Modeling patients' choice between a primary care physician or a diabetes specialist for the management of type-2 diabetes using a bivariate probit analysis.

Abstract

To determine the prevalence of type-2 diabetes patients that were initially and currently being treated by primary care physicians (PCPs) or diabetes specialists and assess predictors influencing their choice. This cross-sectional study was conducted in nine Greek primary healthcare units. Patients' choices were modeled using a bivariate probit analysis. A total of 225 patients participated (84% response rate). Only 15.9% and 11.9% of the respondents acknowledged having chosen a diabetes specialist for their initial or current treatment, respectively. The family history of diabetes in siblings and the screening for diabetic retinopathy during the past year were significantly positively associated with choosing a diabetes specialist (initially p=0.033 or currently p=0.007), and resulted in a statistically significant reduction of the joint probabilities of choosing a PCP by 15.6% and 13.6%, respectively. Younger age (p=0.040), female sex (p=0.017), higher HbA1c (p=0.004), experience of hypoglycemia (p=0.029) and low cardiovascular morbidity index (p=0.016) emerged as important predictors for choosing a diabetes specialist for their treatment. Our findings provide a better insight in diabetes patients' choices regarding the category of their treating physicians and their predictors. More studies are required to replicate them and identify patient subpopulations that may favor diabetes specialists' guidance.