Access to clinical pharmacy services in a pharmacist-physician covisit model

A pharmacist-physician covisit model in which patients see both a pharmacist and physician on the same day was established in a primary care practice. Previously, patients were seen in a referrals-based model in which providers referred patients for clinical pharmacy services on a different day.

To assess access to clinical pharmacy services in a pharmacist-physician covisit model compared to a referrals-based model.

A retrospective chart review was completed for patients who were seen by physicians on pre-specified half-days of clinic before and after implementation of the covisit model. Covisit model half-days between June 29, 2018 and September 30, 2018 and matched half-days from 2015 were included. Charts were reviewed to determine if patients scheduled to see the physician would benefit from clinical pharmacy services, including being seen for chronic disease management, eligible for a Medicare Annual Wellness Visit (AWV), prescribed medications that required counseling, had an adverse medication-related event, or had adherence concerns. Those eligible for clinical pharmacy services were further reviewed to determine if the patient interacted with a pharmacist within three months of their visit.

Prior to implementation of the covisit model, 123 patient visits were completed on the pre-specified half-days. Of these, 61 patients (49.6%) were deemed eligible for clinical pharmacy services. In the covisit model, 149 patients were seen by the physician, of which 69 patients (46%) were eligible for clinical pharmacy services. More patients in the covisit cohort went on to interact with a pharmacist (56 patients, 81% vs. 10 patients, 16%, adjusted OR = 32.98, 95% CI [8.89-122.39]). The most common reasons patients were identified for clinical pharmacy services were eligibility for AWV, hypertension, and diabetes.

A pharmacist-physician covisit model significantly increased accessibility to clinical pharmacy services compared to a referrals-based model.