Ortho vs Radiology

[nibhighfootballrules]

Hi all I am currently a medical student torn between applying into orthopedics and radiology. I came into medical school wanting to pursue orthopedics and worked hard to get a lot of research experience and good grades. Throughout medical school I continually heard if you like anything equally as much you should do that instead. The only other field I have enjoyed is radiology. I enjoy the fact you get to see the entire spectrum of pathologies and that you become an expert in a specific skill (image interpretation). I like that you drive management and that there is so much to learn and keep you stimulated.


The more I have immersed myself in orthopedics I realized I loved the patient population and being in the OR. Additionally, I like the manual work involved in ortho. I would imagine myself getting a lot of satisfaction from going in and fixing peoples mechanical problems and being able to look at my work post-op and have hard proof that I made a difference. My main issue with orthopedics is the sacrifices required for this field. I am not someone who is adverse to hard work but I have begun thinking just because I am capable of enduring training etc doesn't mean it’s for me. I am a non-trad and started school later in life. I plan on having my first child prior to residency and having done orthopedic rotations working 80 hour weeks the thought of having such little time to spend with my future family seems extremely depressing. I know residency/fellowship is only a moment in time but those 6 years I'll never get back and starting later in life l think makes them even more valuable. Apart from residency I am not naive to think life is significantly that much more relaxed as an attending. First you have to build a practice, deal with OR booking time, and then maybe 10-15 years after training you can create a schedule that is more amenable to your personal goals but surgeons are always responsible for their patients and their schedule is inherently unpredictable to a certain extent. I am also concerned about the toll training and subsequent career will have on my physical and mental health and how it will impact my ability to be a good partner/father when I am not working. I cant tell if this is a healthy fear or something that will only get worse throughout my time training and make my life miserable. I find it difficult to have candid conversations with attendings about this because from my experience these concerns are taboo to talk about.

My thought is if I might find a bit less purpose in my work (which may or may not be the case in radiology) and have to sacrifice way less time then why would I not chose that option, but its very difficult to know how I will feel long term. I have read forums of people with similar situations and wanted to see what peoples opinions were now in 2024.

I know the ceiling for earnings is higher in orthopedics and there is less of a threat for industry disruption considering rads is tech heavy and who knows what the field will look like in 30 years. I try to have an optimistic outlook on AI and think radiology will evolve alongside tech advancement and create new opportunities and areas of utility. I know no-one can tell the future but I am concerned that after 6 years of residency/fellowship I will end up regretting my decision due to changes in radiology that make it no longer a viable way to make a good living long term. A side note I want to remain in a large metropolitan area and from what I have researched it seems that radiology jobs in these markets tend to be lower paying and I have also seen that a lot of private practice jobs, especially in these areas, are becoming few and far between because of large healthcare systems and PE firms buying up practices. Money is not my motivation but with a ton of student debt and wanting to start a family it feels naive to not have this factor into my decision.

Additionally, although I enjoy radiology as a medical student it’s very difficult to fully appreciate the life of a radiologist attending. I spent a year scribing in FM prior to school and my time in the hospital during clinicals have made me realize im ok with not having so much patient interaction but I dont know if completely eliminating that will be an issue as well. Ive been told you can perform procedures in rads and get some patient interaction. Also you are likely to be communicating with colleagues about reads which I like but I am still uncertain if that will be enough. It’s difficult as a medical student to fully appreciate patient interactions when you are predominantly seeing patients in-house and extrapolating that to outpatient. I am curious what peoples perspectives are and any advice would be greatly appreciated. Thanks!



TLDR: deciding between ortho and rads. Know I enjoy ortho work but concerned about sacrifices required. Enjoy radiology but cannot get a great feel for attending grind and am concerned about job prospects longterm.

---

Members don't see this ad.

 

[mistafab]

Forget the future job prospects part of this. Neither specialty will be out of a job.

Do you want to be in the OR, or do you want to be at a computer?

That’s the real question. Though the topic of interests overlap (anatomy), the lifestyle and day to day does not overlap at all in these fields.

Decide on that and enjoy your life. Amen.

Good luck.

