Continuous subcutaneous infusion of insulin in the management of diabetes mellitus.

Abstract

Control of glycemia by means of continuous subcutaneous infusion of insulin (CSII) was examined in insulin-dependent diabetes mellitus (IDDM). A battery-powered portable infusion pump was employed to deliver insulin at a constant basal rate designed to maintain plasma glucose within the normal range in the postabsorptive state, and empirically determined supplementary infusions of insulin were delivered before meals by the same system with the objective of maintaining excursions of plasma glucose within the normal range during absorption of a meal. The effects of CSII were compared with those of optimized conventional treatment using twice daily injections of mixed intermediate-acting and crystalline insulins in the same groups of volunteers in short term (1 wk) studies in hospital and in longer term (1 to 7 mo) studies with unrestricted activity under normal conditions after leaving hospital. The control of glycemia obtained with CSII in hospital was closely approximated by optimized conventional treatment under these conditions. In studies in hospital, continued infusion of insulin by the CSII program at basal rates with omission of meals and with vigorous exercise did not result in hypoglycemia. Abrupt cessation of CSII for 12 h overnight resulted in moderate ketoacidosis, which was rapidly corrected on resumption of insulin infusion. After the subjects were discharged from hospital, their mean postabsorptive plasma glucose concentrations and mean excursions of plasma glucose after ingestion of meals were maintained within normal limits with CSII but not with optimized conventional therapy. The control of glycemia with CSII was associated with significant reduction of daily dose of insulin by comparison with conventional treatment. During CSII, plasma free immunoreactive insulin (IRI) concentrations were in the normal range in the postabsorptive state. The supplementary doses of insulin before meals with the CSII program were shown to be essential for control of glycemia and were associated with dose-related increments of plasma free IRI, which were detectable within 15 min of delivery, rose to peaks at between 60 and 90 min, and declined to baseline at between 3 and 4 h. It is concluded that the management of IDDM by CSII is effective and safe and provides normalization of mean diurnal plasma glucose concentration through prolonged unrestricted activities under everyday conditions. The treatment can be maintained by the unaided subject and permits the undertaking of long term studies of its effects on the complications of diabetes mellitus.