Burned Out Physicians: AI is Here to Help
For physicians, it’s been a brutal decade working in U.S. healthcare. In many ways, they’re at the crucible of everything that’s wrong with our fractured, uncoordinated, dysfunctional system.
They’ve borne the brunt of sweeping regulatory changes, disruptive new technology systems, confusing payment model transitions, a global pandemic that strained them to the breaking point, and, more recently, have assumed daunting responsibilities around population health and health equity. And through it all, their time with patients has decreased while their time handling the business and administrative side of healthcare has ballooned.
Fortunately, I believe that real help for reducing the burden on physicians is finally here. And it’s come in an unexpected package – Artificial Intelligence has arrived.
The Burnout Crisis
Physician burnout peaked in 2021, at the height of the global pandemic, when 63% reported burnout, up from 38% the year before.
More recently, the 2023 Medscape physician burnout and depression report entitled, “I Cry but No One Cares” found that 53% of surveyed physicians experience burnout while 23% experience depression. 62% of those physicians said their burnout had lasted longer than a year. The AMA and a 2023 Physician Foundation report found similar numbers.
The personal and physical impact on physicians is hugely concerning and even tragic (including health and relationship problems and increases in substance abuse and suicide) while the impact on care affects us all.
It is debatable that burned out physicians are prone to make more errors and face challenges in working collaboratively with others on their care teams. However, they actually experience less empathy and engagement with their patients. Importantly, burned out physicians are also more likely to quit their jobs or the practice of medicine entirely, at a time when physician shortages are ratcheting up the pressure on healthcare organizations everywhere.
All of this profoundly impacts care quality, outcomes, experience, and costs.
We Know Who Did It
Systemic problems always have complex interwoven causes. But in the case of physician burnout, the biggest culprit is relatively obvious and easy to identify.
Physicians are telling us quite plainly that they are overwhelmed by administrative tasks. In the Medscape report, 61% make that claim.
Recommended by LinkedIn
According to Becker’s, physicians spend an average of 15.5 hours per week on administrative paperwork with 9 of those hours devoted to EHR documentation. They’re scheduling appointments and referrals, recording notes, and clicking boxes on charts to ensure they’re reimbursed correctly.
Technology was supposed to improve workflows and processes to enhance productivity but the EHR systems have done the exact opposite. Designed to support billing systems, they’re simply ungainly and inefficient at the care side of the equation where physicians are forced to spend many hours completing documentation. And they also fail to provide physicians the insights, relevant data, or records for evidence-based best practices that could actually help them while engaging with patients.
If we could take the EHR documentation burden away from physicians, we would significantly free up their time, reduce burnout, and help them focus on patient care.
So, How Do We Fix It?
Over the past decade, a patchwork digital tech sector has arisen to remove pop-ups, reduce alerts, automate entries, speed up searches, and tie processes together. But it’s asking a lot of physicians to add to their technology load with new tools and interfaces.
Unless those tools make their interface as easy as talking to another human.
Human scribes have proven helpful in reducing physician administrative burden. According to the AMA, having a scribe in the room can reduce burnout by as much as 27%. But this is not as easy a solution to implement as one might think. Have you seen the average size of the examination room? Getting another person in there makes for a crowded elevator. Then, there’s the intimacy problem. Health conversations can be difficult with another person listening in and taking notes. Not to mention the challenge of getting enough trained people available whenever the physician needs them.
Alternatively, a well-trained AI model is exceptionally good at processing the conversation between a physician and patient, transcribing it, ordering the conversation into appropriate notes, performing the necessary documentation for claims, initiating referrals for lab tests, images, specialists, and so on.
Plus, an AI model takes up very little space in the room and will be an unnoticed presence!
Few technological advances have taken the world by storm like generative AI. The reason is simple. Unlike other technology interfaces, AI uses the most basic one of all – human language. Anyone can tap the power of generative AI simply by asking questions. In turn, generative AI models can take in human language and instantly organize the information, initiate processes, augment gaps, conduct searches, and generate insights.
That’s what makes AI so useful to physicians – not just a lift for administrative burden, but a support without parallel. It’s the most powerful, useful, and transformative digital technology healthcare has ever seen. Its adoption rate will be head-spinning. The early adopters of AI will outpace their peers and these physicians and their patients will be the beneficiaries.
Empowering Physicians to Build Wealth Beyond Medicine | Real Estate Fund of Funds Manager
6moWhile AI will make us more efficient, my only concern is that it is really just going to enable hospitals to make us more “efficient” hamsters on a wheel, i.e. see 80-100 patients a day instead of the “comfortable” 20-25.
Perceptivity, Problem-Solving, Pattern Recognition 💭💻📊| DATA COLLECTION | DATA MANAGEMENT remote only
6moI recently had appointment with my PCP and she had an AI device with her that recorded her notes so she wouldn’t have to do them at the end of the appointment. I thought that was so cool. I think one of the main complaints people have about doctor’s visit is not having enough time to talk with their doctor, so hopefully if this device is used more, we will see less long wait times to see the doctor and an improvement in the patient-doctor connection and communication. Also I hope this allows doctors the opportunity to leave their office on time so they don’t have to stay late to finish their patient’s notes.
Vice President, Chief Medical Director of Quality, Performance and Innovation, CHRISTUS Health- RETIRED
7moThanks, Abhinav. Well articulated about a symptom but I still believe that the main driver of burnout is the fee-for-service payment system keeping, especially primary, care on the treadmill just to financially survive. Your solution can help reduce some of the pressure but since most physicians are now competing with or employed by large groups requiring productivity, I believe that is the key driver. We need to be in a pay for performance/outcome value-based world so that primary care can concentrate on the patient and not the next task, regardless of what it is. Thanks for listening. roger
Principal leader in AI , Data and Scaling| Pytorch, Python, NLP, Robotics | Manage ML teams to build incremental ROI
7moNice article Abhinav! Definitely articulated the pain point from a very optimistic viewpoint. Some of the challenges that i think exist are 1) source of notes for training the ai- are the notes optimized to get maximum payments from payers or are they representative of true clinical setup to better patient care? Data quality is very important. 2) security and privacy- models should be deployed in the same architecture where data is stored. Models leak so much information. 3) make sure the burden of hallucinations from gen ai models do not negatively impact physicians workload to correct them.
Business Development Associate | Urban Company | Building supply strategies to ensure coverage of demand by > 98%
7moAppreciate the insight! Yet, can AI truly grasp the complexities of patient interactions and emotional nuances?