Randomized comparison of metabolic control achieved by intraperitoneal insulin infusion with implantable pumps versus intensive subcutaneous insulin therapy in type I diabetic patients.

Abstract

To compare intraperitoneal implantable insulin infusion (IP) to subcutaneous (SC) intensive insulin therapy. Twenty-one insulin-dependent (type I) diabetic patients aged 24-61 yr underwent a 3-mo treatment optimization using multiple SC daily injections or external pumps. Patients were then randomized (time 0 mo) to IP infusion using Infusaid-programmable pumps or continuation on SC intensive insulin for 6 mo. No differences were noted between study and control group data. However, longitudinal within-group comparisons from baseline showed that glycosylated hemoglobin improved to near-normal in both groups: IP, 9.0 +/- 0.5 vs. 7.8 +/- 0.6% (P less than 0.05) and SC, 8.4 +/- 0.5 vs. 7.5 +/- 0.3% (P less than 0.5) at 0 and 4 mo, respectively (normal less than 6.9%). The percentage of blood glucose tests greater than 11 mM at 0 and 6 mo was 28 +/- 5 vs. 16 +/- 4% in the IP group (P less than 0.05) and 22 +/- 5 vs. 24 +/- 7% in the SC group (NS). At 0 and 6 mo, the standard deviation of blood glucose values, an index of glycemic fluctuations, was 4.3 +/- 0.4 vs. 3.2 +/- 0.5 mM in the IP group (P less than 0.05) and 3.7 +/- 0.3 vs. 4.0 +/- 0.4 mM in the SC group (NS). Weight, insulin dosages, circulating lipid levels, and the frequency of severe hypoglycemic reactions and biochemical hypoglycemias were similar and did not change in the two groups. IP-implantable pumps compared with SC intensive insulin therapy have similar effects on most metabolic variables and are equally effective at achieving near-normal glycemic levels. Only longitudinal data suggest that IP treatment may be more effective at limiting glycemic fluctuations.