Physician-dispensing as a determinant of clinical and process measurements in patients at increased cardiovascular risk: A cross-sectional study in Swiss general practice

Aims: In some healthcare systems, physicians are allowed to dispense drugs; in others, drug-dispensing is restricted to pharmacists. Whether physician-dispensing affects patient health is unknown. Thus, we aimed to investigate associations between physician-dispensing and clinical and process measurements in patients with selected long-term conditions indicating increased cardiovascular risk.

Methods: Retrospective cross-sectional study in 2018 based on data from electronic medical records of 22405 patients (73.6% physician-dispensing) in Switzerland with medications for diabetes mellitus, arterial hypertension, or lipid-related disorders. We used multilevel regression models to determine the associations between physician-dispensing and clinical measurements (glycated hemoglobin [HbA1c], systolic blood pressure [sBP], low-density lipoprotein cholesterol [LDL-C]) or process measurements (number of annual clinical measurements, consultations, and drug prescriptions).

Results: Median (interquartile range) HbA1c value was 6.8% (6.3-7.5) both for the physician-dispensing and pharmacist-dispensing group, sBP was 137 (126-150) and 136 mmHg (126-149), and LDL-C was 2.3 (1.8-3.0) and 2.5 mmol/L (1.9-3.2). After adjustments, the physician-dispensing group had 4% lower LDL-C levels (p = 0.041), 12% more frequent HbA1c measurements (p = 0001), 16% higher annual consultation rates (p < 0.05 for all conditions), and equal number of different drugs, compared to the pharmacist-dispensing group.

Conclusions: We found no relevant differences in selected clinical measurements between physician- and pharmacist-dispensing, and mixed results in process measurements. Our results do not indicate that one drug-dispensing channel is superior to the other.