How AI Will Bring Joy Back into Clinical Practice

Did you know that post-Covid, physician burnout has reached epidemic levels? Here's how AI could be the game-changer.? In 2013, during the throes of implementing Electronic Health Records (EHRs) in the US, Sinsky published an important paper entitled “In Search of Joy in Practice: A Report of 23 High-Functioning Primary Care Practices.”? The study found that physician burnout was a growing problem at that time.? They made site visits to 23 high-performing primary care practices to determine the best practices to mitigate burnout. They determined that high-performing practices distribute functions among the team, use technology to their advantage, improve outcomes with data, and make the job of primary care more feasible and enjoyable. Innovations identified included (1) proactive planned care, with pre-visit planning and pre-visit laboratory tests; (2) sharing clinical care among a team, with expanded rooming protocols, standing orders, and panel management; (3) sharing clerical tasks with collaborative documentation (scribing), non-physician order entry, and streamlined prescription management; (4) improving communication by verbal messaging and in-box management; and (5) improving team functioning through co-location, team meetings, and workflow mapping. In essence, these best practices represented a shift from a physician-centric and physician-dependent model of workflow design to one that is patient-centric, team-based, distributed, driven by robust communication, and designed around a shared-care model. This approach allows physicians to return to what is most meaningful and enjoyable: spending most of their time directly caring for patients. This factor was emphasized in a recent article, “Rekindling Joy in Medicine Through Thoughtful Communication: A Practical Guide” by Drs Sanft and Winer. They concluded that creating space for more moments of feeling deep interconnectedness with patients and colleagues can rekindle feelings of joy in oncology practice.? The common theme of both articles is that reducing non-value-added administrative work for physicians and enabling all members of the team to participate in care and work “at the top of their license” is critical.? Advanced technologies can now greatly help achieve this aim.

With the maturation of Artificial Intelligence (AI) and the advent of Generative AI (GenAI), we may finally have the tools to have a meaningful impact on this important problem. We are entering the age of “cognitive computing”.? The concept that is important to appreciate is that GenAI establishes the computer as a thinking machine with a massive knowledge base that understands your language, learns with feedback, and can create new outputs based on your direction.? This capability has never existed before.? This capability automates many intellectual tasks that previously could only be completed by humans.? This opens the door for important use cases that reduce administrative burden for physicians and all members of the clinical team.? One of the highest value use cases is the use of virtual scribes.? The GenAI based virtual scribe listens to the physician-patient encounter and transcribes the interaction.? The scribe automatically generates a note that the physician can review, edit and approve.? Early assessments show this can save as much as 3 hours per day on medical documentation alone.? Importantly, this allows the physician to focus on the patient, as they did in the “old days,” instead of typing on the computer.? But this is not the transformative element.? The ability to introduce automated intelligence into the encounter in real-time is what has true transformative potential.? Because virtual scribing happens in real-time, and the system can multitask, it can simultaneously deploy Al algorithms to analyze patient data, evaluate the conversation, search the medical literature, and provide contextual information.? It can provide real-time intelligent clinical decision support, feedback, and analysis.? With this level of real-time support, you can imagine everyone on the team being able to work at a higher level of their license.? The gains in enhanced physician-patient interaction, clinical decision-making, quality, patient safety, and efficiency could be dramatic.? Another important use case is monitoring for inappropriate, biased and non-empathetic interactions.? A company called mPathic demonstrated how culturally responsive AI could improve mental health equity for racial and ethnic minorities by incorporating cultural context in care and addressing challenges in creating sensitive AI.? Reducing mental health inequities can save billions annually by decreasing emergency department visits and managing chronic conditions more effectively. The estimated burden of mental health disparities is $477.5 billion per year.? Leveraging cognitive computing in the office visit will completely change the game in healthcare.

Beyond helping in the actual encounter, AI will become the new team member for population health management because of cognitive computing.? One of the aspects of bringing joy back into practice is team-based care.? This approach takes as much of the clinical care and administrative burden as possible off of the physician, and distributes it among team members.? This gives the physician space to build the deep interconnectedness with patients that makes the practice of medicine so special and enjoyable.? Many of the team-based tasks associated with high performing practices, such as proactive planned care with pre-visit planning, sharing clinical care, panel management, non-physician order entry, and streamlined prescription management, can be accomplished with an intelligent AI agent.??

Of course, the use of AI in healthcare has several significant challenges. One of the biggest challenges is ensuring that AI systems are fair and unbiased, as biases in algorithms can exacerbate disparities in patient care. Developing AI systems that are transparent, accountable, and respectful of patient privacy is essential. Additionally, clinicians must be properly trained to understand how AI works and how to use it effectively in their practice. There is also the challenge of integrating AI systems into existing healthcare infrastructures, which can be complex and costly. Ensuring interoperability between AI tools and current electronic health records (EHRs) is crucial to avoid disruptions in workflow. Furthermore, there are ethical considerations, such as maintaining the human element in patient care and addressing potential over-reliance on AI for decision-making. Finally, there is a need for continuous monitoring and updating of AI systems to ensure they remain accurate and effective as medical knowledge and technologies evolve.

Despite these challenges, the potential benefits of AI in healthcare are enormous. AI has the potential to make clinical practice more efficient, effective, and enjoyable for everyone involved, and most importantly improve quality, efficiency and experience for patients.