Job Market in Radiology

[metview]

Hello Radiology friends,

I come from the rad onc forum. Given the current issues in our field with overexpansion and lack of jobs, I was wondering how Radiology fixed their issues? Correct me if I'm wrong, but I think there was concern about an oversupply of Radiologists/lack of jobs about 10 years ago? What was the cause of these issues and how did you guys turn it around? It seems like your guys' job market is very robust.

Thanks

---

Members don't see this ad.

 

[maxxor]

Hello Radiology friends,

I come from the rad onc forum. Given the current issues in our field with overexpansion and lack of jobs, I was wondering how Radiology fixed their issues? Correct me if I'm wrong, but I think there was concern about an oversupply of Radiologists/lack of jobs about 10 years ago? What was the cause of these issues and how did you guys turn it around? It seems like your guys' job market is very robust.

Thanks

We didn’t expand residencies to the degree Radonc or ER did.

Midlevels in other specialties over order, increasing volume/demand.

Retirements

 

[SeisK]

This is in part speculation, but my suspicion is that radonc academic leadership can have their productivity/salary subsidized by trainees in a way that radiology can’t, which eliminates the incentive of lower tier or community programs to form radiology training programs.

Our “oscillating job market“ was just the growing pains of radiology evolving from a lifestyle friendly specialty to one with hours/call closer to a surgical subspecialty. The ‘08 housing crisis delayed retirements, with additional reimbursement decreases (driven mostly by the ACA) resulting in a lot of people having to cut the fat and focus much more on getting reads out as quickly as possible. Radiology started to become a volume enterprise, and so needed fewer radiologists to read the same number of scans in a day.

However now the volumes are so high and we have no way of improving efficiency. And they’re just going to keep climbing and climbing. A lot of people are hoping AI is going to help ease the workload, but my suspicion is it’s just making our reads slower, albeit probably more accurate sometimes.

 

[2brads]

Rad volumes keep rising and resident pool did not expand at the same rate. I don’t know much about radonc but it seems like volumes are declining while resident spots are expanding….
It is just supply and demand.

As far as being a lifestyle specialty, it is still one of the best IMO. I don’t know any rads pulling surgical type shifts.

 

[deleted379552]

The job market is insane. There are way more jobs than graduating fellows and tons of residents I know have opted to go straight into general/emergency/tele radiology and forgo fellowship. This has left tons of fellowship spots unfilled (somewhere between 28-30% of all MSK fellowship spots went unfilled and it may look worse this coming match year). If this continues, there will definitely be a push for residency expansion soon. Hopefully, our leadership will learn from the mistakes of EM/RadOnc residency expansion and stem the tide. Private equity groups (sadly a big player in all of medicine now) will make a strong push for expansion as a flooded market will allow them to dictate pay and recruit easier (no one wants to work for RadPartner unless they have to and in this market most won't have to).

As Maxxor said, for years now as NP/PAs expand into other fields, they've ordered a ton of imaging, which has been a boon for radiology. As a field, radiology has been a bulwark against midlevel creep, but pressure is mounting more and more each year as the volume get higher and higher. We will see how long we can hold the floodgates. AI so far has been no help. We have both AIDoc and Viz.ai at our hospital and I'm still waiting for the day when it tells me something that I don't already see myself. Also, its cloud-based operation is very spotty and there are tons of misses, which makes it highly unreliable at the moment. I'm sure that will change in the future, but in the most immediate future midlevel creep is a much bigger threat in my opinion (think MACRA and the private equities leading the charge for its passage).

 

[qwerty89]

The job market is insane. There are way more jobs than graduating fellows and tons of residents I know have opted to go straight into general/emergency/tele radiology and forgo fellowship. This has left tons of fellowship spots unfilled (somewhere between 28-30% of all MSK fellowship spots went unfilled and it may look worse this coming match year). If this continues, there will definitely be a push for residency expansion soon. Hopefully, our leadership will learn from the mistakes of EM/RadOnc residency expansion and stem the tide. Private equity groups (sadly a big player in all of medicine now) will make a strong push for expansion as a flooded market will allow them to dictate pay and recruit easier (no one wants to work for RadPartner unless they have to and in this market most won't have to).

