Trajectories of Glycemic Control with Clinical Pharmacy Specialist Management of Veterans with Type 2 Diabetes

Abstract

Purpose: Prior evidence suggests that early, improved control of glycemic control likely lowers the risk of some combination of microvascular and/or macrovascular complications. Furthermore, growing evidence suggests that not only achieving a certain degree of control, but how that control varies also may matter in terms of outcomes. Finally, there is some initial evidence that clinical pharmacy specialist (CPS) services can, on average, lower glycosylated hemoglobin (HbA1c) and improve glycemic control. However, there is insufficient evidence characterizing variation in glycemic control outcomes in patients receiving CPS management of diabetes and factors associated with greater chance of success with that management. Methods: This is an observational, multicenter, retrospective cohort study of Veterans with type-2 diabetes in Veterans Integrated Service Network 15 managed by clinical pharmacy specialists between 7/1/2013 and 7/1/2017 with a baseline HbA1c level ≥ 8%. Glycosylated hemoglobin measurements were collected for two years following the index date and used to group patients into distinct patterns of HbA1c trajectories over time using group-based trajectory modeling and posterior probabilities of group membership. Characteristics associated with successful HbA1c trajectories and association of assigned trajectories with all-cause and diabetes-related hospitalizations were analyzed using logistic regression. Results: A total of 4,119 Veterans were included and successfully divided into six distinct HbA1c trajectory groups: High Gradually Decreasing (n=325, 7.9%), Moderate Early Decline (n=1692, 41.1%), Large Early Decline (n=231, 5.6%), Uncontrolled Stable (n=1468, 35.6%), Early Decline / Subsequent Increase (n=266, 6.5%), and Very Uncontrolled Stable (n=137, 3.3%). The Large Early Decline, Moderate Early Decline, and High gradually decreasing groups were classified as successful. Successful trajectories were more likely to reach a target HbA1c of ≤ 7%, have shorter duration of pharmacist management, greater utilization of nutrition clinic services. The distinguishing factor between successful and less successful trajectories appears to be the progress made within the first six months of pharmacist management. Conclusion: Patients managed for diabetes can be grouped into distinct patterns of change in glycemic control over time. The first six months of the clinical pharmacist/patient relationship may be important in determining a patient’s overall success. Future research is needed to identify pharmacist interventions that increase the likelihood of achieving successful glycemic control trajectory patterns.