The authors from Catholic University of Leuven, Belgium performed a prospective, randomized, controlled study of intensive versus conventional glucose control among adult patients admitted to a medical intensive care unit (ICU) who at the time of admission were expected to need ICU care for at least 3 days and who did not have advance directives limiting their care. The intensive treatment group received insulin when their blood glucose levels exceeded 110 mg/dL with a targeted glucose range of 80–110 mg/dL. The conventional treatment group received insulin when their blood glucose exceeded 215 mg/dL with a targeted glucose range of 180–200 mg/dL. Glucose was managed using a paper-based protocol previously reported,1 and enteral feeding was directed by care guidelines.
Patients with diabetes were eligible for study and constituted 16.9% of patients. In-hospital mortality was the primary outcome while ICU mortality, days of mechanical ventilation, new kidney injury (defined as a doubling of serum creatinine or peak creatinine of >2.5 mg/dL), and the need for dialysis were secondary outcomes. Hypoglycemia was defined as a glucose level <40 mg/dL.
Between March 2002 and May 2005, 1200 patients were enrolled, 605 in the conventional treatment arm and 595 in the intensive treatment arm, with 381 conventional patients and 386 intensive patients remaining in the ICU for at least 3 days. The investigators analyzed data from both the original cohort and the sub-group of patients with actual ICU stay of >3 days. Among the original cohort of 1200 patients, significantly more intensive treatment patients had at least 1 hypoglycemic glucose measurement (19% vs 3%). Of note, the mortality rate from all causes among patients with hypoglycemia was high, 62% in the intensive treatment group and 73% in the conventional treatment group.
Among the original cohort, in-hospital mortality did not differ significantly between intensive and conventional treatment groups overall: 37% vs 40%. Similarly, 28- and 90-day mortality rates did not significantly differ. However, in the subgroup of patients requiring ICU care for ≥3 days, the mortality rate was consistently lower in the intensive treatment group: the in-hospital mortality rate was 42% in that group compared with 53% in the conventional group (P=.009) and 90-day mortality was 43% in the intensive treatment group vs 53% in the conventional group (P=.06).
Morbidity was also evaluated. In the original cohort of 1200 patients, those receiving intensive treatment were significantly less likely to develop new kidney disease (6% vs 9%, P=.04), were weaned from mechanical ventilation faster, and were discharged from the ICU significantly sooner than those in the conventional treatment arm. However, unlike prior reports,1 there was no difference in rates of bacteremia between treatment groups (8% vs 7%). In the targeted sub-group of patients with ICU stay ≥3 days, the authors again reported significantly shorter duration of mechanical ventilation, shorter ICU stay,...