 

[Esmolol]

Just based on the way this is written - ORTHO. Rads reports have paragraphs, indentation, punctuation... /s

 

[nibhighfootballrules]

Just based on the way this is written - ORTHO. Rads reports have paragraphs, indentation, punctuation... /s

Would you say rads now???

 

[Esmolol]

Would you say rads now???

Only if the TLDR would be presented as an enumerated list.

 

[deleted1065184]

Do IR. Very procedural, but still image interpretation.

 

[voxveritatisetlucis]

If all else equal, I would go ortho since there’s less risk of automation, even if it’s 10-15 years down the line

 

[nibhighfootballrules]

Do IR. Very procedural, but still image interpretation.

Thank you for your response. My concern with IR route is a) its no guarantee to get into the early specialization spots within a DR program (although from my understanding not impossible and comes down to how many people in your class are interested). b) the lifestyle of IR seems more comparable to gen surgery vs DR but I'm not sure if you can find a job where you split time between IR and DR. c) multiple mentors have told me that radiation exposure in IR is not something to take lightly

 

[nibhighfootballrules]

If all else equal, I would go ortho since there’s less risk of automation, even if it’s 10-15 years down the line

Appreciate the response. There is a lot of unpredictably with regard to DR and regardless of what you can read online it's impossible to tell what that field will look like within the next 30-40 years.

 

[odyssey2]

Thank you for your response. My concern with IR route is a) its no guarantee to get into the early specialization spots within a DR program (although from my understanding not impossible and comes down to how many people in your class are interested). b) the lifestyle of IR seems more comparable to gen surgery vs DR but I'm not sure if you can find a job where you split time between IR and DR. c) multiple mentors have told me that radiation exposure in IR is not something to take lightly

It's not difficult at all to do IR if you match DR, especially if you match somewhere with ESIR spots. Most DR people don't want to do IR (for lifestyle reasons you mentioned) so it's usually no issue at all landing those spots.

 

[RunsOnCaffeine]

Do Rads. I'm in Ortho and love what I do, but if your heart isn't in it 100%, then do something else. It's worth it if you wouldn't be happy doing anything else, and it sounds like you would be.

 

[nibhighfootballrules]

Do Rads. I'm in Ortho and love what I do, but if your heart isn't in it 100%, then do something else. It's worth it if you wouldn't be happy doing anything else, and it sounds like you would be.

Thanks I appreciate your insight. I wasn't sure if what I was feeling is commonplace or if people who get through it really don't have these concerns

 

[CJhooper123]

Do Rads. I'm in Ortho and love what I do, but if your heart isn't in it 100%, then do something else. It's worth it if you wouldn't be happy doing anything else, and it sounds like you would be.

Counter - I did not have much of an ortho experience in med school. I took a leap and applied. I absolutely love ortho now.
Working with your hands, and learning the biomechanics of the MSK system is awesome.

 

[orbitsurgMD]

Radiology is diverse. There are procedural practice areas, interventional neuroradiology being prominent, where your practice might seem more like neurosurgery than reading images. I would expect that area will grow. Ortho is diverse also. The hand and spine people have their fiefdoms.

 

[nibhighfootballrules]

Counter - I did not have much of an ortho experience in med school. I took a leap and applied. I absolutely love ortho now.
Working with your hands, and learning the biomechanics of the MSK system is awesome.

Thanks for sharing. I’m also wondering if my experiences are biased in ortho just because I was on the trauma service at a busy center. Would you say there is a lot of variability in hours worked per week in ortho depending on the service or it’s consistently 70+ a week for all of training?

 

[CJhooper123]

Thanks for sharing. I’m also wondering if my experiences are biased in ortho just because I was on the trauma service at a busy center. Would you say there is a lot of variability in hours worked per week in ortho depending on the service or it’s consistently 70+ a week for all of training?

In general, ortho works pretty hard and 60-70 hrs/week is pretty common for interns (sometimes more).

I do think there is variability. Academic centers with level I trauma can be pretty rough. There are also community programs that operate a ton and work fewer hours compared to the academic places. Unfortunately, ortho is pretty notorious for working a lot.

