Objective: To determine whether glycemic control is less feasible when hydrocortisone is given as a bolus compared with continuous application in septic shock patients.
Design: Observational prospective pilot study.
Setting: Fourteen-bed surgical university hospital ICU.
Patients: Sixteen consecutive patients with septic shock receiving a continuous infusion of 200 mg hydrocortisone/day and an infusion regime of insulin keeping blood glucose below 150 mg/dl.
Intervention: Blood glucose and insulin infusion were adjusted to steady state before intervention. At baseline, the continuous hydrocortisone infusion was replaced with a single bolus of 50 mg hydrocortisone. During a subsequent 6-h period, blood glucose was monitored hourly and insulin infusion was kept constant. Afterwards, hydrocortisone application and adjustment of blood glucose was resumed according to standard treatment.
Results: Mean blood glucose in steady state at baseline immediately prior to intervention was 128 mg/dl (range 114-141 mg/dl; 95% confidence interval). After bolus injection of hydrocortisone, blood glucose increased significantly within 6 h with peak levels of 154 mg/dl (range 132-178 mg/dl; p<0.01). Blood glucose returned to baseline with restoration of continuous hydrocortisone infusion. There were marked inter-individual variations with peak glucose values up to 254 mg/dl, but no significant difference in intra-individual glucose variability before and after bolus injection of hydrocortisone.
Conclusions: Bolus injections of hydrocortisone may induce significant increases of blood glucose in patients with septic shock. The individual response is highly variable and we speculate that repetitive boluses would induce marked undulation of blood glucose. In terms of glycemic-control strategies, a continuous infusion of hydrocortisone seems to be preferable.