As Maxxor said, for years now as NP/PAs expand into other fields, they've ordered a ton of imaging, which has been a boon for radiology. As a field, radiology has been a bulwark against midlevel creep, but pressure is mounting more and more each year as the volume get higher and higher. We will see how long we can hold the floodgates. AI so far has been no help. We have both AIDoc and Viz.ai at our hospital and I'm still waiting for the day when it tells me someone that I don't already see myself. Also, its cloud-based operation is very spotty and there are tons of misses, which makes it highly unreliable at the moment. I'm sure that will change in the future, but in the most immediate future midlevel creep is a much bigger threat in my opinion (think MACRA and the private equities leading the charge for its passage).

Radiology also cannot make the mistake of NOT expanding residency spots at all. If we continue to have a massive shortage the govt and private equity will look for "alternatives". Whether that is promoting RPAs and expanding their scope of practice (i.e. MARCA) or allowing midlevels sign off independently on some types of studies. You want an undersupply but it can't be dramatic.

 

[SeisK]

Rad volumes keep rising and resident pool did not expand at the same rate. I don’t know much about radonc but it seems like volumes are declining while resident spots are expanding….
It is just supply and demand.

As far as being a lifestyle specialty, it is still one of the best IMO. I don’t know any rads pulling surgical type shifts.

50-60hrs/wk and Q6-Q10 call is average for rads, and average for some surgical sub specialties. The days of 45hr wks no call or weekends is over.

 

[qwerty89]

50-60hrs/wk and Q6-Q10 call is average for rads, and average for some surgical sub specialties. The days of 45hr wks no call or weekends is over.

A few jobs still are. The VA for one. Mine is 8-5 M-F with about 4 weekends from home per year. All outpatient. Never work after 5pm. 60 rvu/day. Can read 90% your own subspecialty. 8 weeks vacation now and about 12 as partner.

Of course you get paid less but some jobs like that are available.

 

[maxxor]

Radiology also cannot make the mistake of NOT expanding residency spots at all. If we continue to have a massive shortage the govt and private equity will look for "alternatives". Whether that is promoting RPAs and expanding their scope of practice (i.e. MARCA) or allowing midlevels sign off independently on some types of studies. You want an undersupply but it can't be dramatic.

I have reluctantly come to agree with this.

If you become too expensive, they find a replacement.

 

[Taurus]

You shouldn’t pick a specialty because of its job market. I genuinely like radiology and my subspecialty. I like my minimal patient contact. Like most radiologists, I literally do no paperwork at all. Patients and paperwork are the two most soul-sucking parts of medicine. I’m a partner in a large private practice group, work mostly 8-5 weekday M-F, weekend call about every 6 weeks, and get 15 weeks vacation. I live in large desirable city. I’m very happy I chose radiology. In my 4th year as a medical student, RadOnc was multiple factors more competitive than radiology to match. Heck, EM was hot at that time too and it was more competitive than radiology to match. It’s ironic how times have changed. I never seriously considered RadOnc though because I felt it was a one-trick pony because of its reliance on radiation treatment. After seeing first hand the “dirty” side effects of radiation treatment, I didn’t think it made sense to pick radiation if a chemo therapy was just as effective and much fewer long term side effects. I didn’t seriously consider EM because it fell in my triad (primary care, EM, and gas) of fields at high risk of midlevel encroachment.

I’m making bank. Make hay while the sun is out as they say. Because of the hot housing market, I think my wife and I have already reached FIRE even though I’m only 6 years out of training but I plan to still work til I’m 70. Who knows how long this gravy train will last. I’m looking into using my gravy train to fund other sources of passive income.

 

[2brads]

I’m pretty much like Taurus. I never work past 5PM and could retire after only 7 years out of fellowship. But the job is too easy/lucrative to walk away now and we get tons of vacay. If you don’t think this is a lifestyle specialty, you’re doing to wrong.

 

[SeisK]

You shouldn’t pick a specialty because of its job market.

Eh, sometimes you should. I loved the idea of being a PhD before med school but saw how horrid the job market was for new grads, and am still exceptionally happy I didn’t do it.

Sometimes the job market can be so tight it isn’t a worthy investment.

 

[scootad.]

Job market is booming now. BUT be careful what job you choose. There’s so much variability and still a fair number of poor quality ones (especially in the highly desired locations) e.g lots of call, busy soul sucking volumes, low $/RVU.

But it’s a different world now than when I finished fellowship ten yrs ago, when jobs were scarce.

Due your due diligence and you will find some gems (high pay and lots of vacay) out there.

