From Wiki Journal Club
Jump to: navigation, search
Preiser JC, et al. "A prospective randomised multi-centre controlled trial on tight glucose control by intensive insulin therapy in adult intensive care units". Intensive Care Medicine. 2009. 35(10):1738-1748.
PubMedFull textPDF

Clinical Question

In critically ill medical and surgical patients, does intensive insulin therapy (target 4.4–6.1 mmol/L) reduce mortality as compared to conventional glycemic control (target 7.8–10.0 mmol/L)?

Bottom Line

In critically ill medical and surgical patients, intensive insulin therapy (target 4.4–6.1 mmol/L) did not reduce mortality as compared to conventional glycemic control (target 7.8–10.0 mmol/L). Intensive insulin therapy was associated with an increased risk of hypoglycemia. The study was stopped prematurely and underpowered.

Major Points

In the Leuven Surgical Trial (2001) of surgical ICU patients, intensive insulin therapy (target glucose 80-110 mg/dL or 4.4-6.1 mmol/L) reduced the in-hospital mortality to 7.2% as compared to 10.9% in patients with conventional treatment (target glucose 180-200 mg/dL or 10-11.1 mmol/L; (P=0.01).[1] This survival benefit was not seen in the follow-up Leuven Medical Trial of medical ICU patients.[2] The NICE-SUGAR trial (2009) randomized 6,104 medical and surgical ICU patients to receive intensive glycemic control (target glucose 81-108 mg/dL) or conventional glycemic control (target glucose ≤180 mg/dL). The intensive glycemic control patients had a higher risk of 90-day mortality (OR 1.14; 95% CI 1.01-1.29; P=0.04) and hypoglycemia (OR 14, 95% CI 9.0-25.9; P<0.001).

The GLUCONTROL study randomized 1,101 medical and surgical ICU patients to receive intensive insulin therapy (IIT) (target 4.4-6.1 mmol/L) and conventional insulin therapy (target 7.8-10 mmol/L).[3] IIT did not reduce mortality (17.2% vs. 15.3% in conventional control group, P=0.41) but the risk of hypoglycemia was increased (8.7% vs. 2.7% in conventional control group, P<0.0001). The trial was stopped prematurely and therefore underpowered.

The study was included in a Cochrane analysis of perioperative glycaemic control for diabetic patients undergoing surgery.[4] Post-hoc analysis showed that intensive glycaemic control may increase the risk of hypoglycaemia.(3 trials, n=724; RR 6.92, 95% CI 2.04-23.41, P=0.002). No significant improvement in all-cause mortality was seen (11 trials, n=1,365; 12.2% vs. 9.6% in conventional group; RR of 1.19, 95% CI 0.89-1.59, P=0.24).


Surviving Sepsis Campaign: International Guidelines for Management of Severe Sepsis and Septic Shock, 2012[5]

  • In ICU patients with severe sepsis, commence insulin when 2 consecutive blood glucose levels are are >180mg/dL. The treatment target is blood glucose level ≤180mg/dL (grade 1B).

ADA Standards of Medical Care in diabetes, 2016[6]

  • Insulin should be initiated for treatment of persistent hyperglycemia when glucose levels are ≥180 mg/dL (10.0 mmol/L). Target glucose level is 140-180 mg/dL (7.8–10.0 mmol/L) for critically ill patients (grade A) and non-critically ill patients (grade C).
  • For selected critically ill patients, more stringent goals, such as 110-140 mg/dL (6.1-7.8 mmol/L) may be appropriate if this can be achieved without significant hypoglycemia (grade C).


  • Multicenter, randomized, controlled trial
  • N=1,101
    • Intensive (n=550)
    • Conventional (n=551)
  • Setting: 21 centers in 7 countries
  • Enrollment: 2004-2006
  • Analysis: intention-to-treat
  • Primary outcome: mortality in ICU


