Ambient AI's impact on clinician experience varies

Though ambient AI was associated with improved clinician experience, new research shows that the positive results vary by sex and specialty.

The study published in JAMA Network Open examined clinician experience before and after implementing an ambient AI platform. These platforms use generative AI to listen to patient-provider conversations during a clinical visit and draft clinical documentation for the clinician to review. Though ambient AI proponents highlight the technology's ability to reduce clinician burnout, research shows mixed results, the study authors wrote. Thus, they decided to study the impact of an ambient AI platform piloted at California-based Sutter Health.

The health system launched the platform in April 2024, selecting 100 ambulatory clinicians to use it. The participants were asked to complete a survey three months before and after the implementation with questions about various factors that impact clinician experience, including burnout and time spent per week completing notes outside clinic hours. Of the 100 clinicians selected, 57 completed both surveys.

Although the difference was not statistically significant, researchers observed a small decrease in burnout, from 42.1% of clinicians to 35.1%.

Still, 54.4% of clinicians noted in the post-implementation survey that they spent 1 hour or less per week on notes compared with 14% who said the same in the pre-implementation survey. A majority of clinicians (71.9%) also agreed or strongly agreed that the ambient AI platform had increased their work satisfaction.

However, a higher percentage of primary care providers (85.8%) agreed or strongly agreed that their work satisfaction had increased with the implementation of the ambient AI platform compared to 36.5% of medical and 50% of surgical specialists.

Further, researchers assessed various EHR metrics, including time in notes per appointment and off-hour EHR activities, to assess the impact of the ambient AI platform. Of the 100 clinicians, 92 had EHR metrics available three months before and after the implementation. 

The study shows that the mean time spent in notes per appointment significantly decreased, from 6.2 to 5.3 minutes. Although off-hour EHR activities were similar before and after implementation, the mean number of characters per documentation and the mean progress note length increased slightly.

Researchers also observed that female clinicians had longer progress notes and documentation than male clinicians, but experienced bigger decreases in the time spent in notes per appointment. 

"Future research is needed to understand specific experiences of these subgroups and overall outcomes after widescale expansion given the context of a rapidly evolving technology, which shows promise in its ability to improve the joy of work for clinicians and their ability to connect with patients," the researchers concluded.

In addition to further analysis of how ambient AI tools impact specific clinician subgroups, healthcare leaders must also ascertain their cost impact.

Last month, a report released by the Peterson Health Technology Institute (PHTI) noted that while ambient AI scribes can help reduce cognitive load and burnout among clinicians, their financial impact remains unclear.

The report, which is based on insights from PHTI's AI Taskforce, states that ambient AI appears to have a largely positive impact on clinician burnout and allows clinicians to spend more time with patients, which in turn enhances patient experience.

However, there is mixed feedback from health systems regarding the impact of ambient AI scribes on patient volumes. It is also too early to assess the impact of ambient AI scribes on the accuracy of medical coding.

Anuja Vaidya has covered the healthcare industry since 2012. She currently covers the virtual healthcare landscape, including telehealth, remote patient monitoring and digital therapeutics.