As part of a randomized, prospective trial in subjects with insulin-dependent diabetes assigned to either continuous subcutaneous insulin infusion (CSII) or to their unchanged conventional insulin treatment (CIT) for 8 mo, patients completed questionnaires dealing with general responses and clinical and technological problems.

Although there was no significant difference between treatment groups with respect to the number of patients experiencing severe hypoglycemia (i.e., requiring intravenous glucose or intramuscular glucagon injection), there were nine episodes in six patients during CSII compared with one episode during CIT. Diabetic ketoacidosis occurred significantly more often in the CSII group (nine episodes in eight patients) than in the CIT group (no episodes).

A number of CSII-related failures occurred, including omission of premeal bolus, needle dislodgement, pump accidentally turned off, and leakage at the infusion site. At 8 mo, 85% of CSII-treated patients wished to continue indefinitely on the pump, and almost all would continue with self-monitoring of blood glucose even if they Stopped CSII.