In the United States, healthcare providers face a big increase in digital communications with patients. The COVID-19 pandemic made patient portals, telehealth, and electronic messaging more common. Research from UC San Diego Health shows that doctors get about 200 patient messages every week. These messages include appointment requests, test result questions, medication concerns, and follow-ups that need careful replies.
While electronic health records (EHRs) and patient portals make access easier, they also increase the mental work for doctors. Replying to messages quickly and thoughtfully can be tiring, especially after long days. This extra communication is one reason why many physicians feel burned out in U.S. healthcare.
To help with this problem, UC San Diego Health started using generative AI models in their Epic Systems EHR platform in April 2023. Generative AI means the system can create text based on the information it gets. Here, the AI writes draft replies to patient messages that doctors can read, change, and personalize before sending. This way, AI acts as an assistant, not a decision maker.
Studies published in the Journal of the American Medical Association’s Network Open give some early proof about how AI helps communication. Although AI drafts do not make doctors reply faster, they do reduce the mental strain by offering ready-made, thoughtful message drafts. These drafts are often longer than usual replies and contain more details and caring words that both doctors and patients like.
Christopher Longhurst, MD, executive director at UC San Diego Health and lead researcher, says AI helps doctors avoid “writer’s block” and lets them spend more time on patient care rather than paperwork. Marlene Millen, MD, a co-author, notes that AI can create kind responses even when doctors are tired, which may help prevent burnout.
A key point from the UC San Diego study is that AI does not take over doctors’ judgment or personal touch. Instead, AI supports doctors by writing first drafts that they can change. This keeps the doctor in charge of communication while making writing easier.
The AI system openly tells patients that the message was first written by AI but then checked by a human doctor. This honesty helps keep patient trust by making sure no one thinks the messages are fully automatic without any check. Being clear about AI’s part in patient communication is important to keep trust and empathy between doctor and patient.
Burnout is a big problem in U.S. healthcare, especially for doctors in primary care and those with many patients. By helping write detailed replies, AI reduces the mental effort needed to answer patient messages. Doctors do not have to start messages from zero, which is hard after busy days.
The study shows that even though AI does not speed up replies, it helps doctors write more thoughtful messages. This can make care better and improve how doctors feel about their jobs. AI can keep the message quality high, even when there are many messages to answer.
Even though AI is helpful in drafting messages, there are real worries about using it too much in healthcare communication. Some reports warn that if AI replaces human contact or if doctors rely too much on technology, patient care might feel less personal.
One issue is the “black-box” problem. This means doctors and patients may not understand how AI creates its answers. This can cause a loss of trust because people might feel the answers are not controlled by a human.
Bias in AI systems is another worry. If AI is trained on limited or biased data, it might treat some patient groups unfairly. Healthcare groups need to make sure AI tools are fair. Doctors must keep control to adjust messages so they fit different patients well.
For those managing medical practices, getting AI tools to work well with daily tasks is very important. Automating routine communication must fit smoothly, cause little trouble, and be easy for busy doctors to use.
Using generative AI inside the Epic Systems EHR at UC San Diego Health shows the benefits of embedding AI in the clinical workflow. Doctors do not have to leave their EHR or learn new apps. This smooth fit helps doctors use AI tools better.
Automation can also be used for other front-office jobs like scheduling, reminders, and taking calls. For example, Simbo AI uses AI to answer phone questions. This can cut wait times and free staff and doctors to do more important work.
Simbo AI’s system understands natural language and answers phone calls with caring responses. By managing routine calls, AI lowers work for staff and makes it easier for patients to get help. Combining phone AI with message drafting creates a fuller system that supports both office work and clinical communication.
When U.S. medical practices think about using AI in communication, they should consider several things:
Using AI phone automation like that from Simbo AI adds to message drafting by handling front-office calls. This cuts the need for staff to answer common questions and gives patients fast, correct information.
As more patients use digital communication, it becomes clear that new tools are needed to balance speed and care quality. Generative AI shows how automating certain tasks can reduce mental strain on doctors. It allows for longer and more kind messages without losing the human touch.
The UC San Diego Health pilot program is an example of putting AI into daily work while keeping transparency and doctor review. These efforts point to a future where AI works with healthcare workers to improve efficiency but also protect the important doctor-patient bond.
Still, research must continue to learn how patients feel about AI-written messages and how it affects care results. Healthcare managers and IT teams must keep watching how well these tools work and whether they follow ethical rules. They need to keep setting standards that put patients and doctors first.
Using AI together with healthcare workers in the U.S. can help lessen doctors’ workloads, improve message quality, and add useful automation. By carefully adding AI into current clinical and office tasks, medical practices can handle more patient messages, lower burnout risk, and keep meaningful patient connections that matter for good care.
The study focuses on the use of generative AI to draft compassionate replies to patient messages within Epic Systems electronic health records, aiming to enhance physician-patient communication.
The study found that while AI-generated replies did not reduce physician response time, they did lower the cognitive burden on doctors by providing empathetic drafts that physicians could edit.
The senior author is Christopher Longhurst, MD, who is also the executive director of the Joan and Irwin Jacobs Center for Health Innovation.
It evaluated the quality of communication and the cognitive load on physicians, suggesting that AI can help mitigate burnout by facilitating more thoughtful responses.
AI is seen as a collaborative tool because it assists physicians by generating drafts that incorporate empathy, allowing doctors to respond more effectively to patient queries.
The COVID-19 pandemic led to an unprecedented rise in digital communications between patients and providers, creating a demand for timely responses which many physicians struggle to meet.
Generative AI helps by drafting longer, empathetic responses to patient messages, which can enhance the quality of communication while reducing the initial writing workload for physicians.
A greater response length typically indicates better quality of communication, as physicians can provide more comprehensive and empathetic replies to patients.
The study suggests a potential paradigm shift in healthcare communication, highlighting the need for further analysis on how AI-generated empathy impacts patient satisfaction.
UC San Diego Health, alongside the Jacobs Center for Health Innovation, is testing generative AI models to explore safe and effective applications in healthcare since May 2023.