INTRODUCTION: Hyperglycemia is associated with increased risk of diabetic complications; however, this can be reduced by the maintenance of healthy blood glucose levels. Substantial evidence suggests that glycemic control can improve health outcomes in diabetic patients; however, a lack of translation to practice of interventions' effectiveness. Several barriers to providing counseling may exist. The likelihood of counseling has been hypothesized to be associated with patient, physician and system characteristics.

AIM: Explore variations in lifestyle counseling for US adult diabetic patients along with patient, physician, and system characteristics during clinical visit using the NAMCS dataset.

METHODS: Clinical patient visit data was obtained from the National Ambulatory Medical Care Survey [NAMCS] 2012-2015 for analysis. Multivariate logistic regression was utilized to calculate estimates and adjusted odds ratios for characteristics associated with the likelihood of intervention for diabetic patients.

RESULTS: Among the characteristics included in the statistical models for counseling, were found to be significantly associated with the likelihood of being offered diet and exercise counseling in the logistic model. These include HbA1c testing [OR=1.615]; Shift, hourly physicians [OR=5.370]; EMR meeting DHHS criteria [OR=6.529]; patient race [Hispanic OR=2.074, Black OR=1.337]; physician specialty [cardiology OR=1.402]; patient’s history of chronic conditions [obesity OR=2.524, OR=4.264].

DISCUSSION: More than patient level characteristics are associated with the likelihood of counseling. Physician and system-level characteristics are also significantly associated with the likelihood intervention. Identifying the sources of variations could not only better understand barriers to executing counseling but also effectively reducing the burden of diabetes and other co-morbidities.