AI enhances physician-patient communication, study reveals

UC San Diego Health, as one of the initial health institutions in the United States to integrate generative artificial intelligence (GenAI) into the Epic Systems electronic health record system, is at the forefront of digital health innovation. This integration facilitates the drafting of replies to patient inquiries, marking a significant step forward in medical communication.

The findings from a recent study conducted by the University of California San Diego School of Medicine, published in the Journal of the American Medical Association’s Network Open on April 15, 2024, reveal that while AI-generated drafts did not shorten the time physicians took to respond, they significantly reduced the mental effort required by doctors by providing an initial empathetic draft. This allows physicians to refine the message rather than composing one from scratch.

This pioneering study represents the first randomised prospective evaluation focused on AI-drafted messages within physician-patient communication, highlighting its potential to alleviate some of the cognitive demands on healthcare providers.

Christopher Longhurst, MD, the senior author of the study and a leading figure at UC San Diego Health, emphasised the utility of AI in addressing challenges within health systems, notably the surge in patient communications contributing to physician burnout. He pointed out that the lengthier AI-drafted messages suggest enhanced quality and that physicians have expressed appreciation for the assistance, which lessens their cognitive load.

The research underscores the role of GenAI in improving the quality, efficiency, and engagement of digital healthcare interactions. By reducing the workload on physicians, AI allows them to concentrate on the more intricate aspects of patient care, thereby aiming to mitigate burnout.

Ming Tai-Seale, PhD, MPH, the lead author and a professor at UC San Diego School of Medicine, described generative AI as a collaborative tool that assists physicians, who may receive around 200 messages per week, in overcoming writer’s block by providing a foundation for crafting empathetic responses to patients.

The enduring high demand for digital communication between patients and physicians, amplified by the COVID-19 pandemic, has led to the widespread use of patient portals like MyUCSDChart. These systems facilitate direct emails to doctors but also increase the pressure for timely responses, a demand that many physicians struggle to meet efficiently.

In response, UC San Diego Health initiated a pilot program with Epic Systems in April 2023 to employ GenAI for drafting responses to non-urgent patient queries. The AI-generated drafts include a disclaimer to inform patients of the AI’s involvement before the messages are reviewed and personalised by the responding physician.

Despite not reducing response times, the AI helps compose more detailed and compassionate replies, which are valued by patients. According to Marlene Millen, MD, another co-author and chief medical information officer for ambulatory care at UC San Diego Health, the AI’s ability to maintain consistent performance, regardless of the time of day, provides significant support in drafting empathetic messages that integrate relevant patient information.

While the study did not find time savings in message drafting, the qualitative benefits suggest a potential shift in healthcare communication practices. Further research is needed to evaluate patient perceptions of the enhanced empathy and detail in AI-assisted replies.

Since May 2023, UC San Diego Health and the Jacobs Center for Health Innovation have been rigorously testing GenAI models, exploring their safe and innovative applications in healthcare settings, thereby paving the way for future advancements in the field.