LOGIC-Insulin Algorithm-guided Versus Nurse-directed Blood Glucose Control During Critical Illness (LOGIC-1)

• Proportion of patients with severe hypoglycemia (<40 mg/dL) during the intervention [ Time Frame: up to 14 days post-randomization ] [ Designated as safety issue: Yes ]
  Proportion of patients to have had one or more episodes of severe hypoglycemia (<40 mg/dL) during the intervention
• Incidence of severe hypoglycemia (<40 mg/dL) during the intervention [ Time Frame: up to 14 days post-randomization ] [ Designated as safety issue: Yes ]
  Number of severe hypoglycemic values as a fraction of all blood glucose measurements during the intervention
• Blood glucose level per treatment group during the intervention [ Time Frame: up to 14 days post-randomization ] [ Designated as safety issue: No ]
• Hyperglycemic index (HGI) during the intervention [ Time Frame: up to 14 days post-randomization ] [ Designated as safety issue: No ]
  Adequacy of reaching and maintaining the target range for blood glucose during the intervention
• Daily maximal blood glucose difference during the intervention [ Time Frame: up to 14 days post-randomization ] [ Designated as safety issue: No ]
  Marker of blood glucose variability
• Proportion of patients with common hypoglycemia (<60 mg/dL) during the intervention [ Time Frame: up to 14 days post-randomization ] [ Designated as safety issue: No ]
  Proportion of patients to have had one or more episodes of common hypoglycemia during the intervention
• Interval between blood glucose measurements during the intervention [ Time Frame: up to 14 days post-randomization ] [ Designated as safety issue: No ]
  Marker of workload
• Percentage of time in target zone (80-110 mg/dL) during the intervention [ Time Frame: up to 14 days post-randomization ] [ Designated as safety issue: No ]
  Adequacy of reaching and maintaining the target range for blood glucose during the intervention
• Length of stay in ICU [ Time Frame: up to 90 days post-randomization ] [ Designated as safety issue: No ]
• Length of stay in hospital [ Time Frame: up to 90 days post-randomization ] [ Designated as safety issue: No ]
• Hospital mortality [ Time Frame: up to 90 days post-randomization ] [ Designated as safety issue: No ]
  Patients who will have been discharged from hospital before 90 days post-randomization will be regarded as survivors
• Incidence of common hypoglycemia (<60 mg/dL) during the intervention [ Time Frame: up to 14 days post-randomization ] [ Designated as safety issue: No ]
  Number of common hypoglycemic values as a fraction of all blood glucose measurements during the intervention

The LOGIC-Insulin computerized software algorithm will be compared with a nurse-directed protocol, both targeting a blood glucose level of 80-110 mg/dL, in critically ill patients

Critical illness typically causes elevated blood glucose concentrations, which have been associated with increased mortality. Strictly normalizing these blood glucose levels by intensive insulin therapy, called tight glycemic control, decreased morbidity and mortality in well-controlled single-center clinical studies. However, multi-center pragmatic trials failed to reproduce those effects, proving that the implementation of tight glycemic control in daily clinical practice is rather difficult.

A computerized algorithm, amongst others, may help the nurses in the titration of insulin to reach normal blood glucose levels and to avoid the particularly worrisome hypoglycemia.

The LOGIC-1 study is a single blinded randomized controlled trial. On admission patients will be randomly assigned to either tight glycemic control (80-110 mg/dL) by the computerized LOGIC-Insulin 3.0 algorithm or to tight glycemic control (80-110 mg/dL) by the nurse-directed protocol. Written informed consent will be asked from the patient in the case of elective surgery requiring post-operative ICU-admission. Proxy informed consent from the closest family member will be asked when the patient was admitted to the ICU in emergency. As blood glucose control by itself is essential in the management of critical illness, random allocation will be done on admission and written informed consent from the closest family member can be deferred to a maximum of 24 hours after randomization. The patient or family member can at all times withdraw from the trial without impact on his treatment. Allocation will be done in blocks (block size is unknown to the care givers responsible for treatment allocation), stratified into cardiac surgery and other reasons for ICU admission, by central computer randomization.

The time window for the study will be 14 days starting from admission to the ICU or when one of the following stop criteria will be met:

• Withdrawal of the informed consent
• Patient starts eating or drinking sugar containing liquids
• Patient is discharged from the ICU (including ICU deaths)
• Removal of arterial line or central venous line

Under the following conditions the study investigator/treating physician will be contacted and, if necessary, patients will be switched from LOGIC-Insulin to nurse-directed blood glucose control:

• Recurrent severe hypoglycemia (<40 mg/dL)
• Refractory hyperglycemia
• Any change in condition that compromises the safety of the patient, as judged by the investigator or treating physician

The common strategy for blood glucose control in both groups involves blood glucose measurements from arterial blood by a blood gas analyzer and the administration of insulin through a central line with a syringe pump.

• Procedure: Nurse directed insulin titration
  Nurse directed insulin titration to establish glycemic control in the target range of 80-110 mg/dL
  Other Name: Active comparator: titration-guidelines supported TGC
• Procedure: LOGIC-Insulin
  Blood glucose control (80-110 mg/dL) guided by the LOGIC-Insulin algorithm
  Other Name: software guided TGC

• Experimental: LOGIC-Insulin
  Blood glucose control (80-110 mg/dL) guided by the LOGIC-Insulin algorithm
  Intervention: Procedure: LOGIC-Insulin
• Active Comparator: Nurse-directed
  Nurse-directed blood glucose control (80-110 mg/dL)
  Intervention: Procedure: Nurse directed insulin titration

• van den Berghe G, Wouters P, Weekers F, Verwaest C, Bruyninckx F, Schetz M, Vlasselaers D, Ferdinande P, Lauwers P, Bouillon R. Intensive insulin therapy in the critically ill patients. N Engl J Med. 2001 Nov 8;345(19):1359-67.
• Van Herpe T, De Brabanter J, Beullens M, De Moor B, Van den Berghe G. Glycemic penalty index for adequately assessing and comparing different blood glucose control algorithms. Crit Care. 2008;12(1):R24. Epub 2008 Feb 26.
• Van Herpe T, De Moor B, Van den Berghe G. Towards closed-loop glycaemic control. Best Pract Res Clin Anaesthesiol. 2009 Mar;23(1):69-80. Review.
• Van Herpe T, Mesotten D, Wouters PJ, Herbots J, Voets E, Buyens J, De Moor B, Van den Berghe G. LOGIC-insulin algorithm-guided versus nurse-directed blood glucose control during critical illness: the LOGIC-1 single-center, randomized, controlled clinical trial. Diabetes Care. 2013 Feb;36(2):188-94. doi: 10.2337/dc12-0584. Epub 2012 Sep 6.

Inclusion Criteria:

• Aged 18 years or more
• Admitted to the ICU
• Already receiving or needing insulin infusion for blood glucose control

Exclusion Criteria:

• Not critically ill (eating, not mechanically ventilated)
• Pregnant or breastfeeding
• Previous inclusion into the trial
• Included in other trial
• Moribund
• Diabetes coma
• No arterial line available

• Agency for Innovation by Science and Technology, Flanders, Belgium
• Fund for Scientific Research, Flanders, Belgium