 

[operaman]

Tough decision. I don’t know the answer, but you’ve framed the dilemmas and much of the subtlety better than most.

The fighting for OR time is a nuance that I never understood as a student/resident and thankfully lucked into some nice block time, but I see life suck for surgeons who have to board as add ons all the time. I will say that as ortho, much of what you do can be done in outpatient surgery centers (many owned by the ortho partners) and it can be a bit easier to get time at good hours. I know quite a few who take no/minimal call as well, though I’m in an area where we have a massive doc shortage so even relatively new doctors can get away with that.

Your desire for a large metro area with a tougher market definitely hits you on either front. There’s an old adage for physician jobs: location, salary, lifestyle - a good job gives you 2 out of 3. Obviously that applies to both fields.

Perhaps another thought - if you ever wanted to switch later, much easier to go from ortho to rads than the other way around.

The AI concerns for rads are legit. Maybe not replacing rads in your career, but will surely increase MD efficiency which means you will be reading a LOT more scans and getting paid less per scan to do so. Just a function of how CMS values codes.

 

[nibhighfootballrules]

Tough decision. I don’t know the answer, but you’ve framed the dilemmas and much of the subtlety better than most.

The fighting for OR time is a nuance that I never understood as a student/resident and thankfully lucked into some nice block time, but I see life suck for surgeons who have to board as add ons all the time. I will say that as ortho, much of what you do can be done in outpatient surgery centers (many owned by the ortho partners) and it can be a bit easier to get time at good hours. I know quite a few who take no/minimal call as well, though I’m in an area where we have a massive doc shortage so even relatively new doctors can get away with that.

Your desire for a large metro area with a tougher market definitely hits you on either front. There’s an old adage for physician jobs: location, salary, lifestyle - a good job gives you 2 out of 3. Obviously that applies to both fields.

Perhaps another thought - if you ever wanted to switch later, much easier to go from ortho to rads than the other way around.

The AI concerns for rads are legit. Maybe not replacing rads in your career, but will surely increase MD efficiency which means you will be reading a LOT more scans and getting paid less per scan to do so. Just a function of how CMS values codes.

Thanks for your help. I agree that the desire to remain in a large metropolitan area makes things trickier. Even from my comparatively limited interactions with physicians it isn’t uncommon to hear people complaining “if I lived xyz I’d have been retired with two houses by now” and I definitely do not want to be that person one day. It speaks to the importance of enjoying the day to day work.

With regards to AI and radiology. Personally I think the concern of increased efficiency/ productivity is present in any field. Reimbursements for joint replacements aren’t what they used to be and require more surgeries to make the same amount of money. I don’t understand why this seems to be particularly emphasized in radiology. Maybe I am thinking about this incorrectly but if AI increased efficiency and you had to read more scans in the same amount of time wouldn’t they offset each other and it remains the same level of work? Yes you could argue well I used to have to read 60 scans a day now I read 80 for the same pay but you need to work less per scan right?

Unfortunately the true impact of AI seems unpredictable and could turn into a net positive for the field as equally as a net negative. Essentially it feels like it comes down to one’s personal comfort level with taking that risk. Which after 4 years of medical school 6 years of training on top of student debt seems like a pretty significant risk.

 

[Dral]

I got to the point of talking to the Ortho PD of my med school. I was a non-trad as well. I didn't go that route and look back and think I'm glad I didn't (due to the lifestyle issues). However, I don't think I was all in on ortho so to speak. If you are, it could still be for you.

 

[SweetBurger]

Thank you for your response. My concern with IR route is a) its no guarantee to get into the early specialization spots within a DR program (although from my understanding not impossible and comes down to how many people in your class are interested). b) the lifestyle of IR seems more comparable to gen surgery vs DR but I'm not sure if you can find a job where you split time between IR and DR. c) multiple mentors have told me that radiation exposure in IR is not something to take lightly

pretty easy to get an ESIR spot most people dont want them they go unfilled all the time

 

[DOVinciRobot]

Lol why are people recommending IR when OPs only reason for not doing ortho are lifestyle reasons… an attending orthopod and attending IR doc are going to have similar lifestyles, they will have the same length of training, and IR training isn’t any easier than surgery training..

