[Management of blood glucose during enteral nutrition in critical patients].

Objective: To investigate an effective and safe protocol for enteral nutrition (EN) patients permitting successfully transmit insulin administration from venous pump-in to subcutaneous injection.

Methods: A prospective randomized control study was conducted. Critical patients admitted to intensive care unit (ICU) of Beijing Tongren Hospital from September 2008 to February 2009 were randomly divided into two groups when the energy provided by EN up to half of the total energy requirement. Experiment group (n=44): the protocol was applied for insulin glargine and regular insulin injection; control group (n=43): protocol was applied for subcutaneous regular insulin injection. Target glucose range was 4.4-7.8 mmol/L (80-140 mg/dl). If blood glucose ≥11.1 mmol/L was maintained twicely, the approach of insulin administration would convert from subcutaneous injection to venous pump-in using the computerized glucose control protocol. If the infusion rate of insulin was less than 3 U/h and lasted more than 6 hours, blood glucose ≤7.8 mmol/L, insulin administration was switched to subcutaneous injection again. The general information and all glucose regulation data were recorded for analysis.

Results: The two groups did not differ at baseline for the general information, mean blood glucose and the glucose variation. A total of 1689 blood glucose records were analyzed. The mean blood glucose in experiment group, and was significantly lower than that in control group(7.58±1.17 mmol/L vs. 9.40±1.74 mmol/L, P<0.05). The rate of glucose values within target range in experiment group was significantly higher than that in control group [49.72% (534/1074) vs. 35.61% (219/615), P<0.01]. The glucose standard deviation (SD) in experiment group was significantly lower than that in control group (1.89±0.52 mmol/L vs. 2.17±0.94 mmol/L, P<0.05). The number of measurements needed per patient per day was significantly reduced in experiment group compared with control group (7.51±1.31 vs. 8.15±0.97, P<0.05). The ratio of patients converted to venous pump-in was significantly decreased in experiment group compared with control group (9.09% vs. 44.19%, P<0.01). Hypoglycemia (≤3.3 mmol/L) did not different between experiment group and control group [0.74% (8/1074) vs. 0.49% (3/615), P=0.75].

Conclusions: Compared with the conventional subcutaneous insulin injection protocol, this protocol with insulin glargine combined regular insulin subcutaneous injection can control the glucose level effectively during EN in critical patients. The glucose variation and the numbers of measurements were significantly reduced by this protocol. It is helpful for the insulin transmission from venous pump-in to subcutaneous injection.