Long-term, ambulatory, subcutaneous insulin infusion versus multiple daily injections in brittle diabetic patients.

We compared the blood glucose control of four intact and eight kidney recipient, metabolically unstable, ketosis-prone, insulin-dependent diabetic patients under two different regimens: (a) intensive conventional treatment with two to four insulin injections daily (48 patient-months) and (B) subcutaneous, portable insulin delivery system (IDS) (54 patient-months). Both regimens included frequent home blood glucose and 24-h urine glucose determinations and daily telephone follow-up to maximize compliance with treatment. Analyzed as a group the fasting blood glucose for intact patients (A: 172 +/- 13 mg/dl; B: 141 +/- 12, P less than 0.02) and the nonfasting blood glucose for kidney recipient patients (A: 165 +/- 10; B: 138 +/- 5, P less than 0.01) were significantly lower during treatment with the IDS than with multiple injections. Six out of 12 patients (2/4 intact and 4/8 kidney recipient patients) showed significant and consistent improvement of blood glucose concentrations. Four showed marginal and inconsistent improvement. Two patients (one intact and one kidney recipient) improved on the IDS but maintained the improvement when changed back to conventional treatment. The 24-h urine glucose, maximal glucose excursions, number of blood glucoses less than or equal to 40 mg/dl, and glycosylated hemoglobin decreased significantly in some patients on the pump. We conclude that subcutaneous, portable insulin delivery devices can significantly improve the metabolic control of some ambulatory, unstable diabetic patients during long-term treatment beyond that obtained with intensive, multiple-injection, conventional treatment. Normalization of the metabolic control, however, is not obtained. These infusion systems still pose several problems during ambulatory use, which could have serious consequences in patients less compliant and/or followed less closely than ours.