This is a good time to be a radiologist

 

[Radology]

The current booming job market, and previous dire job market, have nothing to do with any strategic decisions from a centralized party

2008 financial crisis and multiple deficit reduction acts targeting radiology caused the market crash a decade ago

The delayed retirements from a decade ago combined with unrelated Exploding imaging volumes caused the market boom we’re seeing now

 

[Shifty B]

Choosing a specialty based on the job market now is like picking a team to win the Super Bowl based on how good the team was in 2016. Yeah, there are some perennial winners, but it's pretty unpredictable.

These residents and fellows getting all these job offers now chose radiology when it was at its bottom. Residency spots were going unfilled and word on the street was AI was going to replace all the jobs. Now people realize they will retire before Tesla even moves full self driving out of beta.

It's a great job market for radiologists right now, but all it takes is one shift in reimbursement formula and things can change almost overnight. If there are two specialties you like equally, MAYBE(?) it's a consideration, but a minor one at that. Choose something you want to do and you enjoy, that way at least you're certain of one thing.

 

[qwerty89]

I have reluctantly come to agree with this.

If you become too expensive, they find a replacement.

Not to long ago there were stories of ER docs getting up to 500/hr for locums gigs at desperate hospitals in the middle of nowhere.

Those desperate hospitals have now found a cheaper alternative.

 

[qwerty89]

Choosing a specialty based on the job market now is like picking a team to win the Super Bowl based on how good the team was in 2016. Yeah, there are some perennial winners, but it's pretty unpredictable.

These residents and fellows getting all these job offers now chose radiology when it was at its bottom. Residency spots were going unfilled and word on the street was AI was going to replace all the jobs. Now people realize they will retire before Tesla even moves full self driving out of beta.

It's a great job market for radiologists right now, but all it takes is one shift in reimbursement formula and things can change almost overnight. If there are two specialties you like equally, MAYBE(?) it's a consideration, but a minor one at that. Choose something you want to do and you enjoy, that way at least you're certain of one thing.

I think that used to be the case for the last downturn in 2009-2013. Reimbursement declines led to declining income. Old guy retires. The group then decides to NOT hire someone else and instead each read a few more studies to make up the difference (and keep income the same).

Groups are already at their max workload. I don't think the average doc in private practice can realistically read much more than they are now. Any decline in reimbursement would mostly just reduce income. There would only be a mild effect on the job market at this point IMO.

Volume shows no sign of slowing down. Both due to increased midlevel use AND the fact that clinical docs of my generation expect and have come to rely on imaging much more than before. Unfortunately that just puts imaging in the sights of CMS who will just cut further. A lot of medicine is just algorithmic at this point.

 

[2brads]

I think that used to be the case for the last downturn in 2009-2013. Reimbursement declines led to declining income. Old guy retires. The group then decides to NOT hire someone else and instead each read a few more studies to make up the difference (and keep income the same).

In my experience, the opposite happened. Stock market crashed, old guys did NOT retire, but they are doing so now because they are really old and have lots more money, which just compounds the hiring crunch.

 

[qwerty89]

You shouldn’t pick a specialty because of its job market. I genuinely like radiology and my subspecialty. I like my minimal patient contact. Like most radiologists, I literally do no paperwork at all. Patients and paperwork are the two most soul-sucking parts of medicine. I’m a partner in a large private practice group, work mostly 8-5 weekday M-F, weekend call about every 6 weeks, and get 15 weeks vacation. I live in large desirable city. I’m very happy I chose radiology. In my 4th year as a medical student, RadOnc was multiple factors more competitive than radiology to match. Heck, EM was hot at that time too and it was more competitive than radiology to match. It’s ironic how times have changed. I never seriously considered RadOnc though because I felt it was a one-trick pony because of its reliance on radiation treatment. After seeing first hand the “dirty” side effects of radiation treatment, I didn’t think it made sense to pick radiation if a chemo therapy was just as effective and much fewer long term side effects. I didn’t seriously consider EM because it fell in my triad (primary care, EM, and gas) of fields at high risk of midlevel encroachment.

I’m making bank. Make hay while the sun is out as they say. Because of the hot housing market, I think my wife and I have already reached FIRE even though I’m only 6 years out of training but I plan to still work til I’m 70. Who knows how long this gravy train will last. I’m looking into using my gravy train to fund other sources of passive income.

Every article written by docs about the downfall of medicine always talks about the latter.

