In the rapidly evolving landscape of healthcare technology, the integration of artificial intelligence (AI) into clinical workflows offers promising avenues to address persistent challenges, notably clinician burnout. A groundbreaking large-scale observational study recently published in JAMA explores the real-world impact of AI-enabled ambient documentation systems, commonly known as “AI scribes,” on clinician time management and productivity across multiple U.S. healthcare institutions. This investigation, conducted over a period surpassing two years and encompassing data from five hospitals, provides crucial insights into how AI scribes modestly optimize electronic health record (EHR) interactions and influence clinical practice patterns.
EHR documentation, a fundamental yet time-consuming component of modern healthcare delivery, has been consistently implicated in the rise of burnout among clinicians. AI scribes automate the generation of draft clinical notes by passively capturing and processing clinician-patient interactions through ambient technology during consultations. After the visit, clinicians review and finalize these AI-generated notes, potentially mitigating documentation burdens. Despite their intuitive appeal and prior evidence linking ambient documentation to reductions in burnout, the mechanisms by which AI scribes affect clinician workflows have remained underexplored in large, diverse clinical settings—an investigative gap this study addresses.
The collaborative research effort, led jointly by teams from Mass General Brigham and the University of California, San Francisco, meticulously tracked ambient documentation utilization by over 1,800 clinicians employing AI scribes, contrasting their experiences against 6,770 control clinicians operating without ambient technology at the same institutions. The analysis revealed that AI scribes correlated with average daily reductions of approximately 13 minutes in total EHR usage and 16 minutes specifically in documentation time. These time savings translate to relative decreases of around 3% and 10%, respectively, metrics that, while modest, are statistically significant and indicative of meaningful workflow efficiency gains.
The study further identified a slight but notable increase in clinical productivity attributable to AI scribe adoption, with clinicians conducting roughly half an additional patient visit per week on average. This productivity uptick, although quantitatively small, suggests that the time salvaged from administrative tasks may enable clinicians to accommodate more patients, addressing workflow demand in high-volume practices. Crucially, the researchers posit that these incremental improvements in time management and productivity represent only a portion of the broader benefits of AI scribes, necessitating deeper exploration of cognitive and behavioral changes induced by ambient documentation technologies.
An intriguing dimension of the findings is the heterogeneity in technology adoption and benefits among clinical subgroups. Primary care physicians, advanced practice providers, and female clinicians demonstrated the most pronounced reductions in documentation and EHR use, highlighting potential demographic and specialty-specific receptivity or workflow compatibility with AI scribes. Moreover, users who integrated AI scribes into more than 50% of their patient encounters experienced disproportionately larger time savings, with twice the decrease in overall EHR time and thrice the decline in documentation duration compared to less frequent adopters. Despite these advantages, only about one-third of clinicians adopted the technology with such intensity, underscoring challenges in user engagement and adaptation.
From an economic standpoint, the study documented statistically significant increases in revenue linked to enhanced patient throughput; however, these financial gains were nominal, averaging around $167 per clinician per month. Importantly, the time clinicians spent interacting with EHRs outside formal work hours did not differ meaningfully between AI scribe users and controls, suggesting that ambient documentation tools primarily optimize work-time efficiency without exacerbating off-hours administrative burden. These nuanced observations provide a foundation for future work to disentangle the complex interplay of technological adoption, workflow dynamics, and clinician well-being.
Senior author Dr. Rebecca G. Mishuris, Chief Health Information Officer at Mass General Brigham, emphasizes that the measurable reductions in documentation time noted by the study likely do not fully explain the significant decreases in clinician burnout previously observed in smaller cohorts. She articulates the necessity for further research to elucidate how ambient documentation reshapes clinician approaches to patient care and cognitive load. This perspective invites a broader consideration of AI scribes beyond mere time-saving devices to potential facilitators of enhanced clinical experiences and satisfaction.
The collective findings emerged from the Ambient Clinical Documentation Collaborative (ACDC), a multi-institutional initiative dedicated to studying AI scribe implementation and impact. This alliance harnessed granular usage data and clinician-reported metrics to achieve a comprehensive picture of ambient documentation in diverse clinical environments, reinforcing the importance of multisite research in validating technology benefits and limitations. The study’s robust design and scale lend critical credibility to discussions about integrating AI scribes into mainstream healthcare practice.
Dr. Lisa Rotenstein, study lead and associate professor at UCSF School of Medicine, accentuates the imperative of real-time evaluation of ambient documentation technologies as their adoption accelerates nationwide. She underscores the value of fostering clinician comfort and proficiency with AI scribes through training and support, aiming to maximize the tools’ benefits. This human-centered approach reflects an understanding that AI integration success hinges not only on technological capabilities but also on clinician engagement and workflow harmonization.
Disclosure statements reveal that Dr. Rotenstein maintains connections with several AI health technology firms, including grants, advisory roles, and travel support, reinforcing the transparency and ethical rigor underpinning the study. Funding for the research was secured from the Advancing a Healthier Wisconsin Endowment, a philanthropic gift from Kathy Hao to establish an AI impact monitoring platform at UCSF, and a significant grant from the Agency for Healthcare Research and Quality (R01HS029470), demonstrating strong institutional commitment to advancing AI clinical applications.
As ambient documentation systems become increasingly prevalent, this seminal study offers a critical evidence base for healthcare organizations contemplating AI scribe adoption. By quantifying tangible workflow improvements alongside nuanced considerations of clinician experience, the investigation charts a path toward optimizing healthcare delivery through intelligent automation. Future research endeavors are positioned to build upon this foundation, exploring longitudinal impacts on burnout, patient outcomes, and healthcare economics.
The implications of these findings reverberate beyond immediate clinical environments, signaling a transformative moment in how artificial intelligence can be harnessed to empower clinicians, streamline administrative burdens, and ultimately foster a more sustainable healthcare ecosystem. As AI scribes evolve to encompass enhanced natural language processing and context-aware capabilities, the potential to redefine clinician-patient interactions and documentation fidelity remains vast and compelling.
Mass General Brigham, a leading integrated academic health care system, alongside UCSF, continues to pioneer research at the intersection of technology and medicine, catalyzing innovations that strive to alleviate physician burnout and elevate care delivery standards. By embracing rigorous, collaborative studies like this multisite evaluation of AI scribes, these institutions contribute substantially to the evidence-based deployment of transformative digital health solutions across the United States and beyond.
Subject of Research: People
Article Title: Changes in Clinician Time Expenditure and Visit Quantity With Adoption of Artificial Intelligence-Powered Scribes: A Multisite Study
News Publication Date: April 1, 2026
Web References: https://jamanetwork.com/journals/jama/fullarticle/10.1001/jama.2026.2253
References: Rotenstein L, et al. “Changes in Clinician Time Expenditure and Visit Quantity With Adoption of Artificial Intelligence-Powered Scribes: A Multisite Study.” JAMA DOI: 10.1001/jama.2026.2253
Keywords: Generative AI, Artificial intelligence, Health care, Health care delivery
