In recent years, many healthcare places in the United States have started using Electronic Health Records (EHR). These systems help manage patient data and communication better. But they also bring new problems for doctors. One big problem is the large number of messages doctors receive through EHR tools. These messages include patient questions, prescription refill requests, test result follow-ups, and administrative notices. The number of messages keeps growing. This adds to doctors’ workload and can cause burnout. Burnout is harmful to both doctors and patient care.
Healthcare managers, owners, and IT staff need to know how artificial intelligence (AI), especially generative AI, can help with this problem. This article looks at how more EHR messages affect healthcare workers. It also reviews recent studies on using AI to manage these messages. Finally, it covers how AI and automation might reduce burnout in medical offices in the U.S.
Since the COVID-19 pandemic started, the number of EHR messages has grown rapidly in the U.S. For example, at NYU Langone Health, the In Basket messages increased by more than 30 percent each year during the pandemic. In some primary care offices, doctors say they get over 150 patient messages every day through EHR portals. These messages include appointment requests, prescription refills, test results questions, and other patient concerns.
Doctors must spend extra time after seeing patients to answer these messages. More and more, they have to do administrative tasks and message work instead of direct patient care. This causes high stress and burnout. Surveys show that about half of U.S. doctors feel burnt out, with messaging workload being a big cause.
Burnout can make doctors less happy at work. It can reduce the time they spend with patients and increase the chance of mistakes. Fixing this problem is important for healthcare groups that want good care and a healthy workforce.
A recent study at NYU Grossman School of Medicine looked at how AI could help doctors handle EHR messages. The study used a generative AI tool based on GPT-4. This AI made draft replies to patient messages inside the EHR system. The AI was added to EPIC’s In Basket feature to create first draft answers to patient questions.
Sixteen primary care doctors reviewed 344 pairs of messages—one written by AI and one by a human. The study found no big difference in accuracy, completeness, or relevance between AI and human replies. This means AI can meet the standards expected from doctor messages.
Also, AI responses were found to be more empathetic and positive than human replies. AI messages were 125 percent more likely to seem empathetic and 62 percent more likely to sound positive. Doctors rated AI replies about 9.5 percent better in tone and clarity than human ones.
AI replies were a little longer and used language at about an eighth-grade reading level. Human replies were closer to a sixth-grade level. This means AI might need to be improved to make messages clearer and easier for patients to understand.
Dr. William Small, the study’s lead author, said AI chatbots “could reduce the workload of care providers by giving quick and caring answers to patients’ concerns.” This shows that medical offices could use AI tools to help with patient messaging and ease the pressure on doctors.
Besides AI, healthcare systems have tried other ways to lower the number of EHR messages and reduce doctor stress. The American Medical Association (AMA) created a nine-step plan to manage messages better and lower workload.
Some important parts of the AMA plan are:
Atrius Health lowered the total message volume by about 25 percent over six years using these steps. This shows that combining better workflows with technology can help manage EHR inbox overload.
As EHR messages keep growing, many healthcare groups are using AI tools for front-office phone and messaging automation. Companies like Simbo AI provide AI solutions that answer patient questions from phone calls and digital messages automatically.
For medical managers and IT staff, AI systems can take on routine work like scheduling appointments, renewing prescriptions, and answering simple patient questions. This lets doctors spend more time caring for patients instead of doing paperwork.
Using natural language processing and generative AI models like the ones at NYU Langone, these tools create accurate and caring replies. They talk with patients in a friendly way without needing many people involved.
Examples of AI workflow automation are:
Using AI helps reduce burnout and improves clinic work. Dr. Devin Mann said, “GenAI message quality will soon match human responses in quality, style, and use.” This shows AI is becoming more accepted in healthcare communication.
AI can help, but there are still challenges before it can fully handle patient messages. AI replies can be longer and more complex than needed for some patients. This means AI needs to be trained better to keep language clear and simple and avoid confusion.
Healthcare groups must watch AI closely and check quality often. This way, errors or wrong clinical info don’t happen. Humans still need to review AI messages during early use.
Also, some doctors may not trust or feel comfortable with AI tools. It is important to include doctors in the AI setup and teach staff about AI’s uses and limits.
Healthcare leaders and IT staff in U.S. medical offices can improve efficiency by adding AI tools like those from Simbo AI. Combining these with the AMA’s workflow redesign plans can lead to fewer messages and less stress for doctors.
Steps to consider are:
These steps can lower after-hours work caused by messages, make doctors more satisfied, and keep patient contact strong.
Physician burnout from more EHR messages is a serious issue in U.S. healthcare. Recent studies show AI tools using advanced language models can create patient message replies as good or better than human ones in clarity, care, and tone. Along with system-level changes like cutting unnecessary messages, adding automation, and sharing work with care teams, AI-driven front-office tools offer real help for managing message workloads.
Healthcare managers and IT staff should think about adding AI and changing workflows to lower burnout, improve efficiency, and keep quality care in U.S. medical offices.
The COVID-19 pandemic led to a nationwide trend of increased use of electronic health record (EHR) tools, with many patients using these systems to ask questions, refill prescriptions, and review test results.
Physicians have seen a more than 30 percent annual increase in the number of messages received daily, with some receiving over 150 In Basket messages per day.
The overwhelming volume of messages contributes to physician burnout, with many spending long hours after work managing their In Basket messages.
The AI tool can draft responses as accurately as human healthcare professionals while being perceived as having greater empathy.
NYU Langone used a private instance of GPT-4, utilizing generative artificial intelligence algorithms trained on patient-specific data.
The study found no statistical difference in accuracy, completeness, or relevance, but AI responses were more understandable, empathetic, and positive.
AI responses were written at an eighth-grade level, while human responses were written at a sixth-grade level according to the Flesch Kincaid score.
With physician approval, AI message quality is expected to equal human responses in quality and communication style, enhancing efficiency in patient communication.
AI responses were found to be longer and more complex, indicating a need for further training to simplify language and improve clarity.
AI has the potential to significantly reduce the number of messages physicians handle by efficiently drafting empathetic responses to patient inquiries, thereby decreasing their workload and potentially reducing burnout.
Simbo is using AI agents to automate all phone related workflows. Connect now to know more.
Schedule AI Agents Demo Now© Since 2019, Simbo AI.
