Riskiest Doctor Specialties


Part of the appeal of becoming a doctor is job security – at least, that’s what everyone says, right? “There will always be more patients that need doctors than there are doctors,” they say. Well, this might not be the case anymore for some specialties.

The AAMC estimates that by the year 2034, we will experience a shortage of between 38,000 and 124,000 physicians; however, these shortages are not uniform across all specialties. Although fields such as primary care will likely continue to be in high demand for the foreseeable future, the future is not so certain for others.

With issues like oversaturation, mid-level encroachment, and advancements in artificial intelligence, the livelihoods of some specialties are not as secure as you might think. Here are the doctor specialties that are most at risk.


1 | Oversaturation

The first issue is oversaturation. Due to concerns regarding future physician shortages, there have been large expansions in the number of first-year medical school positions over the last couple of decades.

Since 2002, there has been an increase in allopathic medical school first-year enrollment from approximately 16,000 positions to over 22,000 positions – an increase of nearly 30%. New osteopathic colleges and expansion of existing schools have also increased the number of first-year DO students from approximately 5,000 in 2002 to now over 8,000. These increases in medical school capacity have been accompanied by expansions in many residency programs as well, albeit to a much lesser extent.

Although this is good news for addressing physician shortages in some fields, it is also creating issues in others. There are now some specialties that are training too many physicians and, as a result, it’s becoming increasingly difficult for physicians in these specialties to find a job.

Radiation Oncology

In radiation oncology, for instance, the number of positions offered annually in the match increased by 227% between 2001 and 2019. Additionally, advancements in technology and our understanding of cancer biology have allowed radiation oncologists to offer more effective therapies in fewer treatments than before.

As a result, the number of radiation oncologists entering the field is now higher than the demand and many new grads are reporting difficulties finding jobs after residency.

One radiation oncologist reported applying to over 50 jobs across the country and was only able to get interviews at three – all in undesirable locations. It will only get harder for new grads to find desirable jobs out of residency too as they will not only have to compete against their peers but also with more experienced radiation oncologists.

If the discrepancy between supply and demand continues, radiation oncologists will experience further tightening of jobs and downward pressure on pay.

Plastic Surgery

Plastic surgery is another specialty that is at risk of oversaturation – specifically in aesthetics. Although the demand for cosmetic procedures has generally been increasing year over year, plastic surgeons are starting to face increased competition from non-plastic surgeons.

There are many non-plastic surgeon physicians and mid-levels who want a piece of the action and attend weekend workshops to learn how to do botox, filler, liposuction, and even some surgical procedures. This is what leads to dangerous and life-altering complications, like when an OBGYN does a tummy tuck and the patient develops necrotizing fasciitis, also known as flesh-eating bacteria, or when an ENT-trained facial plastic surgeon does a thigh lift with similarly devastating complications.

Greed is driving many non-qualified practitioners to want a piece of the aesthetic pie, and it’s a major patient safety issue. After a botched job, the patient comes to an actual board-certified plastic surgeon to fix the damage, but many of these complications result in lifelong issues that are impossible to entirely reverse. I’ve heard too many of these stories from my plastic surgeon friends and colleagues that I’ve now lost count.

There’s a reason why plastic surgeons spend 6 or more years in training to perfect the nuances of these surgical and non-surgical procedures. They are far more qualified and perform these various procedures and do so more effectively and more safely.

However, it comes down to a marketing issue. Patients should seek properly trained plastic surgeons who are certified by the American Board of Plastic Surgery, which is a member of the American Board of Medical Specialties. There are several other similar-sounding organizations that are not recognized by the American Board of Medical Specialties that are deceiving uninformed patients.

As it stands now, the aesthetic market is becoming oversaturated, and only through concerted efforts to educate the public will plastic surgeons be able to effectively advocate for patient safety and reclaim volume.

Emergency Medicine

Perhaps the most surprising specialty at risk for oversaturation is emergency medicine. Although emergency medicine physicians have been integral during the pandemic, many new EM doctors are reporting difficulty finding jobs.

The number of accredited emergency medicine programs has nearly doubled over the last 15 years, going from 133 in 2005 to 265 in 2019. As a result, the demand for EM physicians is not increasing fast enough to keep up with the number of new doctors entering the field.