Inclusion Criteria

  • age>18 years
  • patients admitted to medical and surgical ICU

Exclusion Criteria

  • life expectancy <24 hours
  • no consent provided
  • pregnant patients

Baseline Characteristics

From the IIT group

  • Demographics: age 64.8 (50.8-74) years, 64.4% male
  • Type of patient: medical 42.2%, scheduled surgery 30.2%, emergency surgery 16.6 %, trauma 7.6%
  • Organ-system involved: cardiac 31.7%, respiratory 17.9%, gastroenterology 16.4%, neurology 16.4%, vascular 1.7%, renal 2.1%, orthopedic 6.7%, hematological 0.4%, other 9,3%
  • Severity and prognosis scoring systems: APACHE II 15 (11-21), SOFA 6.9±3.1, Glasgow coma score 15 (8-15)
  • Respiratory support: invasive ventilation 67.7%, non-invasive ventilation, spontaneous breathing 26.1%
  • Vasopressors/inotrope support: 37.5%
  • Patients febrile to >38.5°C: 9.7%
  • Pre-existing diabetes: 16.2%


  • Medical and surgical ICU patients were randomized to intensive glycemic control (target 7.8-10.0 mmol/L) or conventional glycemic control (target 4.4-6.1 mmol/L)
  • Regular insulin (Actrapid) was delivered by continuous IV infusion pump in ICU. After patients were discharged from the ICU or when they were able to tolerate full oral intake, IV insulin was changed to subcutaneous administration
  • Hypoglycemia was defined as blood glucose <2.2 mmol/L


Comparisons are intensive vs. conventional glycemic control.

Primary Outcomes

Mortality in ICU
17.2% vs. 15.3% (P=0.41)
13.4% vs. 11% (P=0.5483) in patients with length of stay (LOS) equal or less than 3 days
18.8% vs. 17% (P=0.5135) in patients with LOS >3 days

Secondary Outcomes

28-day mortality
18.7% vs. 15.3% (P=0.1438)
34 and 33 patients were still in ICU at day 28
In-hospital mortality
23.3% vs. 19.4% (P=0.1136)
6 (3–13) vs. 6 (3–13) days (P=0.238)
Hospital LOS
16 (11–29) vs. 16 (11–29) days (P=0.708)
Mean SOFA score during ICU stay
6.0±2.9 vs. 5.9±3.1 days (P=0.583)
mean SOFA score during the days with hypoglycemia was higher (7.3) as compared to days without hypoglycemia (6.1), P<0.01
Number of febrile days
392 vs. 384 days (P=0.98)
Number of days requiring anti-infective agents
0 (0–5) vs. 0 (0–5) days (P=0.573)
Hypoglycemia <2.2 mmol/L
8.7 vs. 2.7% (P<0.0001)
Duration of mechanical ventilation
1,155 vs. 1,179 patient days (P=0.562)
Duration of renal replacement therapy
519 vs. 523 patient days (P=0.753)
Duration of vasopressor/inotropic support
1,521 vs. 1,350 patient days (P<0.0001)
Number of packed red blood cell
0 (0–2) vs. 0 (0–2) units (P=0.607)

Other Outcomes

Patients treated with IV insulin
96.3 vs. 66.2% (P<0.0001)
Duration on insulin (hours)
36 (13–96) vs. 10 (0–43) (P<0.0001)


  • The study was not blinded.
  • Potential reasons for hypoglycemia were not explored nor recorded.
  • Different devices were used for measuring blood glucose in different centers, therefore this may have introduced inaccuracy.
  • The material for insulin infusion was not standardized across all centers, this may have introduced inaccuracy in the rate of infusion.
  • The study was stopped prematurely and therefore underpowered.[7]


  • Communauté Française Wallonie-Bruxelles (Belgium) grant

Further Reading

  1. van den Berghe G, et al. "Intensive insulin therapy in critically ill patients." The New England Journal of Medicine.' 2001;345(19):1359-1367.
  2. Van den Berghe G, Wilmer A, Hermans G, et al. Intensive insulin therapy in the medical ICU. N Engl J Med 2006;354:449-461
  3. Preiser JC 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. 35:1738-48.
  4. Buchleitner AM et al. Perioperative glycaemic control for diabetic patients undergoing surgery. Cochrane Database Syst Rev 2012. :CD007315.
  5. Dellinger RP, et al. Surviving sepsis campaign: international guidelines for management of severe sepsis and septic shock: 2012. Crit Care Med. 2013;41(2):580-637
  6. American Diabetes Association. 13. Diabetes Care in the Hospital. Diabetes Care 2016 39:S99-S104; doi:10.2337/dc16-S016
  7. Schultz MJ et al. Glucontrol, no control, or out of control?. Intensive Care Med 2010. 36:173-4; author reply 175-6.