The technology won’t be valuable if it isn’t interoperable
Editor’s Note: Missy Scalise, MD, is an internal medicine physician, associate program director of the IM residency program, and chair of the Clinician Well-Being Committee at Ascension-St. Thomas in Nashville.
The medical profession faces a looming crisis on two related fronts: clinician burnout is at an all-time high, driving many to leave medicine or seek early retirement, while a shortage of clinicians choosing to practice medicine (specifically, primary care specialties) is causing patient access imbalances, especially in lower-income areas across the country.
Generative AI can significantly improve the quality of life of clinicians and help make primary care more appealing. Its ability to quickly and accurately document visits is crucial, but there are difficulties in integrating this technology into the current clinical platform.
Since electronic medical records (EMRs) became standard with the Health Information Technology for Economic and Clinical Health (HITECH) Act of 2009, clinicians have spent more and more time on documentation and other administrative tasks, which in turn contributes to decreasing clinician well-being. The integration of AI into EMRs will free clinicians from these tasks and allow us to use our knowledge and energy on the part of medicine we enjoy most: Connecting with patients, thinking about their problems, and helping solve them.
Before adopting an AI tool for EMR documentation, I would spend at least two hours per clinic day documenting patient visits, often during my free time (that’s a lot of “clicks” per day). On the weekends, I would wake at 5 a.m., before my children were up, to finish my documentation.
Now, with help from AI, I can sleep AND still spend time with my family.
Missy Scalise, MD, is an internal medicine physician, associate program director of the IM residency program, and chair of the Clinician Well-Being Committee at Ascension-St. Thomas in Nashville. Photo courtesy Dr. Scalise.
After getting my patients’ permission, I can activate an AI clinical assistant and then speak to my patients naturally. The tool, listening in the background, will pull the most relevant parts of the conversation and generate notes, which are then pushed into the EMR. Instead of staring at my computer while speaking with my patients, I can maintain eye contact with them the entire time. After we talk, I review the AI-generated note, editing a few words (turning plain language into medical terminology, for instance). My notes are automatically uploaded into the EMR with one click, and I am done.
In my estimation, this technology has reduced my paperwork load by 80% to 90%. This allows me more time to focus on patients as well as spend more time with my family. I’m not exaggerating when I say it’s changed my life.
If these AI tools are adopted at scale, I believe clinician efficiency and well-being will improve across the board. With increased well-being, clinicians are less likely to make mistakes and more likely to be engaged, thereby improving healthcare outcomes.
However, not all EMRs allow for smooth integration with AI-powered tools. Support for the integration of third-party applications ranges from not working at all, to only enabling schedule access, to allowing the two-way sharing of information between tools and the EMR in real-time.
EMRs are complicated software that must prioritize the privacy and safety of the patient data they contain, and thus third-party integrations are complex and often protracted. But these legacy EMRs need to adapt to the times. Integration of AI tools must be a priority going forward. At the same time, AI vendors must find ways to partner with EMR providers to build the level of integration they need.
The commitment to integrating these tools, regardless of the time and resources required, is a non-negotiable step toward better healthcare. Solutions that can fully integrate with EMRs promise enhanced patient care through more accurate and comprehensive data and increased efficiency in clinical workflows. This level of integration is more than a technological advancement, it’s a critical part of an evolution in healthcare.
AI has shown me a better way to work, and there’s no going back. I now advise residents to ask about AI tools while negotiating for positions. Anecdotally, I’ve heard from other doctors pushing their employers toward more open EMRs so that they, too, can leverage AI-powered tools in their day-to-day work. One gastrointestinal practice, for example, shifted from a gastroenterology-specific EMR to a specialty-agnostic EMR because the latter allowed for integration with third-party tools.
As the U.S. healthcare system faces shortages and clinician burnout, adopting technology can act as a salve for many of its ills. I’ve experienced firsthand how using AI tools improves my workflow and makes me a better doctor. As more doctors learn about these tools, they will surely push for their integration, and EMR companies and vendors alike must act swiftly to listen.
KEY TAKEAWAYS
AI technology can help health systems and hospitals reduce physician stress and improve clinical and administrative outcomes
Not all EHRs are able to seamlessly integrate third-party apps, making it difficult for the technology to be effective
Healthcare leaders need to pressure EHR providers to improve interoperability to ensure that AI can be used to its best potential