AI-Powered Pharmacy Benefit Optimization: A Framework for Value-Based Care

Abstract

Pharmacy benefits accounts for a significant share of health care spending, and recent calls for a reduction in pharmacy benefit plans’ administrative and drug acquisition costs have intensified interest in optimizing these benefits for all stakeholders. However, existing approaches mainly reduce spend without considering quality or access. Artificial intelligence opens new pathways for pharmacy benefit optimization that seek to improve cost, quality, and equity simultaneously. A conceptual framework is proposed that connects AI methods to multi-domain optimization goals through an integrated architecture of data, models, decision-making interfaces, and feedback loops. Potential value outcomes include reduced total cost of care, improved medication adherence and clinical endpoints, increased patient satisfaction, and strengthened equity; primary and secondary metrics are defined along with proposed data sources. A range of implementation considerations encompasses governance, regulation, ethics, change management, integration with existing workflows, and alignment with the priorities of pharmacy benefit managers and health systems.

Growing interest in the application of artificial intelligence (AI) technologies in health care is revolutionizing many aspects of clinical management. Unfortunately, pharmacy benefit managers (PBMs)—the companies that administer the pharmacy benefits of the majority of American health insurers—are facing increased scrutiny. The criticism centers on the perception that PBMs are collecting excessive administrative fees, increasing the spread between the pharmacy acquisition costs of drugs and the reimbursement paid by insurers, and lacking functional transparency. As a result, pharmacy benefits are being optimized mainly for cost reduction, rather than for simultaneous improvements in quality and equitable access. The integration of artificial intelligence into PMB operations represents an opportunity for more effective exploration of the tradeoffs among spending, quality, and accessibility.