The Asheville Project: factors associated with outcomes of a community pharmacy diabetes care program

OBJECTIVE: To assess the importance of environment, patient characteristics, and health behavior in explaining differences in clinical, economic, and humanistic outcomes of pharmaceutical care services (PCS) for patients with diabetes. DESIGN: Quasi-experimental, pre-post cohort-with-comparison group study using multivariate logistic regression. SETTING: Twelve community pharmacies in Asheville, N.C. PATIENTS AND OTHER PARTICIPANTS: Eighty-five patients with diabetes who were employees, dependents, or retirees from two self-insured employers; community pharmacists who completed a diabetes certificate program and received reimbursement for PCS. INTERVENTIONS: Scheduled consultations with pharmacists involving education and training, assessment, monitoring, follow-up, and referral. MAIN OUTCOME MEASURES: Change in glycosylated hemoglobin (A1c) value, diabetes diagnosis and all-diagnosis utilization and cost of medical care, quality of life, and satisfaction with pharmacy services. RESULTS: The strongest predictors of improvement in A1c following PCS were the patient characteristics baseline glycemic control and type 1 diabetes. All patients with type 1 diabetes had reduced their A1c concentrations at follow-up. Patients in one employer group (an environmental characteristic) were significantly more likely to have a 10% reduction in diabetes diagnosis costs, compared with employees in the other group. They were also more likely to report improved satisfaction with pharmacy services. No other statistically significant relationships were found. CONCLUSION: The greatest improvement in A1c occurred among patients with type 1 diabetes and/or higher baseline A1c concentrations. When controlling for other factors, PCS did not emerge as a significant factor in lowering A1c, but it was imprecisely measured, and our proxy measure did not capture the full complement of PCS provided to patients. Success in terms of cost savings and patient satisfaction differed by employer group.