Factors influencing initial choice of insulin therapy in a large international non-interventional study of people with type 2 diabetes

Abstract

Aim: To use baseline characteristics of the Cardiovascular Risk Evaluation in people with type 2 Diabetes on Insulin Therapy study population to identify factors that could explain the choice of insulin therapy when beginning insulin.

Methods: The source, non-interventional, longitudinal, long-term study involves 314 centres in 12 countries in five regions. People were enrolled having started any insulin regimen in the previous 12 months. To identify factors associated with the choice of insulin regimen, multivariable backward logistic regression was performed on eligible physician and participant explanatory variables.

Results: Participants (N = 3031) had mean age 62 years, diabetes duration 11 years, body mass index 29.3 kg/m² and an HbA1c of 9.5%. Participants in Japan had less hypertension, smoked more and used fewer concomitant medications than those of other regions. Only physician location (rural or urban) influenced the choice of insulin in Japan. In the other four-regions-combined, physician location, specialty, sex and practice type influenced choice of insulin as did participant location, baseline HbA1c, use of glucose-lowering therapies and prior insulin secretagogue use.

Conclusion: Choice of initial insulin regimen was influenced by several physician and participant characteristics in Canada and Europe, but only by physician location in Japan.

Figure 1

Factors influencing choice of initial insulin therapy in people with type 2 diabetes, as determined by multivariable analysis of premix vs. basal insulin (A) and other insulin treatments vs. basal insulin (B) in all regions except Japan. The model was based on 2200 observations in A and 469 observations in B. There was no evidence of over-dispersion as the ratio of deviance to degrees of freedom was 0.86. HbA1c odds are per 1.0% units increase. For each pair of factors (e.g. rural vs. urban), odds ratios (ORs) > 1 indicate that the first of the pair (e.g. rural) favours premix insulin over basal insulin (A) or other insulin over basal insulin (B); ORs < 1 indicate that the first of the pair (e.g. general practitioner) favours basal insulin.