But the real secret of medicine is that the former is also equally the problem. You just aren't allowed to talk about it.

Being a clinical doc is like working customer service in retail. Or like being a teacher in a fancy zip code where all the parents think their kids deserve As.

Good luck getting a med student to understand that though.

 

[Hzreio]

“Patients are one of the most soul sucking parts of medicine”

-radiologist

Lol

 

[DOFOSHO]

If you are absolutely certain you will do a fellowship, should you start applying for jobs as a senior resident or wait until you're a fellow?

 

[2brads]

If you are absolutely certain you will do a fellowship, should you start applying for jobs as a senior resident or wait until you're a fellow?

If there is a place you need to be then no harm looking as soon as possible. I know people who booked jobs as 4th years and even got some financial support during their fellowships.

 

[DoctorMD1]

50-60hrs/wk and Q6-Q10 call is average for rads, and average for some surgical sub specialties. The days of 45hr wks no call or weekends is over.

I disagree. You can easily find 3-4 days work week without calls and evenings if you are not interested in partnership.

 

[2brads]

A fair number of groups still do 4 day work weeks. Have to work an occasional weekend day if you’re a partner though.

 

[SeisK]

I disagree. You can easily find 3-4 days work week without calls and evenings if you are not interested in partnership.

Sure you can! You can get a breast surgery gig 4 days a week no call no weekends.

Averages are equal, that’s the point

 

[Neuronix]

If there are two specialties you like equally, MAYBE(?) it's a consideration, but a minor one at that. Choose something you want to do and you enjoy, that way at least you're certain of one thing.

From the perspective of starting my career and living in the bad job market of radiation oncology, I disagree with you.

It is impossible to know for sure what will happen with the job market, that part is true. Still, it's hard to see rad onc getting better. There are a lot of reasons it will only continue to worsen. If radiology takes a dive in say 5 years, I think we'll all not see it coming.

So for a medical student, I think it's wise to pick a specialty based on the best information available and make a informed choice. It's rare that a medical student would only be happy in radiation oncology (and if so, welcome aboard!). I assume that's true in other specialties as well. So I disagree and would say the job market should be a major factor.

I've come to this conclusion after watching supply-demand mismatch affect things in ways I never realized. I've also been watching how med oncs and radiologists are treated, which is totally different than rad onc. When I was younger and more naive, I used to think that academics wanted to give faculty the things necessary to succeed. I used to think that the private world was all about equal partnerships. I was a fool. They are all out to get as much money as they can out of you. The only way you get anything is to negotiate. At the very least groups look around at what others are doing and what they think they can get away with in the current job market. It's not even all about money. Everything from time to partner, location choices, ability to change jobs for any reason, how you are treated in your job, how your clinics run, vacation time, resources to do research (in academics), pay raises/cuts in employed positions, etc etc etc... It's all predicated on how much competition there is for jobs, how easily replaceable you are, and what other job offers you can get.

Employers are still making a ton of money off employed rad oncs. Stagnant and declining pay are not directly a function of reimbursement cuts. The money generated by rad oncs is just not going into their pockets. It's increasingly common that 10% or less of the collections generated by the rad onc go to the rad onc. These jobs would not be filled 10-20 years ago, but now they are often the only thing available anywhere near a metro area.

The grass is always greener on the other side, it is true. But I just wanted to give the perspective from someone living on this side.

PS: as I was writing this, this was posted:

Ghosting during the job hunt

This is happening all over rad onc right now. Institutions are ****ing us over simply because they can. If that doesn't scare med students away, I don't know what will.

 

[maxxor]

I've come to this conclusion after watching supply-demand mismatch affect things in ways I never realized. I've also been watching how med oncs and radiologists are treated, which is totally different than rad onc. When I was younger and more naive, I used to think that academics wanted to give faculty the things necessary to succeed. I used to think that the private world was all about equal partnerships. I was a fool. They are all out to get as much money as they can out of you. The only way you get anything is to negotiate. At the very least groups look around at what others are doing and what they think they can get away with in the current job market. It's not even all about money. Everything from time to partner, location choices, ability to change jobs for any reason, how you are treated in your job, how your clinics run, vacation time, resources to do research (in academics), pay raises/cuts in employed positions, etc etc etc... It's all predicated on how much competition there is for jobs, how easily replaceable you are, and what other job offers you can get.

