Leveraging the Clinician’s Expertise with Agentic AI

TL;DR: Ambient AI assistants are already saving clinicians significant time by automating documentation, with accuracy rates in the “high 90s”. Agentic AI promises to further transform healthcare by creating unified workflows across disconnected systems, reducing the administrative burden that contributes to burnout affecting nearly half of US physicians.

Administrative documentation occupies at least eight hours of a US physician’s 59-hour work week, according to the American Medical Association (AMA), with many spending additional time on electronic health records outside working hours. This burden contributes to burnout affecting nearly half of US physicians, but ambient AI assistants are already offering relief.

Current Impact of Ambient AI

Ambient AI assistants, which gained mainstream traction in 2023, can record, structure, and summarise patient encounters in real time. Dr. Ed Lee, Chief Medical Officer at health care AI company Nabla, reports: “For complex patients, it could take me up to 45 minutes to complete the documentation. Nabla makes that task infinitely better and allows me to give each patient my full, undivided attention.”

The technology achieves accuracy rates in the “high 90s” according to Lee, with clinicians maintaining responsibility for reviewing and approving final records. Contrary to initial concerns, patients respond positively—they appreciate receiving their physician’s full attention and hearing technical language that signals comprehensive care.

The Agentic AI Evolution

Agentic AI represents the next phase, promising unified workflows across healthcare’s fragmented systems. Alexandre LeBrun, co-founder and CEO of Nabla, explains: “Rather than forcing doctors and nurses to click through a dozen separate systems, our platform will provide specialised, customisable, and composable agents that turn disconnected tools into a single, continuous workflow.”

LeBrun envisions a cardiologist using voice commands to have one agent pull vitals, lab results, and imaging reports from the EHR, another generate patient summaries, and a third flag missed follow-ups—all before the patient arrives.

Implementation and Trust

Applying AI to high-stakes healthcare requires balancing productivity with accuracy. “Trust is everything in medicine,” says LeBrun. Nabla uses techniques like adversarial training models to check outputs and defaults to conservative responses, optimising precision over recall.

Successful integration requires education and cultural change. Lee emphasises: “We need to start with the basics of AI, making sure everyone understands what it is and how it works. Not how the programming takes place but more around what it can do, what it can’t do, the risks and pitfalls.”

Looking Forward

The technology’s potential lies not in replacing clinicians but in supporting them to focus on patient care. “AI should focus on supporting decisions and automating everything downstream,” says LeBrun. “The first role of AI is to get physicians back to the state where they make medical decisions.”

By simplifying complex systems and automating routine tasks, agentic AI holds promise for further augmenting ambient AI assistants whilst improving the process without requiring wholesale replacement of legacy infrastructure.

Source Attribution:

Share this article