Surprisingly, the coronavirus pandemic has only added to this issue. In many areas of the country, emergency departments experienced significant decreases in patient volumes during the pandemic.

Many people were avoiding hospitals for fear of contracting COVID-19. The transition to remote learning and working from home also meant that fewer people were leaving the house, leading to fewer accidents. Many emergency departments were even forced to close their doors due to a lack of inpatient beds on overrun hospital floors.

Some emergency departments reported as much as a 40% decrease in patient volume early in the pandemic and some continue to be down as much as 20% compared to pre-pandemic patient volumes. This is a big issue as patient volume matters a lot more in emergency medicine than it does in other specialties. There aren’t as many high-cost procedures in emergency medicine, so EM physicians rely on seeing a high volume of patients as opposed to seeing just a few complicated ones.

It’s also very expensive to keep an emergency department open all day, every day, and have the necessary staffing and resources. Since physicians are the highest-paid members of the medical team, hospitals are incentivized to stretch each physician as far as they can.


2 | Mid-level Encroachment

This brings me to the next risk category. The growth of mid-level providers such as nurse practitioners, physician assistants, and CRNAs has exploded compared to physicians and many specialties are now at risk of mid-level encroachment.

Between 2016 and 2019 alone, there was a 34% increase in the number of employed nurse practitioners. At that rate, students starting medical school this year can expect over a quarter of a million more nurse practitioners to be employed by the time that they graduate.

In addition, many mid-levels are lobbying for independent practice so that they can practice without a supervising physician. As of August 2021, 24 states have granted independent practice to nurse practitioners and this number is likely to continue to increase over the coming years.

Many physicians are strongly opposed to this due to the difference in training and experience between mid-levels and physicians. For instance, a doctor that is fresh out of residency has more than 15,000 to 20,000 hours of clinical training. At the point of certification, a new nurse practitioner only has less than one-tenth of that at between 500 and 1,500 hours of clinical training.

Although that level of training may be sufficient to handle much of the bread and butter straightforward cases, things can quickly become dangerous without the greater expertise of a supervising physician. While mid-levels are great physician extenders when working as a part of a team, they are certainly not a replacement for physicians, and treating them as such is an issue of patient safety.

Emergency Medicine

Emergency medicine is one of the specialties at higher risk for mid-level encroachment. Mid-levels are often utilized to decrease costs by extending each physician. Now instead of needing four doctors to see one-hundred patients, they may be able to have 2 doctors and 2-3 mid-levels instead.

We are already starting to see the effects of this with increasing numbers of mid-level providers in the ED. Between 2012 and 2018, the total growth and use of nurse practitioners and physician assistants in the ED increased by 66% – and this trend is likely to continue.


Mid-level encroachment into the field of anesthesiology is also a growing concern. Many hospitals are now adopting an anesthesia care team model whereby an MD anesthesiologist simultaneously supervises multiple CRNAs – each in a different operating room.

This has raised concerns about decreasing employment opportunities for anesthesiologists as each anesthesiologist is stretched further by overseeing multiple mid-level providers.

Plastic Surgery & Dermatology

Similarly, plastic surgery and dermatology are at moderate risk for mid-level encroachment – particularly in aesthetics. With more states allowing independent practice for mid-levels, many are opening their own medical spas and performing minor cosmetic procedures.

Although dermatologists and plastic surgeons are better trained and qualified to perform these procedures, and discerning patients will understand this, mid-level run practices are often able to charge less for their services. Over the coming years, without a concerted effort to inform patients about safety, we are likely to see further mid-level encroachment.

What Specialties Are Safe from Encroachment?

That being said, there are many specialties that are relatively resistant to mid-level encroachment. For instance, surgical specialties tend to be on the safer side of the spectrum as mid-levels are typically limited to the first assist in the operating room. While some mid-levels may experience a strong Dunning-Kruger effect with regards to anesthesia, emergency medicine, or in other clinical settings, you’ll be hard-pressed to find one that tries to convince you they’re qualified to perform surgery.

