Personalised Glucose Therapy: Glucose Targets in Critically Ill Patients with Pre-Existing Poorly Controlled Type 2 Diabetes — ASN Events
  1. Schedule
  2. ADS Clinical Young Investigator Awards
  3. Personalised Glucose Therapy: Glucose Targets in Critically Ill Patients with Pre-Existing Poorly Controlled Type 2 Diabetes

Personalised Glucose Therapy: Glucose Targets in Critically Ill Patients with Pre-Existing Poorly Controlled Type 2 Diabetes (#90)

Palash Kar 1 2 , Mark P Plummer 1 2 , Rinaldo Bellomo 3 , Alicia Jenkins 4 , Andrzej S Januszewski 4 , Kylie Lange 5 6 , Marianne J Chapman 1 2 5 , Michael Horowitz 5 6 , Adan M Deane 1 2 5
  1. Intensive Care Unit, Royal Adelaide Hospital, Adelaide, South Australia, Australia
  2. Discipline of Acute Care Medicine, University of Adelaide, Adelaide, South Australia, Australia
  3. Intensive Care Unit, The Austin Hospital, Melbourne, Victoria, Australia
  4. NHMRC Clinical Trials Centre, University of Sydney, Camperdown, New South Wales, Australia
  5. NHMRC Centre for Research Excellence, University of Adelaide, Adelaide, South Australia, Australia
  6. Discipline of Medicine, University of Adelaide, Adelaide, South Australia, Australia

In patients without pre-existing diabetes, hyperglycaemia during critical illness is associated with adverse outcomes. Paradoxically, recent observational data indicate that for patients with pre-existing poorly controlled type 2 diabetes (defined as an HbA1c >7 %), blood glucose <10 mmol/l is associated with harm1. Accordingly more liberal glucose targets could reduce hypoglycaemia and benefit these patients. This study aims to determine whether more liberal glucose targets in critically ill patients with pre-existing poorly controlled type 2 diabetes increases time weighted mean glucose concentration, attenuates hypoglycaemia, and appears safe.

This is a prospective, open-label, sequential-period (6-month periods), pilot study of 84 patients with inadequately controlled type 2 diabetes (admission HbA1c ≥7.0 %) and with blood glucose >10 mmol/l. Peak blood glucose targets differed between the ‘control’ (n=53) and ‘intervention’ (n=31) groups (6-10 vs. 10-14 mmol/l). Time weighted mean glucose was calculated and hypoglycaemic episodes recorded (blood glucose <4.0 mmol/l). Blood biomarkers for glycaemic variability (1,5-AG) and inflammation (hsCRP) were also collected. Data are mean (SE), median [IQR] or n (%). 

The groups were well matched in terms of age (Control: 64.0 (2.0) vs. Intervention: 62.8 (2.1) years), admission HbA1c (8.5 (0.2) vs 8.7 (0.3) %) and APACHE II score (20.4 (1.0) vs. 19.9 (1.1)). In the intervention group, blood glucose (TWglucose 9.3 (0.3) vs. 10.3 (0.4) mmol/l, P=0.03) and enrolled hours within the study (64 [96] vs. 115 [99] hours, P=0.04) were increased. Additionally, there was a trend towards fewer hypoglycaemic patients (P=0.09) and proportion of time within hypoglycaemic range (P=0.07). There was no difference in ICU mortality (11 [21%] vs. 4 [13%], P=0.37) or 90-day mortality (19 [36%] vs. 10 [32%], P=0.74). Blood biomarkers showed no difference between the groups (1,5-AG 3.6 [4.5] vs. 2.7 [2.8] ug/ml, P=0.30, hsCRP 132.1 [172.9] vs. 148.7 [179.8] mg/l, P=0.78). ICU length of stay was shorter in the ‘control’ period (3.5 [4.0] vs. 6.7 [10.0] days, P=0.01).

In critically ill patients with pre-existing poorly controlled type 2 diabetes more liberal glucose targets increases blood glucose, may reduce the incidence of insulin-induced hypoglycaemia, and appears overtly safe. Prospective studies using larger cohorts are indicated.

  1. Egi M, Bellomo R, Stachowski E, French CJ, Hart GK, Taori G, Hegarty C, Bailey M. The interaction of chronic and acute glycemia with mortality in critically ill patients with diabetes. Critical Care Medicine. 2011; 39:105-11