Making pharmacogenomic-based prescribing alerts more effective: A scenario-based pilot study with physicians

Area of Science:

• Pharmacogenomics
• Clinical Decision Support Systems
• Personalized Medicine

Background:

• Personalized drug dosing (PDD) requires effective integration of pharmacogenomic (PGx) information into clinical workflows.
• Electronic health record (EHR) embedded clinical decision support (CDS) systems offer a potential solution for delivering PGx insights to physicians.

Purpose of the Study:

• To assess the communication effectiveness and clinical impact of a prototype PGx-CDS system in an EHR.
• To explore how physician and technology characteristics influence the system's utility and impact on prescribing decisions.

Main Methods:

• A pilot study using a prototype PGx-CDS system with simulated clinical case scenarios.
• Twenty-two physicians evaluated the system, with data collected on physician characteristics, technology implementation, and prescribing outcomes.
• A conceptual framework and measurement model were used to analyze communication effectiveness and clinical impact.

Main Results:

• Physicians reported a high perceived relative advantage (83%-94%) and utilized semi-active alerts (74%-88%).
• Confidence in prescribing decisions significantly decreased post-system use (p=0.02).
• Sixty-five percent of physicians reduced drug dosing, notably for capecitabine and mercaptopurine/thioguanine, aligning with evidence.

Conclusions:

• The prototype PGx-CDS system demonstrated potential for influencing drug dosing but requires optimization to enhance physician confidence and perceived usefulness.
• Future improvements should focus on PGx-CDS content, delivery methods, and tailoring to physician characteristics for greater communication effectiveness and clinical impact.