Quality Forum Review: Integrating Pharmacy into AI Innovation and Governance

The PQA Quality Forum Webinar is a regular, recurring series on healthcare quality topics with a focus on medication use and medication services. It is a forum for educating and engaging with PQA members and quality-focused healthcare professionals. 

The July 9 PQA Quality Forum Webinar welcomed Anthony DiDonato, Program Manager at Coalition for Health AI (CHAI) and Jeff Rochon, PharmD, FAPhA, FWSPA, Executive Director at Pharmacy Health Information Technology Collaborative (PHIT). Both organizations work to support the responsible use of AI technology in pharmacy and healthcare systems.

Deep Dive into CHAI: Work Groups of Interest to the Pharmacy Community

CHAI is home to numerous work groups that are open to various levels of contribution, with several new groups launching this year. Active and upcoming Work Groups from CHAI’s Use Case Work Groups and Cross-Cutting Work Groups are outlined below, with pharmacy-specific applications. PQA members are invited to engage with CHAI by sharing use cases with Work Groups, contributing expertise to the Agentic AI Skills group, and helping to shape the guardrails and frameworks being developed around AI.

Agentic AI Work Group

The Agentic AI Work Group seeks to build consensus-based protocols and expectations for how AI agents should behave within healthcare systems. AI agents are an emerging artificial intelligence technology that can perceive, reason and perform tasks semi- or fully autonomously with minimal human supervision. PQA members may have the opportunity to provide input regarding AI agents and pharmacy use cases in conversations with the Work Group.

Example use cases:

  • Medication adherence and refill-reminder voice agents
  • Prior authorization criteria matching for medications
  • Clinical decision support for drug-drug interaction and dosing checks
  • Formulary and contract search agents for health plans and PBMs

Once healthcare protocol language and expectations are consensus-defined, a subgroup may further build agent “skills” and develop guardrails (or guidelines for how an agent can operate) to be published and tested.

Example pharmacy skills and guardrails:

  • Medication reconciliation skills with safety guardrails
  • Drug-drug interaction & allergy checking guardrails
  • Adherence-outreach skill with human-escalation triggers

Ambient AI Work Group

Ambient AI is a technology that can be leveraged to detect information and generate insights based on the people around it. It is designed to function based on automatic triggers instead of active user inputs, such as a smart device automatically detecting elevated heart rate and sending subsequent notifications. Ambient AI may one day be deployed and used in practical settings such as emergency departments, pediatric units, behavioral health, and nursing workflows. The work group aims to develop implementation guidance that any organization may use, regardless of vendor or product.

Example pharmacy use cases:

  • Ambient documentation of pharmacist-patient counseling
  • Medication reconciliation during transitions of care

Post-deployment Monitoring Work Group

Consensus guidance is being developed for AI monitoring practices, which include auditing, logging, transparency metadata, and operational oversight to keep AI accountable over time.

Example pharmacy applications:

  • Monitoring drift in medication adherence prediction models
  • Audit trails for AI-assisted prior authorization decisions
  • Transparency metadata for medication-related chatbots
  • Ongoing performance tracking for AI feeding quality measures

AI Governance and Risk

CHAI has developed AI Governance Playbooks spanning four core domains of AI Policy, Organizational Structures, Organizational Resources, and Organizational Processes for healthcare organizations of all sizes who wish to implement AI. The Risk Categorization Tool (v3) is another resource for healthcare organizations and teams who wish to implement early AI governance.

Deep Dive into PHIT’s CAD Work Group

PHIT’s Clinical Workflow, Artificial Intelligence and Digital Innovation (CAD) Work Group aims to improve quality in patient care across the healthcare continuum by empowering pharmacists with responsible and trustworthy AI tools. The work group has previously published a resource titled “Artificial Intelligence Considerations for Pharmacy,” a document that outlines recommendations and resources around AI and machine learning applications in pharmacy dispensing and clinical services. Their upcoming publication titled “Overview of the Pharmacist’s Role Integrating Digital Health and AI Technologies” will include a JCPP Patient Care Process AI Use Case which identifies where AI could be appropriately utilized to increase efficiency in healthcare technology. Interested PQA members are encouraged to join PHIT’s work groups to join the conversation regarding AI in pharmacy technology environments.

You can listen to the full recording of this Quality Forum Webinar, a presentation with audience Q&A, on PQA’s YouTube channel. PQA members can access the presentation slides in the Member Resources Library.

,