Will Healthcare Let AI Fix The Slow-Motion Crisis of Physician Burnout?

When I ask primary care physicians about their biggest problems today, they share the excruciating pain they experience as the American healthcare system pushes them towards administrative work rather than letting them focus on true patient care and clinical excellence. 

Some doctors described themselves as “square pegs in a round hole” for bridling under the system’s bureaucratic constraints. Others feel stuck in a “transactional world” of recordkeeping. They yearn to spend more time providing care so that they can improve their patient’s health outcomes. “Let doctors be doctors and don’t expect them to perform like machines,” they say. But, with hundreds of conditions, ever-changing clinical guidelines, and thousands of predictive patient attributes to process within minutes, it has been humanly impossible for doctors to deliver truly comprehensive care to patients. 

AI offers a lifeline

One important suggestion I’ve heard repeatedly is to give doctors AI-enhanced solutions to deliver hyper-personalized care to the patient. “Not every check-in needs to be done by a PCP, RN, or PA,” a doctor told me. “AI can determine what the next check-in for this patient is about and who should do it. If AI detects the potential for a rapid disease progression in someone, should it be with a specialist? Or if it is an affordability problem, should it be with a social worker? Or is it something simple, with an AI chatbot? If it’s a PCP visit, AI can also recommend if it should be a tele-visit or an in-person visit.”

Similar AI-driven solutions could address other problems in healthcare today, too. “If we had a chronic condition score that popped up on the chart showing which conditions this patient is at risk for and why exactly, we would have a better idea of which patients would need more time and support,” another physician told me.

Others noted that AI could significantly enhance care by helping doctors keep up with the most recently published research in their fields. “Doctors try to stay up to date on evidence-based guidelines but, in reality, it’s a huge problem,” a doctor said. “When seeing more than 20 patients a day, physicians fall behind. Lots of clinically unwarranted variation in care is due to gaps in knowledge.”

Sadly, these complaints don’t surprise me. In the last decade, I’ve heard versions of them again and again as I’ve moved through the healthcare ecosystem as a chief clinical product officer at Humana and chief operating officer at UpToDate. I’ve seen how doctors give their best and keep treating patients because they care. I’ve come to the conclusion that these trends are unsustainable.  

Physician burnout: A slow-motion crisis

I see three root causes for physician burnout. We talk about one a lot. The second is only now receiving attention. The third is rarely discussed at all.

Tactical and administrative workload is the most well-known cause of physician burnout . More than 63 percent of physicians now experience burnout, according to the American Medical Association. Many doctors experience “emotional exhaustion” and “de-personalization” at least once per week, the Mayo Clinic Proceedings found. Recent Becker’s research shows that doctors are now more prone to suicide than any other professionals. We’re seeing a crisis in slow motion. 

Looking at how doctors spend their time explains the dilemma. On average, they see patients for only 18 minutes per visit, according to Medical Care. But for every hour they provide direct clinical face time, they spend nearly two additional hours on electronic health records, or EHRs, and other desk work, Annals of Internal Medicine researchers found. Doctors who have studied for years to save lives are spending more than half their time on data entry instead of focusing on improving holistic health outcomes for their patients. 

We’ve discovered that burnout among doctors is a moral crisis, too. To fully understand the depth of the crisis, look no further than a recent story in the New York Times that shook the medical world and personally touched me. The story described how doctors today in the US are experiencing the kind of “moral injury” that inflicts shell-shocked soldiers on the battlefield. The article quotes anguished doctors who “bemoaned having to fight with insurers about whether a person with serious illness would be preapproved for medication and the doctors who felt torn between their Hippocratic oath and the realities of making a profit from people at their sickest and most vulnerable.”

Now we must discuss the other driver of burnout in healthcare: The tasks we expect of doctors are humanly impossible. Three factors accelerate clinical conditions: social determinants of health, mental health conditions like depression, and lifestyles that include smoking, obesity, etcetera. Slowing disease requires physicians to deliver integrated care factoring in all these aspects of their patient’s history. The integrated system of care needs to factor in medical, social, and mental conditions combined with family histories, comorbidities, and other information. Primary care physicians are expected to play a critical “quarterback” role in this process. However, it’s humanly impossible to read thousands of data points within minutes to synthesize, understand, and deliver the best possible personalized treatment. This impossible expectation has fueled the crisis.

As Dr. Jordan Asher, chief clinical officer of Sentara Health, one of the largest not-for-profit health systems in the country, said, “Burnout is much more related to me not being able to do my calling as best as I can. Yes, administrative issues get in the way, but not being able to have all the information at my fingertips, trying to keep up with all the science that comes at me with ever-increasing frequency, and realizing I am a human and can’t know everything all the time are far more likely to be the root causes.”

Delivering this modern, integrated, and personalized healthcare to patients requires powerful new technological tools.

A new unified, AI-powered intelligent cockpit for doctors  

We need to confront physician burnout with a better appreciation for the three causes of burnout. Doctors’ frustration stems from the myriad bureaucratic responsibilities that consume their clinical time and negatively impact their experience and patient outcomes. To target the root causes of this situation, we need to pursue the latest technological approaches that are designed with our physicians’ experience and holistic health outcomes in mind, specifically generative AI combined with predictive AI and integrated data. We have talked about adopting new technology to revolutionize doctors’ experience and primary care for decades. Now AI has the potential to do it. 