DR sure, IR no.

Personally in your situation OP I would pick ortho. There is no substitute for operating, and your post is pretty obvious that you want to do ortho. Attending life IS different than training.

And you can have a family and be in surgery training, trust me. I have multiple kids and spend time with them every day.

 

[nibhighfootballrules]

Lol why are people recommending IR when OPs only reason for not doing ortho are lifestyle reasons… an attending orthopod and attending IR doc are going to have similar lifestyles, they will have the same length of training, and IR training isn’t any easier than surgery training..

DR sure, IR no.

Personally in your situation OP I would pick ortho. There is no substitute for operating, and your post is pretty obvious that you want to do ortho. Attending life IS different than training.

And you can have a family and be in surgery training, trust me. I have multiple kids and spend time with them every day.

Thanks for sharing, this is uplifting to hear. I think my experience with orthopedic training thus far has been biased by single program/service/role of a medical student which definitely complicates seeing the big picture especially when you are concerned about the sacrifices required.

 

[collegestud2013]

Appreciate the response. There is a lot of unpredictably with regard to DR and regardless of what you can read online it's impossible to tell what that field will look like within the next 30-40 years.

If your concern is long-term job security ortho will likely be the better choice over DR. There's a big push to integrate AI into a lots of fields in medicine, and AI have gotten a lot better at reading images over the past few years. Now many are least good enough to aid radiologists to read more quickly. Job market for radiology is good right now only because the volume of imaging ordered by other clinicians is higher than ever (partly for medicolegal reasons and partly because less and less physicians do a thorough physician exam nowadays and no one really trusts the accuracy of a physical exam documented in the medical record anyways) but AI hasn't been fully integrated yet in most practices. Once it does inevitably, demand for radiologists will drop significantly which will drive down pay, and the remaining radiologists will be expected to read at even faster speeds to make the same amount of pay (though with assistance from AI).

IR is another option but lifestyle will be much worse than DR and mores similar to ortho.

Ortho job market is subject to some encroachment by midlevels but as a surgical field PAs/NPs will have limited role in the OR and most of the role is helping with the non-surgical stuff (eg rounding on post-op patients on the floor or clinic, answering nursing patients or talking to families, seeing consults and writing consult notes). This will also decrease the number of ortho attendings needed on the market, but until we have robots that can perform orthopedic surgeries nearly independently the job market should not be as affected as by DR.

Of course its possible in any specialty to open way too many residency spots and saturate the job market (EM and rac onc being the most recent examples of this).

 

[nibhighfootballrules]

If your concern is long-term job security ortho will likely be the better choice over DR. There's a big push to integrate AI into a lots of fields in medicine, and AI have gotten a lot better at reading images over the past few years. Now many are least good enough to aid radiologists to read more quickly. Job market for radiology is good right now only because the volume of imaging ordered by other clinicians is higher than ever (partly for medicolegal reasons and partly because less and less physicians do a thorough physician exam nowadays and no one really trusts the accuracy of a physical exam documented in the medical record anyways) but AI hasn't been fully integrated yet in most practices. Once it does inevitably, demand for radiologists will drop significantly which will drive down pay, and the remaining radiologists will be expected to read at even faster speeds to make the same amount of pay (though with assistance from AI).

IR is another option but lifestyle will be much worse than DR and mores similar to ortho.

Ortho job market is subject to some encroachment by midlevels but as a surgical field PAs/NPs will have limited role in the OR and most of the role is helping with the non-surgical stuff (eg rounding on post-op patients on the floor or clinic, answering nursing patients or talking to families, seeing consults and writing consult notes). This will also decrease the number of ortho attendings needed on the market, but until we have robots that can perform orthopedic surgeries nearly independently the job market should not be as affected as by DR.

Of course its possible in any specialty to open way too many residency spots and saturate the job market (EM and rac onc being the most recent examples of this).

