Ambient AI Scribe Technology in Healthcare

There is a lot of buzz across industries about Artificial Intelligence (AI). CIOs and organizational leaders investigate and debate what is hype versus reality.

In healthcare, there is one AI solution that is gaining traction across organizations and appears to be delivering results- ambient voice technology or ambient AI scribes.

Ambient AI scribe technology leverages generative AI to listen, interpret, analyze, and organize conversations between patients and providers which are then inserted into the electronic health record (EHR). Why is that such a big deal? Let’s find out …

So, what is ambient AI scribe technology all about?

Ambient AI scribe technology records clinical conversations, typically via an application on a mobile device, to draft provider notes and patient visit summaries using generative AI. It combines speech recognition, natural language processing, and LLMS to transcribe and organize conversations into structured notes. Integration to a healthcare organizations EHR is key to maximizing the benefit. New capabilities are evolving at a rapid pace.

According to KLAS research, it is on track to become one of the fastest adopted technologies in healthcare.

What problem is it trying to solve?

A challenge facing healthcare organizations today involves physician and provider dissatisfaction with the amount of time spent documenting in the EHR (aka: documentation burden). For primary care physicians, they often site “pajama time” (time spent outside of 7am-7pm) catching up on admin tasks, such as EHR documentation, as a chief complaint.

The primary reasons organizations adopt AI scribe technology include promoting physician wellbeing, reducing physician burnout, enhancing the patient experience, and improved compliance/accuracy, among others.

How are organizations measuring success and what are the results so far?

Adoption has been rapid across healthcare organizations. While primary care and outpatient settings led the way initially in terms of adoption, this is now expanding into other specialties and inpatient settings.

Many organizations are focused on measuring:

·        Reduction in documentation time

·        Physician burnout rates

·        Accuracy and completeness of medical records

·        Time lag between patient encounters and note closure

·        Patient satisfaction

Early results have been encouraging.

Several studies noted improvements like:

·        Clinical note time decreased from 10.3 mins to 8.2 mins

·        “Pajama time” dropped from 50.6 to 35.4 mins.

·        One system reported 8% improvement in patient satisfaction scores.

·        Reduce clinician burnout with some health systems reporting up to 40% improvement.

·        According to a study published in the NEJM Catalyst, 56% of patients feel more engaged with physicians during visits, 84% of physicians felt using AI scribe had a positive impact on their visit interactions, with a notable reduction in minutes of pajama time.

What are the challenges and risks to be aware of?

·        The cost can be a challenge with most vendors using a per user/per month (PUPM) subscription fee model

·        Other challenges include integration issues and training and support costs.

·        Risks can involve clinical inaccuracies, over-reliance on AI, technical and integration issues and compliance with regulations and data protection laws

·       Like all AI solutions, it is important to have guard rails and governance in place, including understanding all compliance and regulatory requirements.

What’s next?

Many vendors have roadmaps in place to extend capabilities to include:

·        Revenue cycle coding

·        Deeper integration to the EHR to automate tasks (ex. Placing pending orders)

·        Multilingual capabilities

·        Added intelligence for more “ChatGPT like” capabilities.

·       Expanded capabilities for other specialties and departments.

Conclusion:

AI scribes continue to show great promise, and it is an area many healthcare organizations are seriously considering if they haven’t already implemented. A healthcare AI trend to watch as adoption and expanded capabilities continue to drive innovation and efficiencies in this space.

At Hoag, the results for our pilot are similar to industry results so far, with positive feedback from the pilot group physicians. We will evaluate expansion into other specialties and departments, pending conclusion of our pilot.

As we know, for AI in general it’s important to have the governance in place to effectively evaluate products and ROI to target your investments.

Having effective governance in place is also essential, specifically:

·        Identify how you will measure success. Define your key metrics and KPIs up front

·        Understand your ROI

·        Pilot before you invest

·        Understand and account for compliance and regulatory requirements

·        Define how you (or the vendor) will monitor models, specifically to assess fairness and bias mitigation, transparency and explainability of models, accountability, privacy and data protection considerations and security and risk management

Author:

Carmella Cassetta , Chief Information Officer, Hoag Health System

 Sources:

·        Beckers Health IT: The fastest –adopted tech in healthcare: Report, Naomi Diaz , March 26th, 2025

·        Beckers Health IT: AI scribes boos clinician efficiency: Study: Report, Naomi Diaz , Feb 20th, 2025

·        Peterson Health Technology Institute: Adoption of Artificial Intelligence in Healthcare March 2025

·        NEJM Catalyst Ambient Artificial Intelligence Scribes: Learnings after 1 Year and over 2.5 Million Uses; Published March 31, 2025; NEJM Catal Innov Care Deliv 2025;6(5)

·        KLAS Research

·        Peterson Health Technology Institute: Adoption of Artificial Intelligence in Healthcare, March 2025

Lance Ralls, Howard Miller, MBA, PMP, CISM, CRISC, Julie Shroyer, Lea Eriksen, Kevin Gray, George Sheth, MBA, FHIMSS, Kathy Linares, Jake Westphal, Michael Olinger, Greg Geary, CISM, Thomas Phelps, David O'Brien, MBA, CHCIO, HCISPP, Tina Machi, MSIT, Joel Manfredo, MBA, MS (Finance), Nancy Kim-Yun, Beth Hilbing, Todd Britton, EdD, Kathy Lomax, Rich Lindberg, Berhanu Tadesse, John Buccola, YuChih Liou, Carmella Cassetta, John Manzanares, Rachel Yegiaian, Cindy McKenzie, Tamara Armstrong, Grant Anderson, Garrett Zorko, Harry Cometa, Atul Anand, Jacqueline Toubes