What a Digital Health Physician Learned While Recertifying His Board – Healthcare Blog
Author: Jean-Luc Neptune
I recently got the good news that I passed my recertification exam with the American Board of Internal Medicine (ABIM). As a little background, ABIM is a national physician evaluation organization that certifies physicians practicing internal medicine and its subspecialties (each other specialty has its own board certification body, such as ABOG for obstetrics and gynecology and ABS for surgery). Physicians practicing in most clinical settings need to be board certified to become credentialed and eligible to work. Board certification can be completed by taking a test every 10 years or participating in a continuing education process called a LKA (Longitudinal Knowledge Assessment). Instead of pursuing the LKA approach, I decided to take the 10-year test. For those ABIM certification doc colleagues who are wondering why I chose 10 years vs. LKA, I'd be happy to have a side discussion, but this is very much a career timing issue.
It's important to note that board certification is different from the USMLE (United States Medical Licensing Examination), which is the first in a series of licensure hurdles doctors face during medical school and residency and involves 3 separate exams (USMLE Steps 1, 2, and 3). After completing the USMLE steps, obtaining a medical license is a separate state-mediated process (I'm active in New York, not Pennsylvania) and has its own set of requirements that need to be met in order to practice in either state. If you wish to be able to prescribe controlled substances (opioids, benzos, stimulants, etc.), you will need a separate license from the DEA (Drug Enforcement Administration, a federal entity). In short, you need to complete a lot of training, score high on a number of standardized tests, and obtain a series of certifications (which cost a lot of money, by the way) to practice medicine in the United States.
What I learned while preparing for the ABIM recertification exam:
1.) There is so much to know as a doctor!
To prepare for the exam, I used the New England Journal of Medicine (NEJM) review course, which includes approximately 2,000 detailed case studies covering all subspecialty areas of internal medicine. If you consider that each case involves the mastery of dozens of pieces of medical knowledge, the exam requires doctors to memorize tens of thousands of different pieces of information for just one specialty (remember, the medical vocabulary alone contains tens of thousands of words). Furthermore, a single fact means nothing without mastering the basic concepts, models, and frameworks of biology, biochemistry, human anatomy, physiology, pathophysiology, public health, etc. Then, there's everything you need to know for your specific specialty: medications, diagnostic frameworks, treatment guidelines, and more. a lot of. There's a reason it takes the better part of a decade to gain competence as a physician. So whenever I hear a non-doctor say they've been reading XYZ and “I think I know almost as much as my doctor!” my response is always “No, you don't. Not at all. Not even a bit. Stop it.”
2.) As doctors, there is so much we don’t know!
What particularly struck me as I was doing the review was how often I came across cases or demonstrations like this:
- It's unclear what causes the disease,
- The natural history of the disease is unclear,
- We don't know how to treat this disease
- We know how to treat the disease, but we don’t know how the treatment works,
- We don't know which treatment is most effective, or
- We don't know which diagnostic test is best.
- Go on, go on, go on…
It is estimated that there are over 50,000 (!!) active journals in the biomedical field, publishing over 3 million (!!!!) articles every year. Despite this emerging knowledge, we still know very little about the human body and how it functions. I think some people think doctors are arrogant, but anyone who really understands doctors and the culture of doctors can tell you that doctors possess a deep sense of humility that comes from knowing how little you actually know.
3.) One day soon, computer doctors will definitely be smarter than human doctors.
The entire time I was preparing for the exam, I kept telling myself that nothing I was doing was possible with a sufficiently advanced computer.
If you abstract away what most doctors do (diagnose disease and prescribe treatment), it’s clear at this point in the history of artificial intelligence that computers will soon be able to do most of what doctors do.
Conceptually, diagnosis is very simple: gather information about a patient's presentation and evaluate complex patterns involving the patient's history, signs, symptoms, and various tests. While human doctors are capable of identifying hundreds or thousands of patterns, there are limitations to our abilities due to our limited memory, prior experience, and how we acquire information. However, existing AI systems, with access to virtually unlimited information and more powerful pattern recognition algorithms, will soon be able to identify disease patterns better than the best doctors.
