AI in Healthcare

Reducing Clinician Burden with AI

Sergiy Sumnikov talks with Kimberly Noel about why clinical AI adoption starts with workflow, what meaningful clinician co-creation looks like, and how to reduce cognitive load without weakening clinical judgment. They cover evidence standards, human checkpoints, and monitoring after deployment.

Watch on YouTube Get the transcript

Clinical leaders

Identify where AI can relieve cognitive load while preserving clinician judgment and accountability.

Health innovation teams

Design tools around real care workflows and the people who use them every day.

AI and innovation teams

Set meaningful baselines, validate clinical value, and monitor systems after launch.

Meet the expert

Kimberly Noel

Kimberly Noel is Global Lead, AI Advocacy & Digital Health at Roche. A physician with experience in clinical medicine, digital health, and AI, she previously served as Deputy Chief Medical Information Officer at Stony Brook Medicine and founded Human in the Loop. She is also an editor at npj Digital Medicine.

Hosted by Sergiy Sumnikov
General Manager, Health Innovation at Halo Lab

Map the clinical workflow and identify the task that creates friction before choosing a model or feature. Agree on a baseline so the team can tell whether the intervention truly reduces work.

Separate routine, lower-risk tasks from decisions that need expert review. Put human checkpoints at consequential or irreversible decisions without adding alerts to every step.

Bring clinicians into design and testing as partners. Include patient, legal, and ethics perspectives, and make it clear who owns the decision when AI contributes to care.

Compare documentation time, cognitive load, and burnout alongside patient experience and clinical outcomes. Monitor both intended benefits and new work the system may create.

Keep it easy to question or override a recommendation. Train teams on failure modes, watch for over-reliance, and continue clinical validation and surveillance in production.