This is 100% true and I’ve lived it. The radiology job market in my city is garbage, mainly because the main groups have all colluded to have identical exploitative partnership tracks. This then affects the academic market because they only have to offer slightly above what the crappy partnership tracks offer.

It’s an amazing market for radiology right now…in other states. In mine, it’s crap.

 

[vm26]

This is 100% true and I’ve lived it. The radiology job market in my city is garbage, mainly because the main groups have all colluded to have identical exploitative partnership tracks. This then affects the academic market because they only have to offer slightly above what the crappy partnership tracks offer.

It’s an amazing market for radiology right now…in other states. In mine, it’s crap.

Sorry to hear that. In other geographic areas its infiltration by P/E that has killed the market. With that said, one can always move. Aside from family, its the same pro's/con's that one has to navigate.

 

[vm26]

A fair number of groups still do 4 day work weeks. Have to work an occasional weekend day if you’re a partner though.

Former job had me working 2-3 days on average as a partner (though with weekend and Q6 weekday call taken from home til about 10pm). Clearly we weren't crushing it financially but had plenty of time to do per diem if I wanted to.

 

[vm26]

Not to long ago there were stories of ER docs getting up to 500/hr for locums gigs at desperate hospitals in the middle of nowhere.

Those desperate hospitals have now found a cheaper alternative.

Really hard to predict how things play out. I suspect that eventually we will be largely replaced by AI+/- mid-levels but this may happen 5 years from now or 20+ years from now. Given the unique nature of our work, we have been largely shielded from mid-level encroachment. I get CMS is always on our a$$ but somehow each of us finds a way of being more "productive". As things stand now I would recommend the field. It's still possible to find a solid PP group that is more than reasonable in terms of work load, vacation, and compensation.

 

[SeisK]

AI+/- mid-levels

30+ years of any day now.

Remember when MRI was going to make us useless?

 

[qxrt]

AI has shown no capability in any field near that needed to do anything close to what is needed in radiology. All the hype is thrown about by programmers and business types who think radiology is just pattern recognition with potential for big profits if it can be automatized and don't actually understand the cognition needed to interpret a CT. Notice how the goalposts have moved from "AI will take over radiology" to "AI will make radiologists more efficient and reduce demand" as people come to grips with the limitations of current AI technology, even then not realizing that radiology is probably already the most efficient specialty in medicine with much less room for improvement in productivity. It may happen sometime in the distant future when AI can think like humans, but not in our lifetimes.

And I don't foresee radiology volumes going down to any meaningful extent in the future. People are obsessed with data and discrete information they can see with their own eyes, and radiology is one of the foundations of that (pathology is the main other, but unlike radiology where you can always build more machines faster to take more pictures of people who may not even have anything wrong with them other than a vague sense of unease or a touch of nerves, there is a more confined logistical limit to how many biopsies/pathology specimen of suspicious tissue can be obtained in a day). Radiology job market is booming because of volumes caused by that. Has nothing to do with the capability or foresight of radiology leadership. Radiology leadership is no more benevolent nor wise than that in other specialties like rad onc.

It also helps that unlike in other fields like EM or maybe even rad onc (I honestly don't know too much about rad onc), residents are functionally labor-extenders who help boost productivity at low cost. Radiology residents, on the other hand, are a net drag to productivity, and their only benefit in terms of cost is their ability to take night/weekend call. And with the increasing demand for attending radiologist coverage even at night and on weekends, even that benefit is faltering. Money talks, and in that regard, there's less monetary incentive to start a radiology residency.

Finally, you can't replace a radiologist with, say, two NPs/PAs like hospitals and private equity has been doing to EM/hospitalists/anesthesiologists/etc. NPs/PAs help out immensely with scutwork and low-liability risk work like seeing the ED patient with a sprained ankle or checking in daily on the inpatient or sedating the healthy young patient for a small procedure. There is almost no scutwork in diagnostic radiology, and almost no radiology exam is considered low-liability, as every exam has potential for a critical finding. Maybe a bit in IR, but that's only a small part of radiology, and the midlevels are not doing anything near a big case like ablations or embolizations.

 

[2brads]

AI has shown no capability in any field near that needed to do anything close to what is needed in radiology. All the hype is thrown about by programmers and business types who think radiology is just pattern recognition with potential for big profits if it can be automatized and don't actually understand the cognition needed to interpret a CT. Notice how the goalposts have moved from "AI will take over radiology" to "AI will make radiologists more efficient and reduce demand" as people come to grips with the limitations of current AI technology, even then not realizing that radiology is probably already the most efficient specialty in medicine with much less room for improvement in productivity. It may happen sometime in the distant future when AI can think like humans, but not in our lifetimes.