Specialties that require deep knowledge also tend to be safe from mid-level encroachment as the specialized knowledge acts as a sort of moat. Specialties such as pathology and radiology would fall into this category given the depth of knowledge and the importance of an accurate diagnosis. Although radiology and pathology may be fairly resistant to mid-level encroachment, they may not be immune to the next point on our list: artificial intelligence.


3 | Advancements in Artificial Intelligence

Over the last decade, there have been significant advances in artificial intelligence. This is leading some to wonder if AI will soon make certain specialties obsolete – namely radiology, pathology, and dermatology.

Several new studies have come out that demonstrate computers can outperform doctors in cancer screenings and disease diagnoses. In one study, an algorithm designed to diagnose skin cancer at Stanford University had a success rate nearly identical to 21 board-certified dermatologists.

Another algorithm developed by Google using 42,000 patient scans from an NIH clinical trial was able to detect 5% more cancers than its human counterparts and reduced false positives by 11%. This is especially of interest as false positives are a big problem with lung cancer – made even more alarming by the fact that lung cancer is the leading cause of cancer death in the United States.

With such positive preliminary results, some people joke that we should stop training radiologists and pathologists now as AI will soon replace them; however, I don’t believe this to be the case.

Although these studies are promising, AI is not without its flaws. No matter how much the technology has advanced, it is still far from where it would need to be before it can be used without physician oversight.

Take EKGs for instance. Although the computer outputs its analysis of the EKG, its interpretation is far from perfect and the magnitude of the risk of being wrong is very high. This is why we need doctors and cardiologists to review them. They understand nuance and take into account the patient’s presentation and other clinical factors. As it stands, AI is simply unable to replicate this level of attention to detail.

Even if an algorithm is 98% accurate and can do so at a fraction of the cost of a physician with zero cost of replication, the effects of being wrong even 2% of the time can be devastating for patient care.

The interpretations made by radiologists and pathologists are the foundation of a large part of a patient’s care plan, so accuracy is incredibly important. An incorrect interpretation of a CT scan or biopsy specimen could be the difference between detecting a patient’s cancer early versus missing it and allowing it to grow and metastasize. You’ll always need a human with specialized knowledge to be able to recognize the nuance.

What I do think is of concern is if artificial intelligence can significantly increase efficiency among radiologists and pathologists. If you can effectively decrease the workload for each physician and make them more efficient, then theoretically they should be able to interpret more images and more slides in the same amount of time. In this scenario, we will likely see decreased demand for these physicians.

Although these issues are important to be aware of for anyone considering pursuing one of these specialties, it is impossible to know exactly how the landscape of medicine will change over the coming years. No matter what field you pursue, changes will occur, and you will need to adapt.

At the end of the day, I would argue that it is still much more important to choose a specialty that you enjoy and can see yourself doing for years to come than it is to choose one based on what we think will be the most stable in the future.

If you enjoyed this article be sure to check out our piece on the Best Doctor Lifestyle Specialties or Doctor Specialties with the Highest Hourly Rate.


Medical School vs Residency Comparison

The transition from Medical School to Residency is often misunderstood. Learn about the key differences between med school and residency so that you can best prepare yourself for both types of training.

Read More »

This Post Has 3 Comments

  1. Rev. Fr. Mark Lichtenstein

    98:2 correct/incorrect for AI diagnosis might be true, but the way the essay is written it feels like you are implying that physicians get the diagnosis right 100% of the time. I doubt that’s true. I doubt that humans get close to 98% correct on diagnoses. Autopsies aren’t as common as they once were, in part due to (over) confidence in MRI and CT etc. providing the right diagnosis, but when they were more common, I’m pretty sure I read somewhere that ~20% of the time the cause of death apparent on autopsy was not what the attending physicians thought was going on.

  2. Gail Neuman MSN FNP-C

    I am not a midlevel. I am a peer because I can do anything a doctor does but surgery. I had 47 years of healthcare as a nurse when I started nurse practitioner school in 2012. I have caught many physician errors in primary care. We don’t replace doctors so they shouldn’t feel threatened. There is a severe and worsening shortage of physicians in primary care and we NPs help alleviate that.

  3. R

    you’ve missed psychiatry under encroachment

Leave a Reply