A randomized pilot study in Type 1 diabetes complicated by severe hypoglycaemia, comparing rigorous hypoglycaemia avoidance with insulin analogue therapy, CSII or education alone.

AIM: To determine potential for amelioration of recurrent severe hypoglycaemia without worsening in overall control in individuals with long-standing Type 1 diabetes (T1DM). METHODS: Twenty-one people with T1DM characterized by altered hypoglycaemia awareness and debilitating severe hypoglycaemia were randomized in a pilot 24-week prospective study to optimized analogue therapy (ANALOGUE; lispro/glargine); continuous subcutaneous insulin infusion therapy (CSII; lispro); or re-education with relaxation of blood glucose targets on existing conventional insulin regimen (EDUCATION). Glycaemic profiles and duration of biochemical hypoglycaemia were measured by continuous subcutaneous glucose monitoring and self-monitored blood glucose. RESULTS: Further severe hypoglycaemia was prevented in five participants (71%) in each group (P = 0.06). Incidence of severe hypoglycaemia was: 0.6 (ANALOGUE), 0.9 (CSII), and 3.7 (EDUCATION) episodes per patient year. Restoration of hypoglycaemia awareness was confirmed by validated questionnaire in three (43%) ANALOGUE, four (57%) CSII and five (71%) EDUCATION patients. Glycated haemoglobin (HbA1c) was significantly improved in the ANALOGUE group between weeks 0 and 24 (8.6 +/- 1.1 vs. 7.6 +/- 0.8%; P = 0.04 for change). Non-significant improvement was seen in the CSII group (8.5 +/- 1.9 vs. 7.4 +/- 1.0%; P = 0.06). No change in HbA1c was seen in the EDUCATION group (8.5 +/- 1.1 vs. 8.3 +/- 1.0%; P = 0.54). There were no episodes of diabetic ketoacidosis or any other adverse events in any group. CONCLUSIONS: In this pilot randomized trial comparing optimized ANALOGUE, CSII or EDUCATION alone in unselected individuals with recurrent severe hypoglycaemia, we show potential for restoring hypoglycaemia awareness and preventing further severe hypoglycaemia with concomitant improvement in glycaemic control in ANALOGUE and CSII groups.