I appreciate your insight. One thing I don’t understand and I’d be interested to hear others thoughts is what happens when AI can suddenly make radiologists so efficient that way less are needed? I don’t think this will be unique to radiology either… there must be fields that require a considerable less amount of critical thinking that are ripe for the taking well before radiology. Will there suddenly be mass unemployment and the economy will get obliterated all in the name of technological advancement? Or are people arguing that healthcare, specifically imaging, is so costly that radiologists, governing bodies etc will just be ok with ( or have no control over) getting so drastically disrupted? Has this happened before where critical thinking professions have become obsolete or are people saying radiology doesn’t require that much critical thinking (which I’d be surprised to hear.)

 

[collegestud2013]

I appreciate your insight. One thing I don’t understand and I’d be interested to hear others thoughts is what happens when AI can suddenly make radiologists so efficient that way less are needed? I don’t think this will be unique to radiology either… there must be fields that require a considerable less amount of critical thinking that are ripe for the taking well before radiology. Will there suddenly be mass unemployment and the economy will get obliterated all in the name of technological advancement? Or are people arguing that healthcare, specifically imaging, is so costly that radiologists, governing bodies etc will just be ok with ( or have no control over) getting so drastically disrupted? Has this happened before where critical thinking professions have become obsolete or are people saying radiology doesn’t require that much critical thinking (which I’d be surprised to hear.)

Definitely not unique to DR, but diagnostic radiology is probably the most commonly cited field to be taken over by AI given it's seen as a non-procedural and non-patient facing specialty. Pathology is also cited as another field that can easily be taken over by AI as like radiology is also involves pattern recognition on images.

The non-procedural cognitive specialties like neurology, rheumatology, heme/onc, nephrology, endocrinology are also suspectable to AI playing a large part in treatment decision making; an in addition to midlevels taking over both combined can make physician demand shrink. However, since those are patient-facing specialties a big part of the job is communicating with and examining patients, which most patients still aren't comfortable in the near future being done by a robot/machine.

I doubt there would be suddenly be mass unemployment. The shift will be more gradual; over 1-2 decades the job markets in fields affected by AI and automation will become more tight and pay mediocre. During that time the role of many physicians will be different than what it is now, and many will probably end up leaving traditional clinical medicine and having to re-train one way or another to meet demands of the future job market.

 

[nibhighfootballrules]

Definitely not unique to DR, but diagnostic radiology is probably the most commonly cited field to be taken over by AI given it's seen as a non-procedural and non-patient facing specialty. Pathology is also cited as another field that can easily be taken over by AI as like radiology is also involves pattern recognition on images.

The non-procedural cognitive specialties like neurology, rheumatology, heme/onc, nephrology, endocrinology are also suspectable to AI playing a large part in treatment decision making; an in addition to midlevels taking over both combined can make physician demand shrink. However, since those are patient-facing specialties a big part of the job is communicating with and examining patients, which most patients still aren't comfortable in the near future being done by a robot/machine.

I doubt there would be suddenly be mass unemployment. The shift will be more gradual; over 1-2 decades the job markets in fields affected by AI and automation will become more tight and pay mediocre. During that time the role of many physicians will be different than what it is now, and many will probably end up leaving traditional clinical medicine and having to re-train one way or another to meet demands of the future job market.

So in your opinion radiology is one of the more if not the most risky field to enter in medicine? Basically a matter of can you get in and earn as much as possible before this shift begins? I guess if this does occur they will start decreasing residency spots in order to make it a smoother transition otherwise there will be a lot of people looking to pivot elsewhere...

 

[operaman]

So in your opinion radiology is one of the more if not the most risky field to enter in medicine? Basically a matter of can you get in and earn as much as possible before this shift begins? I guess if this does occur they will start decreasing residency spots in order to make it a smoother transition otherwise there will be a lot of people looking to pivot elsewhere...

I suspect the pattern will be similar to other technical advances. AI will let radiologists read faster and faster as the AI gets better and better. As this occurs over a number of years, the human will do less and less work per scan until eventually the AI drives most of it.

Nobody really knows the exact timeline of the above, and there’s usually a middle period where people can make serious bank from the new efficiency before reimbursements adjust. Hard to say what this means for a new grad. Possible you graduate into the golden era and crush it.

