healthsystemCIO
Penn State Health’s DeFlitch & Hijjawi Say Past Lessons Can Help Guide Today’s AI Implementations
Why It Matters
Understanding the barriers and best practices for AI integration is critical as hospitals face a flood of new tools that can impact patient safety and clinician burnout. By applying lessons from earlier EHR rollouts, health systems can deploy AI more responsibly, ensuring it improves care delivery rather than creating new inefficiencies.
Key Takeaways
- •Clinician awareness and early involvement crucial for AI adoption
- •AI should augment, not replace, physician decision‑making
- •Human‑centric design reduces workflow friction and administrative burden
- •Past EHR lessons warn against unchecked AI implementation
- •Rapid AI evolution challenges informatics teams’ tool expertise
Pulse Analysis
In this Health System CIO Show episode, Penn State Health’s chief and associate CMIOs, Dr. Chris DeFlich and Dr. Shadi Hijawi, unpack findings from their recent systematic review on physician perceptions of artificial intelligence. They stress that clinician awareness, early engagement in design, and clear legal frameworks are foundational for AI buy‑in. By linking these insights to real‑world pilots at Penn State, they illustrate how a patient‑centric mindset can turn abstract algorithms into actionable bedside tools, echoing broader trends in clinical informatics and decision‑support system adoption.
The conversation draws a direct parallel between today’s AI rollout and the chaotic early days of electronic health records. Both technologies arrived with lofty promises but often added administrative strain when poorly integrated. DeFlich emphasizes that AI must serve as a supplemental “second brain,” offering relevant data without dictating care. Hijawi adds that human‑centric design—tailoring interfaces to clinicians’ workflows and minimizing clicks—can transform AI from a disruptive novelty into a productivity enhancer. This perspective aligns with industry calls for workflow‑aware AI that supports, rather than replaces, physician judgment, reducing burnout and improving patient outcomes.
Finally, the duo acknowledges the speed at which AI capabilities evolve, outpacing traditional informatics expertise. They advocate for iterative pilots, continuous feedback loops, and cross‑disciplinary collaboration to keep pace with daily model updates. By leveraging lessons from past EHR migrations—such as the need for robust training, clear governance, and realistic expectations—Penn State Health aims to embed generative AI responsibly within its seven‑hospital system. Their pragmatic roadmap offers a template for other health systems seeking to balance innovation with safety, ensuring technology ultimately serves the core mission: people taking care of people.
Episode Description
Most health systems are deploying AI tools faster than they can evaluate them. Penn State Health’s informaticists explain why piloting with discipline and designing for the clinician workflow separates lasting gains from expensive disappointments.
Source: Penn State Health’s DeFlitch & Hijjawi Say Past Lessons Can Help Guide Today’s AI Implementations on healthsystemcio.com - Interviews & Webinars with Health System IT Leaders
Comments
Want to join the conversation?
Loading comments...