HCPLive Five - Internal Medicine Updates at ACP 2025

Thousands of internal medicine specialists gathered in New Orleans for the 2025 American College of Physicians (ACP) Internal Medicine Meeting to share clinical insights, examine emerging data, and attend hundreds of sessions.

As part of HCPLive’s on-site coverage, the Internal Medicine Team conducted expert interviews and reported on timely topics across hospital medicine, artificial intelligence, health policy, and more. In this edition of the HCPLive Five, the team highlights key conversations on opioid-sparing regimens in hospitalized patients, the clinical risks and benefits of AI tools, bias mitigation in AI modeling, AI's diagnostic performance in urgent care, and the ACP’s call to action on national health policy issues.

Validating Opioid-Sparing Pain Management Regimens, with Angel Goenawan, MD

Angel Goenawan, MD, a hospitalist at Bayhealth in Delaware, discussed a meta-analysis of eight randomized clinical trials comparing opioid-sparing regimens to traditional opioid-based pain management in hospitalized patients.

The analysis showed that patients on non-opioid regimens had better pain control and used significantly lower total doses of opioids across both surgical and nonsurgical groups.

Balancing Benefits and Risks of AI Mitigation of Clinician Burden, with Ellen Gelles, MD

Ellen Gelles, MD, from the American College of Physicians, discussed how AI tools can reduce clinician workload by handling tasks like ambient documentation and message triage.

She noted pilot studies showing improved efficiency but emphasized the need for data transparency and monitoring to avoid unintended consequences like bias or safety lapses.

Reducing Racial Bias in AI Models, with Nawsabah Noor, MBBS

Nawsabah Noor, MBBS, presented on an AI-driven tool that modifies image saturation to improve dermatologic diagnosis accuracy in patients with darker skin tones.

Early results suggest that this intervention enhances visual clarity of skin conditions and could reduce diagnostic disparities across racial groups.

AI Bests Clinicians in Urgent Care Treatment Recommendations, with Zehavi Kugler, MD, and Ran Shaul

Zehavi Kugler, MD, and Ran Shaul discussed a study showing that AI-generated recommendations for virtual urgent care visits were rated as optimal 72% of the time, compared to 61% for clinician recommendations.

The AI system also produced fewer suboptimal recommendations, suggesting a role for decision support tools in improving care quality.

ACP Urges Concern for Health Policy Issues, with Shari Erickson, MPH

Shari Erickson, MPH, from the American College of Physicians, highlighted key health policy threats including prior authorization burdens, data privacy concerns, and clinician burnout.

She referenced ACP survey data indicating high levels of frustration among internists and warned of potential legal action to protect access to evidence-based care.

Kugler and Shaul's disclosures include K Health. Goenawan, Gelles, Noor, and Erickson had no relevant disclosures to report.