All that said, there’s definitely some security in a highly procedural field. I think surgery will eventually be automated too, but definitely not in my career.

 

[odyssey2]

I’ve never subscribed to the idea that AI will increase radiology efficiency to the point that it will significantly impact the job market. If the radiologist is still legally on the line for the scans that they read, they’re still going to go through every image no matter what the AI-generated report says. It might cut down on more mundane tasks like nodule-counting or filling in dose reports (God willing) but that would hardly make radiologists efficient to the point of job market concerns. The ceiling of a really good AI would be about the same as a good senior radiology resident, and even then academic attendings don’t blindly sign off on whatever they write. The only way AI makes a significant dent in supply and demand is if certain modalities are completely taken over by AI and the radiologist never touches them, which I won’t say isn’t within the realm of possibility but I don’t think we’re anywhere close to that.

 

[nibhighfootballrules]

I’ve never subscribed to the idea that AI will increase radiology efficiency to the point that it will significantly impact the job market. If the radiologist is still legally on the line for the scans that they read, they’re still going to go through every image no matter what the AI-generated report says. It might cut down on more mundane tasks like nodule-counting or filling in dose reports (God willing) but that would hardly make radiologists efficient to the point of job market concerns. The ceiling of a really good AI would be about the same as a good senior radiology resident, and even then academic attendings don’t blindly sign off on whatever they write. The only way AI makes a significant dent in supply and demand is if certain modalities are completely taken over by AI and the radiologist never touches them, which I won’t say isn’t within the realm of possibility but I don’t think we’re anywhere close to that.

This makes a lot of sense. What are your thoughts about the utility of AI then if liability (signing off) is still placed on the Radiologist? Like you said it can increase the efficiency of mundane tasks but if it requires Radiologists to re-read all scans it already read seems like it would actually slow efficiency. Also it could potentially bias Radiologists if it narrowed their thinking to only read certain elements. From my understanding there would have to be major studies examining all modalities and reviewing all pathologies and showing that AI is better or AI + Radiologist is better than Radiologists alone. What is also interesting is a ton of money is being poured into AI for this purpose so tech must think that AI could eventually replace or significantly augment Radiologists to the point it makes sense on a cost basis for healthcare systems to implement this tech. I wonder what they are betting will happen. Might just be my lack of knowledge in this area but seems like AI could potentially increase accuracy but would only decrease efficiency while Radiologists are still re-reading scans unless they aren't reading them the same way they would read a non-AI read scan.

Your last point is interesting. As others have mentioned though if we got to a point where we were utilizing technology that only AI was touching it seems like the market for Radiology would be the least of our concerns.

 

[odyssey2]

This makes a lot of sense. What are your thoughts about the utility of AI then if liability (signing off) is still placed on the Radiologist? Like you said it can increase the efficiency of mundane tasks but if it requires Radiologists to re-read all scans it already read seems like it would actually slow efficiency. Also it could potentially bias Radiologists if it narrowed their thinking to only read certain elements. From my understanding there would have to be major studies examining all modalities and reviewing all pathologies and showing that AI is better or AI + Radiologist is better than Radiologists alone. What is also interesting is a ton of money is being poured into AI for this purpose so tech must think that AI could eventually replace or significantly augment Radiologists to the point it makes sense on a cost basis for healthcare systems to implement this tech. I wonder what they are betting will happen. Might just be my lack of knowledge in this area but seems like AI could potentially increase accuracy but would only decrease efficiency while Radiologists are still re-reading scans unless they aren't reading them the same way they would read a non-AI read scan.

Your last point is interesting. As others have mentioned though if we got to a point where we were utilizing technology that only AI was touching it seems like the market for Radiology would be the least of our concerns.

So we already use AI in my residency and it's mostly used for Yes/No questions about a limited number of specific life-threatening pathologies like PEs or cervical spine fractures. It's good for confirming negative studies but false positives are common and when they happen they actually slow you down as you struggle to find something that isn't there. Overall it doesn't really increase speed since you still have to check it yourself, and at most gives you peace of mind if you call something negative. But when the AI is wrong, sometimes it's wildly wrong. There was a gigantic saddle embolus the AI didn't pick up since it basically opacified the entire pulmonary trunk. As for "biasing" radiologists these programs don't really do that since we're so heavily trained to look for pathology ourselves.

