Insulin therapy in type 2 diabetes patients : balancing between evidence and experience

In 2005 we investigated in a questionnaire survey in 1621 general practitioners (GPs) whether the practice organisation in primary care is equipped to provide optimal diabetes care. This was expressed in the Structure Index (SI), based on the presence of practice characteristics (score 0-4.5). The mean SI was 2.10.9 and was higher in training practices and in practices in which GPs started insulin treatment. About 75% of the GPs worked in practices that were average in organisation necessary to provide optimal diabetes care; the rest worked in practices poorly equipped or in practices well equipped to provide such care. In the questionnaire survey, we investigated the organisation of insulin therapy and assessed factors that were associated with providing insulin therapy in type 2 diabetes patients. 67% of the GPs start insulin therapy, especially male GPs and those older than 40, GPs working in a health centre and those working with a practice nurse. GPs working in urban regions less often start insulin. The most often mentioned barriers for starting insulin therapy were lack of knowledge, lack of time and insufficient financial incentives. We compared psychological well-being between type 2 diabetes patients using oral glucose lowering agents and patients on insulin therapy with or without oral agents. Insulin-users reported worse outcomes on Short Form-36 subdomains physical functioning and general health; and on Diabetes Health Profile (DHP) total score, barriers to activity and psychological distress. Insulin regimen and diabetes duration did not influence these outcomes. We reviewed 78 randomised clinical trials regarding insulin preparations in patients with type 2 diabetes. They showed that continuing oral glucose lowering agents after the start of basal insulin therapy resulted in better glycaemic control with less insulin requirements, less weight gain and less hypoglycaemic events than monotherapy. Long-acting insulin analogues were associated with similar glycaemic control, but less hypoglycaemia than NPH insulin. Premix insulin provided better glycaemic control than a basal insulin, but caused more hypoglycaemia. Analogue premix insulin provided similar glycaemic control, but lower postprandial glucose levels compared to human premix insulin. Drawing conclusions from the studies concerning basal-bolus regimen seemed not possible. In a randomised clinical trial 39 type 2 diabetes patients with inadequate glycaemic control on metformin and sulphonylurea, were randomised to continue or discontinue sulphonylurea, once they started insulin glargine therapy and continued metformin. After one year, the combination with sulphonylurea provided no significant differences in C-peptide-glucose ratio, an estimator for ?-cell function, compared to not using sulphonylurea. In the group continuing sulphonylurea the number of hypoglycaemic events was higher and also weight gain was, not statistically significant, higher in this group. We reviewed the effects of insulin monotherapy versus insulin-oral glucose lowering agents combination therapy in type 2 diabetes patients already treated with insulin in 38 randomised clinical trials. The addition of oral agents has positive effects on glycaemic control and insulin requirements. Combination with sulphonylurea or thiazolidinediones causes more hypoglycaemic events, thiazolidinediones more weight gain and metformin less weight gain. Side effects such as gastro-intestinal complaints and oedema also occurs.