An AI-driven intelligent platform could help doctors in the same way that pilots operate through a cockpit to control their aircraft. A streamlined clinical cockpit in one platform would give them the right data at the right time personalized to every patient they serve. The readouts on this cockpit would focus the doctor’s attention squarely on the safe health journey of patients while minimizing time spent on the administrative demands of employers and insurers. AI would parse thousands of data points and the latest guidelines, predict what needs to be done next, and augment their care intelligence to suit each patient. Doctors would no longer be expected to do the humanly impossible to deliver the best possible integrated care. Airline pilots don’t use one cockpit for taking off, another for landing, and a third to monitor air traffic all while expected to keep up with the latest FAA regulations. They’ve got all the essentials at their fingertips in one place. Doctors, more than ever, need the same kind of centralized, intelligent tool. 

Instead of contending with unfiltered mountains of data that they see every day, physicians want to access that data in an intelligent clinical application that helps determine the right care, at the right time, in the right channel for each patient.. As the cockpit metaphor suggests, this approach would use technology not as a substitute for doctors – it’s not an autopilot – but a means to empower them. Doctors wouldn’t use it otherwise. That’s why the American Medical Association recently  that it is formulating recommendations for deploying AI to augment – not replace – physicians in clinical settings. 

We can’t afford not to

Why not use AI to transform electronic record keeping and other responsibilities from an annoyance that frustrates doctors to a vehicle that seamlessly orchestrates patients’ health journey? Why not design the user experience to satisfy the daily needs of burned-out physicians and their patients? The outcome would be a considerably streamlined, singular clinical workflow engineered to provide the appropriate care, precisely timed and tailored for the individual, administered through the most suitable channel that aligns with their unique persona, a comprehensive approach covering the healthcare journey from cradle to grave. 

An AI cockpit would need to adhere to HIPPA and other privacy rules and security requirements, of course. But it would likely follow those rules more, not less, closely than the layers of legacy systems whose backdoors are often the culprits in data leaks and other issues today. Because humans are behind it, without proper training, AI can fall prey to biases and ethical lapses, too. Broad generative AI can create so-called hallucinations or misinformation. These risks are real. Doctors, however, have been mitigating all sorts of risks while adopting new technologies since the dawn of time. That’s why the American Medical Association said doctors would be central to its forthcoming recommendations. 

Those are open questions that will be answered. In the meantime, research shows that AI can save time without compromising quality. A study in Jama Network Open, for instance, showed that AI cut 18 percent of the time necessary for gastroenterologists to answer clinical questions based on medical histories as accurately as they would without AI. That’s a game-changer for burnt-out doctors. 

Bringing it all together 

In what ways would an AI-enabled cockpit help doctors address day-to-day demands on their time while increasing access to high-quality patient care? 

Longitudinal Data – Healthcare data is currently too siloed across claims, labs, EHR, social determinants of health, remote patient monitoring data, etc. An AI-enabled cockpit could amass the necessary longitudinal patient data to help understand the 360-degree aspects of their care. This approach makes life easier for doctors, but, more importantly, it also unlocks enormous opportunities for doctors to work with patients more closely and holistically and with a clear view of the intertwined causes of the patient’s health challenges. 

Predictive AI – Foreseeing outcomes based on historical data, this AI type uses machine learning techniques to analyze past data, detect patterns, and use mathematical models to predict future events and behaviors. For example, longitudinal data and AI algorithms together can help physicians craft treatments that are hyper-individualized, including medications, dosages – all important aspects of clinical decision-making quickly support – at a broad and deep level that is almost impossible for a human to accomplish fast enough given the hundreds and millions of data records they would need to parse to garner such intelligence. 

Generative AI – Generative AI and predictive AI serve distinct yet interconnected functions. Generative AI synthesizes existing information to summarize or create new content or data. It uses large language models (LLMs) and algorithms to understand and mimic data structure and patterns to generate output that is novel but still within the confines of its learned data patterns. This particular skill makes generative AI exceptionally beneficial for emulating interactions with physicians. It can automate repetitive tasks like scheduling routine follow-up appointments, lab tests, and prescriptions. 

Clinical collaborations – I believe healthcare needs the human touch. Patients – and doctors, too – crave human interaction. The true solution to helping doctors excel clinically while reducing burnout comes from helping them collaborate with patients and each other – not replacing them. By reducing all three causes of burnout and providing powerful tools to foster more and better contact between people, AI-enabled cockpits widen the lines of communication between doctors, their patients, and their colleagues. One can envision, for example, doctors building on their work together – absent the burnout – to develop new ideas that could help healthcare providers, technology companies, and policymakers further revolutionize various aspects of care. 

If we drive innovation by integrating and using all the above, we can not only reduce burnout but also change the trajectory of healthcare positively. 

Overcoming the slow-motion crisis afflicting America’s doctors will require the healthcare system to move much faster and be a lot more open-minded about AI innovation. It’s crucial to tackle all three root causes of physician burnout with a deeper understanding of their ethical struggles, innovative technology centered around revolutionizing the care they deliver and providing comprehensive health outcomes for their patients. We’ve been discussing this issue for years. Now is the time to take action. Generative AI, continued advancements in predictive AI, and integrated data present us with an opportunity to significantly alter the course of the healthcare industry. Let’s together move with a greater sense of urgency. 

About Deepthi Bathina
Deepthi Bathina is the CEO of RhythmX AI, a new health AI company, owned by the SAIGroup which includes Concert AI and Symphony AI

ai
ehrs
machine learning
remote patient monitoring
healthcare
medicine
medical