Developing a treatment plan is also very simple: Based on the patient's characteristics, disease, nature/stage of the disease, patient preferences, etc., recommend the treatment that is most effective and least harmful as shown by the literature (clinical guidelines, peer-reviewed journal studies, etc.). As humans, we are limited in the journal articles we can read and the information we can store in our brains. AI systems can access the accumulated knowledge of all of humanity and will soon be able to review all literature at once to guide treatment decisions.
Recently published research has shown that AI systems can match or exceed the performance of human doctors. Many will argue that machines don’t really reason, and that’s currently true, but inference technology may not be far off. Given that these technologies are advancing at an exponential rate, it is clear that a significantly better machine will surpass the cognitive abilities of human doctors in a very short time (up to 10 years). I believe that one day soon, patients will ask their doctors “What does the AI system recommend?”
4.) What computers can’t do yet is “be human” (at least not yet).
What is often overlooked in studies showing comparability between computer and physician performance is that the computer works based on a well-summarized case presentation (like the ones I used to study for the committee) and all relevant data. What these studies ignore is that one of the most important roles of a physician is to communicate with another human being to obtain the information needed to make a diagnosis and recommend treatment. As physicians, we rarely receive a great summary that contains all the relevant information. The other person is often agitated, or under the influence of a substance, or lying, or unconscious. As human doctors, most of what we can do is use our human senses (sight, smell, touch, hearing – thankfully not taste) to put together a story to guide our judgement. A big part of medical training is learning about human psychology, human culture, and human history, which we then use to inform the science we know.
Another important aspect of being a human physician is our role as advisors, advocates, and care managers for individual patients and the broader patient community. Ultimately, patients need someone to help them understand a serious diagnosis or support them in making difficult choices about treatment options. The modern healthcare system has evolved into a transactional model where doctors and patients are often deprived of deeper human interactions, but new technologies offer opportunities that may reduce the administrative burden on doctors and patients so more time can be spent on therapeutic human-to-human interactions.
One day we will have machines technologically advanced enough to fully mimic humans (interestingly, the original Blade Runner Tyrell Corporation Nexus-6 “replica” exists in the fictional year 2019.) But for now, there's nothing better than humans.
5.) Technology can now help us become better doctors.
What a lot of people don't know is that the day-to-day job of being a doctor kind of sucks. Assume that for every hour of direct clinical care a physician spends an additional 2 hours on administrative tasks. Most doctors don't spend their careers entering data into poorly designed electronic medical records, waiting for prior authorization from insurance, or asking patients for the same information over and over again. I'm excited that my role at Commure gives me the opportunity to contribute to technology that makes the lives of doctors and patients better.
Environmental delineation is a transformative technology that helps physicians reduce the administrative burden of documenting care by up to 80%, thereby reducing physician burnout and allowing physicians to regain the joy of seeing patients. Co-Pilot technology puts all published medical research at doctors' fingertips, reminding me of how access to the Internet (and sources like UpToDate) 25 years ago changed the way we deliver care. Finally, Agentic AI helps reduce the “tedious” work of physicians by automating and routinizing repetitive tasks that do not deserve human attention.
I know that the introduction of new technologies has many people worried about the future of employment, and in these uncertain times, these are legitimate concerns. That said, we're not providing a lot of care because we don't have the resources, and I think the story over the next few years is going to be using technology to catch up with what we were supposed to be doing in the first place. I encourage my fellow doctors and sisters not to fight technology, but to work to make technology meet our needs. Modern electronic medical records were developed at the expense of physicians to make life better for other stakeholders who are not at the bedside. We cannot allow this to happen this time.
(AI Certification: I certify that this post was written without the use of any AI assistance, but with some editing by my very human wife.)
JL Neptune is an internist in New York City and executive medical director of Commure.