My understanding is yes, you would need to show increased or similar sensitivity and specificity for AI alone for basically every pathology with significant morbidity to get rid of radiologists entirely. This would be a massive undertaking even if the technology is there, and you would need a heterogenous data set from many different radiology sites with different scanners, different techs, different patient populations, etc, to effectively show this. This isn't happening anytime soon, even if AI somehow becomes advanced enough that this becomes a credible threat.

There's plenty of money to be made for AI in radiology without the explicit goal of replacing radiologists (not that that's not on the minds of many of these people). As they exist now they're mostly sold as discrete products to radiology practices/hospitals to assist the radiologists. Radiologists now are basically working at maximum efficiency by using templates, speech to text, macros, etc, that unless you take actually reading the image out of the equation, which, if the radiologist is still legally liable, won't happen, you can't really improve on efficiency as much as people seem to think.

I agree with your last point and this was basically the line I used whenever AI came up on my residency interviews. Radiology is a lot more subjective than people think, and often requires outside clinical context that sometimes doesn't exist. People love to make "correlate clinically" jokes, and there are definitely some overly hedgy rads out there, but the fact is you often can't tell exactly what that gray blob on the CT represents and need outside clinical context to point you in the right direction, or at least provide a differential. For an AI to have the sophistication to do this, to not only incorporate clinical context, information from previous reports, information from the clinician themselves (somehow), but also express uncertainty when the answer just isn't there, not to mention adjust for any mistakes the technologist made or mistakes in protocoling (which sometimes requires a call back) would require something that basically approaches general intelligence. If we get that, the radiologist job market will be the least of our worries.

 

[nibhighfootballrules]

Thanks for the thorough response. Was wondering for all of the residents and or attendings who have posted on this thread what is the ability of a DR to include procedures in their day to day? I know for MSK they can do joint injections and cyst aspirations and I've been told neuro will do lumbar taps. I was wondering what other procedures are possible and if you are substituting reading scans for patient facing procedures does your earnings take a significant hit because the volume just isn't comparable?

 

[GadRads]

Thanks for the thorough response. Was wondering for all of the residents and or attendings who have posted on this thread what is the ability of a DR to include procedures in their day to day? I know for MSK they can do joint injections and cyst aspirations and I've been told neuro will do lumbar taps. I was wondering what other procedures are possible and if you are substituting reading scans for patient facing procedures does your earnings take a significant hit because the volume just isn't comparable?

It depends on the practice. Generally everyone gets paid the same since we are providing a service. The most procedure-oriented will be IR (the most procedure-driven), breast, body imaging (biopsies) and MSK. In some practices the non-mammo procedures are done by IR except for MSK-ultrasound-guided procedures.

I was concerned about AI when I applied for radiology residency, but I couldn’t see myself doing anything else so I stayed the course. But once I started training and began to see the intricacies and nuances of imaging, I was no longer concerned about AI.

In my practice, our AI software specifically states it is not for diagnosis but for triaging, even though it is about 90% accurate (largely due to it beinging overly sensitive and making us waste time looking hard for what the AI system “sees” that is not really there). The system is however very good at excluding a very limited number of specific pathologies (intracranial large vessel occlusions, head bleeds, spine fractures and pulmonary emboli). But we must still look, and I have seen it miss, for example, subdural hemorrhages. It is these things that can get the radiologist in front of a jury. It is that rare critical miss that can ruin your career and will get lawyers chasing you.

Radiology volumes keep increasing since medicine is too addicted to imaging and we are reading insane volumes at max efficiency. The work days are now tiring and long. I think that’s the worst thing about the field, as well as the threat from corporatization and private equity. AI is not an issue.

 

[scubaboy]

As someone who just matched ortho (no bias) go into ortho!