And I don't foresee radiology volumes going down to any meaningful extent in the future. People are obsessed with data and discrete information they can see with their own eyes, and radiology is one of the foundations of that (pathology is the main other, but unlike radiology where you can always build more machines faster to take more pictures of people who may not even have anything wrong with them other than a vague sense of unease or a touch of nerves, there is a more confined logistical limit to how many biopsies/pathology specimen of suspicious tissue can be obtained in a day). Radiology job market is booming because of volumes caused by that. Has nothing to do with the capability or foresight of radiology leadership. Radiology leadership is no more benevolent nor wise than that in other specialties like rad onc.

It also helps that unlike in other fields like EM or maybe even rad onc (I honestly don't know too much about rad onc), residents are functionally labor-extenders who help boost productivity at low cost. Radiology residents, on the other hand, are a net drag to productivity, and their only benefit in terms of cost is their ability to take night/weekend call. And with the increasing demand for attending radiologist coverage even at night and on weekends, even that benefit is faltering. Money talks, and in that regard, there's less monetary incentive to start a radiology residency.

Finally, you can't replace a radiologist with, say, two NPs/PAs like hospitals and private equity has been doing to EM/hospitalists/anesthesiologists/etc. NPs/PAs help out immensely with scutwork and low-liability risk work like seeing the ED patient with a sprained ankle or checking in daily on the inpatient or sedating the healthy young patient for a small procedure. There is almost no scutwork in diagnostic radiology. Maybe a bit in IR, but the midlevels are not doing anything near a big case like ablations or embolizations.

Agreed. AI is a nonfactor in radiology for the foreseeable future for a lot of reasons. I do foresee expansion of residency positions though. The elderly academics would take a pay cut to have more warm bodies generating reports if it comes to that.

 

[Taurus]

The EM docs don’t like this observation when I posted it in their forum but sometimes the truth hurts.

“Images don’t change and they can last forever.”

This is what ultimately protects diagnostic fields like radiology and pathology from midlevel encroachment. In our highly litigious country, it creates this big beautiful insurmountable moat separating radiologist from midlevel. Wall Street understands this and that’s why they are trying the AI route. They are hoping they can significantly increase the efficiency of each radiologist and hence decrease overall demand to hire them. From my experience also, I think that AI will prove to be a useless hyped-up feature that companies will waste millions on.

 

[vm26]

AI has shown no capability in any field near that needed to do anything close to what is needed in radiology. All the hype is thrown about by programmers and business types who think radiology is just pattern recognition with potential for big profits if it can be automatized and don't actually understand the cognition needed to interpret a CT. Notice how the goalposts have moved from "AI will take over radiology" to "AI will make radiologists more efficient and reduce demand" as people come to grips with the limitations of current AI technology, even then not realizing that radiology is probably already the most efficient specialty in medicine with much less room for improvement in productivity. It may happen sometime in the distant future when AI can think like humans, but not in our lifetimes.

And I don't foresee radiology volumes going down to any meaningful extent in the future. People are obsessed with data and discrete information they can see with their own eyes, and radiology is one of the foundations of that (pathology is the main other, but unlike radiology where you can always build more machines faster to take more pictures of people who may not even have anything wrong with them other than a vague sense of unease or a touch of nerves, there is a more confined logistical limit to how many biopsies/pathology specimen of suspicious tissue can be obtained in a day). Radiology job market is booming because of volumes caused by that. Has nothing to do with the capability or foresight of radiology leadership. Radiology leadership is no more benevolent nor wise than that in other specialties like rad onc.

It also helps that unlike in other fields like EM or maybe even rad onc (I honestly don't know too much about rad onc), residents are functionally labor-extenders who help boost productivity at low cost. Radiology residents, on the other hand, are a net drag to productivity, and their only benefit in terms of cost is their ability to take night/weekend call. And with the increasing demand for attending radiologist coverage even at night and on weekends, even that benefit is faltering. Money talks, and in that regard, there's less monetary incentive to start a radiology residency.

