Hybrid ambient documentation decreases after-hours work, note delays for physicians, researchers report

Physicians often need to work outside of typical clinic hours to complete patient visit notes, and studies have shown this “pajama time,” along with other factors driven by documentation burden, increases rates of burnout.

In a new study, Mass General Brigham researchers have shown that generative artificial intelligence (AI), ambient documentation technologies that listen to appointments to generate visit notes, can play a major role in reducing physician “work outside of work.”

Their study, published in the Journal of General Internal Medicine, also shows that the technologies, which were paired in a hybrid program with a virtual human scribe, reduced delays in note completion, and increased physician productivity.

“Our findings directly inform ongoing efforts to deploy AI responsibly to improve clinician experience, reduce burnout, and enhance care delivery efficiency,” said lead study author Lidia Moura MD, Ph.D., MPH, director of population health for the Department of Neurology at Mass General Brigham.

The retrospective study included 181 primary care physicians and advanced practice providers (APPs) across 14 adult primary care practices at Massachusetts General Hospital, a founding member of the Mass General Brigham health care system. They were enrolled in a hybrid program where ambient documentation technologies recorded the note and generated text, and a virtual scribe manually entered the final note into the electronic health record for the clinician to review.

For their analyses, the researchers collected data on the total minutes of “work after work” a provider spent per day in the electronic health record outside of standard work hours, documentation delays (the proportion of visit notes that were not completed two days post-visit), and measures of clinical financial productivity, over an 80-day period (that included a 30-day adjustment period to the program).

The research team evaluated the program’s effects at 50 days post-implementation and found a sustained 41% reduction in after-hours work and a 66% reduction in delayed note closures.

Clinician financial productivity, as measured by work relative value units (wRVUs, a points-based measurement that reflects how much time and labor a physician puts into medical tasks), increased by 12%, suggesting efficiency gains that can translate into measurable value for health systems.

“This research reflects the value these technologies offer to health systems and provides actionable insights into optimizing AI tools to support clinicians by targeting their after-hours work, which ultimately can benefit their overall well-being,” said Moura, who is also director of the Center for Value-Based Healthcare and Sciences and Associate Professor of Epidemiology at the Harvard T.H. Chan School of Public Health and Neurology at Harvard Medical School.

This new study of a hybrid program served as a precursor to Mass General Brigham’s current efforts deploying and scaling AI-driven ambient documentation technologies. Mass General Brigham’s ambient documentation technologies initiative has grown to where they are now routinely used by more than 4,000 providers throughout the health care system.

Since the launch of a July 2023 proof-of-concept pilot study and expansion of the pilot to more than 800 providers one year later, researchers in the health system’s Digital Department have continued to study the tools and their impact on clinicians and system operations.

Future studies will track the technologies longer-term, including a comparison of hybrid and full-AI programs, to determine other measurable factors that might reflect burnout rates and clinical productivity, and whether the data they’ve collected to date may change over time.

Limitations in this observational study include selection bias as participation was voluntary. The study also lacked job design factors in the analyses, and therefore “work after work” hours may inadvertently capture individual schedule variations for some physicians. This study took place at one health system, and more research is needed to determine generalizability across other health care settings.

This research adds evidence that the technologies can significantly impact physician burnout. A study published in August that included more than 870 Mass General Brigham physicians using ambient documentation technologies reported a 21.2% absolute reduction in burnout prevalence at 84 days, and qualitative reports from physicians about the technologies restoring their “joy” in medicine by reducing the documentation burden.