 

[JuicedToTheSocks]

The more I have immersed myself in orthopedics I realized I loved the patient population and being in the OR. Additionally, I like the manual work involved in ortho. I would imagine myself getting a lot of satisfaction from going in and fixing peoples mechanical problems and being able to look at my work post-op and have hard proof that I made a difference. My main issue with orthopedics is the sacrifices required for this field. I am not someone who is adverse to hard work but I have begun thinking just because I am capable of enduring training etc doesn't mean it’s for me. I am a non-trad and started school later in life. I plan on having my first child prior to residency and having done orthopedic rotations working 80 hour weeks the thought of having such little time to spend with my future family seems extremely depressing. I know residency/fellowship is only a moment in time but those 6 years I'll never get back and starting later in life l think makes them even more valuable. Apart from residency I am not naive to think life is significantly that much more relaxed as an attending. First you have to build a practice, deal with OR booking time, and then maybe 10-15 years after training you can create a schedule that is more amenable to your personal goals but surgeons are always responsible for their patients and their schedule is inherently unpredictable to a certain extent. I am also concerned about the toll training and subsequent career will have on my physical and mental health and how it will impact my ability to be a good partner/father when I am not working. I cant tell if this is a healthy fear or something that will only get worse throughout my time training and make my life miserable. I find it difficult to have candid conversations with attendings about this because from my experience these concerns are taboo to talk about.

I don't know why ppl in this thread are not reading ur post???

Full disclosure: I'm applying ortho this yr but I would not do ortho/even surgery if I was a woman. I personally want a family (and not everyone might) but I know being a woman comes with a lot of unique circumstances that, as many woke minded ppl may try their darndest to argue, are unavoidable if u want a family.

Hell, I'm a single guy but can't even commit to a girl due to all the commitments and things u need to do to be a strong ortho applicant. I can't imagine being a woman trying to have kids, doing IVF, failing at it bc I shadowed IVF clinics and that sht was straight-up depressing, and trying to complete a brutal residency and then tough attendinghood for years after certification.

If I was a chick, I'd do roads.

I am also concerned about the toll training and subsequent career will have on my physical and mental health and how it will impact my ability to be a good partner/father when I am not working.

I don't want to ramble on abt this but this is an extremely important point. Not to mention finding a partner - which is definitely harder for ortho women v peds women.

 

[Splenda88]

I vote for ortho.

Operate only and then dump your patients into the hospitalists' lap and said well, I did my job; you guys have to fix the other issues.

 

[Cognovi]

Thanks for the thorough response. Was wondering for all of the residents and or attendings who have posted on this thread what is the ability of a DR to include procedures in their day to day? I know for MSK they can do joint injections and cyst aspirations and I've been told neuro will do lumbar taps. I was wondering what other procedures are possible and if you are substituting reading scans for patient facing procedures does your earnings take a significant hit because the volume just isn't comparable?

The program director association has agreed that the following procedures are expected to be within the skillset of a general radiologist:
Paracentesis
Thoracentesis
Thyroid FNA
Breast Bx
Lumbar Puncture (e.g. CSF drainage, myelography)
Arthrogram/Joint Aspiration/Injection
Image-guided core biopsies
Exchange catheter over a wire
US-guided venous access
Image guided abscess drainage

If you are substituting scans for procedures, your revenue generation takes a hit, mainly because of the inefficiency of the practice that are outside of your procedure time (eg, rooming the patient, cleaning the room, checking in on the patient before and after the procedure, waiting for anesthesia if necessary, etc.). In most practices, you can't go from room to room constantly engaged in doing a procedure, so there's down time. You could fill that downtime with reading scans but running back and forth can also be challenging. Whether that translates into your earnings takes a hit depends on the compensation model of your practice.

 

[Cognovi]

The AI concerns for rads are legit. Maybe not replacing rads in your career, but will surely increase MD efficiency which means you will be reading a LOT more scans and getting paid less per scan to do so. Just a function of how CMS values codes.

It's funny because when mammography started adding computer aided detection, it decreased efficiency (takes more time to review the results), it didn't improve accuracy (woops!), and CMS agreed to pay more per scan. Don't be so sure things will turn out one way!