Finally, you can't replace a radiologist with, say, two NPs/PAs like hospitals and private equity has been doing to EM/hospitalists/anesthesiologists/etc. NPs/PAs help out immensely with scutwork and low-liability risk work like seeing the ED patient with a sprained ankle or checking in daily on the inpatient or sedating the healthy young patient for a small procedure. There is almost no scutwork in diagnostic radiology, and almost no radiology exam is considered low-liability, as every exam has potential for a critical finding. Maybe a bit in IR, but that's only a small part of radiology, and the midlevels are not doing anything near a big case like ablations or embolizations.

Having recently competed a few CME courses on breast imaging, it does seem that AI is getting pretty good at screeners. Throw in a mid-level to prelim or finalize and that would drastically decrease the need for breast imagers. Plus its easy enough to train a mid-level to perform most breast procedures. Breast imaging (particularly tomo screeners) is a money maker and breast imagers are in high demand to the point where they can once again obtain 9-5 no call jobs. Private equity/large HC systems could save/make a lot with the AI+ mid-level combo. I def see some barriers to this (eg. breast surgeons insisting on rads performing locs, and overall optics) but who knows. I am hoping this doesn't happen for some time. On the positive side, the complex/fast-paced nature of image interpretation precludes mid-level encroachment.

 

[2brads]

Yea, this is the only area I predict AI will have a significant impact since mammos are relatively anatomy agnostic. I could see a machine generating a report and getting it signed off by a human. I wouldn’t even be that bothered by it. I find mammos mind numbing.

 

[deleted379552]

I think there's some underlying presumption that radiology/medicine will stay static as AI evolves. Who is to say that mammo/tomo will remain the screening modality of choice?

Also, who is to say mammo screening won't be replaced by other non-imaging screening exams? GRAIL Receives New York State Approval for Galleri Multi-Cancer Early Detection Blood Test

Our hospital has a truncated MR screening protocol that cuts the magnet time to less than 10-15 minutes. It would be quite ironic if the moment the mammo screening AI becomes slightly useful, a different exam becomes standard of care.

This is a cool article for those interested on some outdated radiology examinations: @RadioGraphics.

 

[Taurus]

I think there's some underlying presumption that radiology/medicine will stay static as AI evolves. Who is to say that mammo/tomo will remain the screening modality of choice?

Also, who is to say mammo screening won't be replaced by other non-imaging screening exams? GRAIL Receives New York State Approval for Galleri Multi-Cancer Early Detection Blood Test

Our hospital has a truncated MR screening protocol that cuts the magnet time to less than 10-15 minutes. It would be quite ironic if the moment the mammo screening AI becomes slightly useful, a different exam becomes standard of care.

This is a cool article for those interested on some outdated radiology examinations: @RadioGraphics.

I agree. It’s very risky to be a one-trick pony. Just look at RadOnc. That’s why I advise against doing only one subspecialty and losing your general skills. The mammo job market will tank badly if other screening methods are approved and are as effective, ie, a blood test. The majority of the profits in mammo is from the high screening volumes and the often negative callbacks. An effective blood test would eliminate most of that.

 

[2brads]

Oh man. The schadenfreud will be strong if that happens. All these subspecialty mammo-only folks have been coddled for so long…

 

[tomballunger]

https://seekingalpha.com/article/4434811-radnet-rdnt-investor-presentation-slideshow

“Almost 20% of all collected global revenue goes towards paying our radiologists.

Material opportunity to lower this cost through artificial intelligence.”- slide 18

Any comments on this?

 

[maxxor]

https://seekingalpha.com/article/4434811-radnet-rdnt-investor-presentation-slideshow

“Almost 20% of all collected global revenue goes towards paying our radiologists.

Material opportunity to lower this cost through artificial intelligence.”- slide 18

Any comments on this?

Radnet is a huge investor in internal AI products. They’ve bought a couple startups and brought them in-house. They are spending big money to try and make it happen.

It will come, often in a way we didn’t expect. I know a couple of the key players in Mammo AI; it’s gotten further along than I expected since their first few publications on automated density classification.

 

[SeisK]

https://seekingalpha.com/article/4434811-radnet-rdnt-investor-presentation-slideshow

“Almost 20% of all collected global revenue goes towards paying our radiologists.

Material opportunity to lower this cost through artificial intelligence.”- slide 18

Any comments on this?

¯\_(ツ)_/¯

People often oversell to investors.

