Hydrocortisone Continuous Drip in ICU Pushed for Better Glycemic Control

LAHTI, Finland, Feb. 16 -- Continuous hydrocortisone infusions for septic shock in the intensive care unit (ICU) may yield fewer hyperglycemia episodes and less work for nurses compared with bolus infusions, researchers here said.

Although tight glycemic control was fairly easily achieved with both methods, there was a small but significant advantage in glucose levels with continuous administration, reported Pekka Loisa, M.D., of the Pijt-Hme Central Hospital, and colleagues, online in Critical Care.

Tight glycemic control improves survival among some ICU patients, according to previous studies. However, low-dose hydrocortisone used to treat vasopressor-dependent septic shock in the ICU can induce hyperglycemia.

"These effects must also be considered as major adverse events in critically ill patients," Dr. Loisa and colleagues wrote, because other studies have linked prevention of hyperglycemia to outcomes.

To see whether continuous infusion could reduce some of the fluctuations in blood glucose levels, Dr. Loisa and colleagues randomized 48 such patients from several ICUs across Finland to continuous or bolus infusions totaling 200 mg of hydrocortisone daily for five days.

Bolus infusions were given in 50-mg intravenous doses every six hours. All patients received maintenance infusions of 5% glucose (30 ml/kg per day) as well as enteral nutrition (500 up to 1,500 ml/day) with standard formulas according to a protocol. The blood glucose goal was 4 to 7 mmol/l. Hyperglycemia (glucose above 7 mmol/l) was treated with 1 IU/ml of Actrapid insulin, then adjusted according to a strict algorithm.

None of the patients had pre-existing diabetes or was receiving glucocorticoids. Baseline glucose levels and other characteristics were similar between groups.

For the primary endpoint of mean blood glucose levels, there was a small but significant improvement for continuous infusion (6.2 versus 6.4 mmol/l, P=0.04).

However, "this difference of 0.2 mmol/l cannot be considered clinically significant," the researchers cautioned.

For the coprimary endpoints of the occurrence of hyper- and hypoglycemic episodes, the researchers found:

• Significantly fewer hyperglycemic episodes per patient in the infusion group (15.7 8.5 versus 10.58.6, P=0.039),
• No significant difference in the rate of hypoglycemia below 3 mmol/l between the infusion and bolus groups (13% versus 4.5%, P=0.609),
• No difference in severe hyperglycemia above 8.3 mmol/l episodes per patient (2.63.2 versus 3.63.4, P=0.383), and
• No severe hypoglycemia below 2.2 mmol/L in either group.

Those in the infusion group also needed significantly fewer changes to their insulin infusion rate to maintain strict normoglycemia (3.41.9 versus 4.72.2 adjustments per patient per day, P=0.038).

Insulin requirements and caloric intake were similar between groups. Adjusting for administered calories indicated a trend toward lower insulin requirements in the infusion group throughout the five-day study period, but this difference was not significant because of large individual variations (P=0.505).

Reversal of septic shock with withdrawal of vasopressor support was similar between groups at 48 hours (50% infusion and 58% bolus) and at five days (63% versus 83%, respectively, P=0.83). Four patients in the infusion group and two in the bolus group died from refractory hypotension.

The researchers noted that the study was not placebo controlled or blinded, but that the major limitation in the study was heterogeneity in nutritional support among patients even though there was a clear protocol for feeding.

Most study patients had septic shock from gastrointestinal disease, such as perforation or acute pancreatitis, so enteral feeding could not always follow the protocol, they said.