An increasing body of literature suggests that poor glycemic control is associated with worse outcome in critically ill patients and that achieving good glycemic control with aggressive insulin therapy reduces morbidity and mortality. In our institution, a fairly simple insulin infusion protocol has been in place for patients undergoing cardiac surgery. The hourly insulin drip rate is calculated as: (patient glucose in gm/dl –60) x factor. The initial factor is usually 0.03 and is adjusted subsequently. However, the source and validity of this protocol were not elucidated despite a literature search. This retrospective study aims to ascertain the validity of the protocol.
Charts of patients undergoing coronary artery bypass graft (CABG) surgery were reviewed and 20 diabetic patients identified; 10 who received the insulin protocol and 10 that did not. Data abstracted included patient demographics, parameters needed to determine the APACHE III score on the day of surgery, and all fingerstick glucose readings on 3 consecutive days starting from the day of surgery.
Nineteen of the 20 patients were males. The mean age (65 ± 15 vs. 65 ± 10) and mean APACHE III scores (14.4 ± 8 vs. 14.6 ± 12) were similar between the control and insulin protocol groups, respectively. The mean glucose levels on each of the three days were 162, 151 and 179 mg/dl in the control group and 157, 148 and 141 in the protocol group. Glucose level was £145mg/dl in 41% (76/184) of readings in the control group and 65% (228/353) of readings in the protocol group (p<0.0001; Chi-square). Glucose levels <60mg/dl occurred in 2.7% (5/184) of control and 0.6% (2/353) of protocol group readings (p<0.05; Fisher Exact Test).
In our cohort, this insulin protocol resulted not only in significantly better glycemic control, but also a significant reduction in the incidence of hypoglycemia.
This relatively simple insulin protocol appears to optimise glycemic control in CABG patients and may also have utility in non-CABG, critically ill patents.
Vinay Sharma, None.