Despite their big Jan 2021 study, I don’t see how their algorithm could be reasonably implemented into a workflow that saves me anytime or near obviates the need for a radiologist. I’m happy to go into details if you like.

 

[2brads]

“Material opportunity” and reality do not always coexist…

I don’t deny that there is a lot of motivation to excise radiologists from radiology, but this song and dance has played out with other technologies for years/decades. (Where’s my flying car?)

My hope is for incremental tech that does things like: pre-read mammos for probability of cancer, automated bone age, automatic lung nodule finder and comparison. If I get these three things, my life is good. But even these relatively “simple” things always seem just beyond the horizon.

 

[qwerty89]

FYI as was also posted on Auntminnie. Radnet has an odd set up where some of their radiologists are employees (and therefore Radnet skims their professional fees) and other rads are part of groups which have contracts with radnet where radnet collects only the technical fees and the groups get the professional fees. The Beverly group in LA are employees and other groups affiliated with Radnet in the Northern LA suburbs and Orange County are independent groups with contracts with Radnet.

 

[Cognovi]

30+ years of any day now.

Remember when MRI was going to make us useless?

I hear that asserted all the time by people who are not old enough to remember that. I have yet to hear it from someone who lived through that era (80s) or see it printed in some journal op-ed.

 

[maxxor]

I hear that asserted all the time by people who are not old enough to remember that. I have yet to hear it from someone who lived through that era (80s) or see it printed in some journal op-ed.

there may have been a brief period of an existential threat to Radiology.

One of the first clinical mris was installed in an old mcdonalds adjacent to a hospital in Florida by a pathologist who later did a Nuc Med residency. I think at the beginning, there was fear that radiology would not “own” the modality.

There was a Forbes article about him a couple years ago.

Want Fries With That? A Brief History Of Medical MRI, Starting With A McDonald's

The fast-food joint was converted into an outpatient MRI center before the technology was FDA-approved. In its story, find the history of a $50 billion industry.

www.forbes.com

 

[vm26]

https://seekingalpha.com/article/4434811-radnet-rdnt-investor-presentation-slideshow

“Almost 20% of all collected global revenue goes towards paying our radiologists.

Material opportunity to lower this cost through artificial intelligence.”- slide 18

Any comments on this?

They lost me at "no sub-specialized radiologist readers" within hospitals. Parking issue is also laughable. AI is coming but many hurdles. One being public perception (particularly with breast imaging), and more importantly medical legal liability.


• Approx. 60% imaging occurs within hospitals. • More expensive for patients and their
insurance companies
• Inferior service
• More difficult access and parking
• Often no sub-specialized radiologist readers

 

[Taurus]

My philosophy is you can't predict the exact future. You can only make educated guesses on the best available data. Based on what I'm seeing, radiology is still in a very good position compared to some other fields. I would still highly recommend medicine, radiology, and my subspecialty to any prospective students. I'm making bank right now and I have practically reached FIRE 6 years out of training. I plan to continue working til 70 though. I would encourage everyone to look outside of medicine for other sources of income. There are so many things outside of your control that can impact your income, job security, and job satisfaction, e.g., govt reimbursement cuts, uncontrolled residency expansion, blood test, changing indications, etc. It's not just radiology but it applies to any medical field. Find a passive income stream that may replace some or most of your doctor income.

 

[TXRads]

My philosophy is you can't predict the exact future. You can only make educated guesses on the best available data. Based on what I'm seeing, radiology is still in a very good position compared to some other fields. I would still highly recommend medicine, radiology, and my subspecialty to any prospective students. I'm making bank right now and I have practically reached FIRE 6 years out of training. I plan to continue working til 70 though. I would encourage everyone to look outside of medicine for other sources of income. There are so many things outside of your control that can impact your income, job security, and job satisfaction, e.g., govt reimbursement cuts, uncontrolled residency expansion, blood test, changing indications, etc. It's not just radiology but it applies to any medical field. Find a passive income stream that may replace some or most of your doctor income.

What is your subspecialty?

 

[painballer27]

I’m pretty much like Taurus. I never work past 5PM and could retire after only 7 years out of fellowship. But the job is too easy/lucrative to walk away now and we get tons of vacay. If you don’t think this is a lifestyle specialty, you’re doing to wrong.

Does the job really feel easy to you? Like I get that its not physically taxing, but it seems like it requires a great amount of sustained focus/thinking all day long? Just curious as a current med student considering radiology