There is current evidence that maintaining ICU patient's blood sugar between 4.4-6.1 saves lives. However, this is difficult to do in the ICU and carries risks of lowering the blood sugar too much. In addition, the best way to achieve this control is not known. Many strict nomograms that provide a standardized approach for nurses have been developed and validated, including one here at SMH. However, these nomograms cannot apply to all patients at all times, especially ICU patients whose needs are rapidly changing. ICU nurses are at the bedside constantly, are very familiar with their patient's needs, and have decades of experience in titrating medication doses without a nomogram to achieve a pre-determined response (i.e. medications to achieve pre-selected blood pressure). Indeed, once the bedside nurse has bought into the importance of the concept of tight glucose control and have been introduced to the nomogram here at SMH, their experience and intuition may be more adaptable to the changing needs of the patient than an inflexible paper nomogram. This study will compare glucose control using our current standard nomogram versus no nomogram (i.e. nurse directed) in order to determine whether the nomogram should continue to be used.

Inclusion Criteria:

Patients:

• adult CVICU pts
• requires insulin therapy for glucose greater than 8 mmol/L
• anticipated ICU stay > 24 hrs
• not in DKA/HHNK
• mechanically ventilated

Exclusion Criteria:

• MD refusal
• no IV access for insulin
• allergy to insulin
• DNR or moribund as determined by the clinical team