A randomized study comparing blood glucose control and risk of severe hypoglycemia achieved by non-programmable versus programmable external insulin pumps

Objective: To compare a non-programmable and a programmable insulin externa l pump using regular insulin on glycemic stability, the risk of severe hypo glycemia and metabolic control in type 1 diabetic patients. Material and methods: Ten type 1 diabetic patients were involved in a rando mized, crossover study comparing two periods of 3 months with continuous su bcutaneous insulin infusion (CSII) either with a nonprogrammable insulin pu mp or a programmable insulin pump. Comparisons were made among mean blood g lucose values before and after meals, at bedtime and at 2: 00 a.m.; the ris k of severe hypoglycemia assessed by the low blood glucose index (LBGI); an d HbA1c. Results: Mean average blood glucose (BG) measurements were significantly lo wer with the programmable in comparison with the nonprogrammable insulin pu mp (respectively 157 +/- 78 vs. 165 +/- 79, p = 0.034). While postprandial values for BG were not different between the two pumps, the use of the prog rammable pump resulted in a significant decrease in mean preprandial BG lev els (140 +/- 68 vs. 150 +/- 73 mg/dl p = 0.039). Conversely mean BG level w as lower at 2 a.m, with the non-programmable pump (125 +/- 81 vs. 134 +/- 9 3 mg/dl, p = 0.02) but with a higher incidence of hypoglycemia. Mean LBGI w as comparable with the two pumps (3.1 +/- 8.6 vs. 2.8 +/- 6.9, p = 0.1). Th ere was a 0.2% decrease in HbA1c during the programmable pump period that d id not reach statistical significance (p = 0.37). Conclusions: The present study suggests that programmable external insulin pumps, although more complex and more expensive than nonprogrammable insuli n pumps, significantly reduce fasting glycemia during the day without incre asing the risk of severe hypoglycemia and are safer during the night.