 2008
  • Med-Surg: Farah et al, 2007; McMullin et al, 2007; Saberi et al, 2004; Henderson et al, 2005; Azevedo, unpublished, 2008.