5 Takeaways from Newsweek's Digital Health Care Forum 2025

Digital Health Care Forum 2025: Practical Lessons Health Systems Can Apply Now

Health systems are under pressure to improve outcomes, reduce friction for clinicians, and keep care affordable. At the Digital Health Care Forum on September 16, leaders from major systems and technology companies shared what's working, what isn't, and how to make AI and digital programs deliver real value.

Below are the most actionable insights, distilled for executives, clinical leaders, and informatics teams.

Trust is the foundation for AI in care

• K Health's Allon Bloch was blunt: without trust, there is no adoption. If your AI is not embedded in clinical workflows and validated by clinicians, patients will default to unreliable sources. Put AI where decisions happen and prove it adds value.
• Nabla's Daphne Groll emphasized customization and transparency. One-size-fits-all models erode confidence. Calibrate models to local workflows, communicate openly about limitations, and co-design with end users.
• Maria Middelares' Peter Dierickx showed how accessibility builds confidence. Make tools fast, user-friendly, and responsive to clinician feedback. Happy clinicians lead to better patient trust.

Governance: fewer trailblazers, more sherpas

• URAC's Dr. Shawn Griffin: bring clinical leaders (nursing, physicians, pharmacy, lab) into AI oversight. The higher the clinical risk, the tighter the monitoring and auditing. Treat model output like any high-stakes clinical input.
• University of Rochester's Dr. Gregg Nicandri: audit AI-generated documentation rigorously. Don't repeat mistakes learned from dictation tools-errors can look authoritative.
• Aneesh Chopra (former U.S. CTO) reminded everyone: experiment, measure, and follow existing rules. AI doesn't void prior regulations; it must operate within them.

Helpful reference: Patient right of access under HIPAA remains core to any data strategy. See HHS guidance on individual access rights for details. Read more

Data strategy: clarify rights, reduce friction

• Chopra highlighted the difference between a patient's right to their data and institutional authorization to reuse it. Build patient-controlled data flows and transparent consent. This supports second opinions and consumer-grade decision support, responsibly.
• Aim for open data models with aligned incentives between payers and providers. Without trust and shared benefits, AI efforts stall or inflate costs.

Vendor selection: security and scale over hype

• Kaiser Permanente's Dr. Daniel Yang: most startups cannot deliver at enterprise scale with required reliability, security, and compliance. Favor partners who fit your strategy and prove value at your volume.
• Microsoft's Joe Petro: security, reliability, and supply-chain validation are non-negotiable. You're not just evaluating one company-you're inheriting their entire ecosystem.

From pilots to production: focus on outcomes

• Palantir's Drew Goldstein: effective operational AI pairs complete clinical context with subject-matter experts closest to the problem. Success criteria should be user outcomes, not "we went live."
• Hospital for Special Surgery's Dr. Ashis Barad created a safe "LAB" environment so staff can learn tools, share prompts, and validate models without fear. This reduces legal/compliance anxiety and speeds real adoption.

Virtual and home-based care: make it seamless

• MetroHealth's Dr. Nabil Chehade: aim for the "Disney monorail" experience-patients move smoothly between virtual and in-person care without getting lost. Don't spend endlessly on patient navigation; design systems that are intuitive by default.
• UMass Memorial Health's Dr. Eric Alper showed how a digital hub can enable remote clinical services and hospital-at-home programs with strong results on mortality, safety, and readmissions.

Operations and workforce: give time back to care

• Jefferson Health's Dr. Baligh Yehia announced a commitment to return 10 million hours to patients by 2028. The path: reduce documentation burden, cut low-value paperwork, and deploy automation where it frees clinical focus.

Rankings and transparency

• Statista and the research team underscored that transparent, refreshed methodologies matter. Annual updates reflect current performance and encourage continuous improvement.

Key takeaways to align your digital strategy

• Build accountability: set owners, timelines, and clinical metrics-and ship improvements.
• Operate transparently: explain models, data use, and performance to clinicians and patients.
• Standardize data: reduce patient confusion and improve your view of patient needs.
• Remember: AI is only as useful as its users. Train and support the people using it.
• Invest with purpose: define clinician-cleared, mission-focused goals before buying tools.

What to do next (a practical checklist)

• Stand up an AI governance council with CNIO/CMIO leadership and clear charters for risk tiers.
• Pick 2-3 high-friction workflows (e.g., prior auth, discharge summaries, in-basket) and set measurable targets for time returned to clinicians.
• Adopt patient-first consent and data access practices; make "download and share" easy.
• Demand vendor proof: security certifications, bias testing, model performance at your scale, and ROI within 6-12 months.
• Build a "safe sandbox" for clinicians to test tools, share prompts, and validate outputs.
• Publish a brief model fact sheet for any AI tool in production: purpose, limits, monitoring plan, and rollback triggers. Consider aligning with the NIST AI Risk Management Framework. Reference

Upskill your teams

AI succeeds when clinicians and operators know how to use it. If you need structured learning paths for roles across your organization, explore curated options here: AI courses by job.

The signal from the forum is clear: make AI trustworthy, govern it like clinical tech, measure outcomes, and give clinicians time back. Do that, and digital health supports care-not the other way around.