I've just started seeing radiology NP CXR reads with a sign off by a supervising radiologist. I'm an M2 at a large regional center in the Midwest who is potentially interested in radiology, in part because people on this board and others have said the field is immune to midlevel takeover. Anyone else seeing this?
You are using an out of date browser. It may not display this or other websites correctly.
You should upgrade or use an alternative browser.
You should upgrade or use an alternative browser.
NPs reading chest X-Rays?
- Thread starter KoolKeith
- Start date
- Joined
- Jan 25, 2007
- Messages
- 1,077
- Reaction score
- 1,962
I think a PA or NP could read CXR perfectly well with some training and some oversight.
Most radiologists agree that midlevel encroachment is the greatest threat to radiology. Next would be potentially negative healthcare policy changes, and reimbursement cuts. The infamously hyped AI “threat” is really no threat at all and will be helpful, if any impact at all.
Midlevel encroachment in radiology is in its infancy as compared to fully maturing in most other fields of medicine. Even in surgical fields, PA/NP’s often see all the clinic patients. In radiology, the threat is more in doing procedures that we allow them to do.
As far as interpreting studies, it would be significantly more difficult for them to gain ground and we have the opportunity to nip it in its infancy, as long as we don’t voluntarily give up ground. It starts with reading CXR under supervision (like a resident), then moves to all radiographs, CT and other advanced imaging. Eventually they will be claiming they are competent to read independently... “look significant errors were caught by the supervising radiologist only 0.1% of cases....” etc. etc. This can only happen if enabled by radiologists agreeing to sign off these studies in the first place.
Midlevel encroachment in radiology is in its infancy as compared to fully maturing in most other fields of medicine. Even in surgical fields, PA/NP’s often see all the clinic patients. In radiology, the threat is more in doing procedures that we allow them to do.
As far as interpreting studies, it would be significantly more difficult for them to gain ground and we have the opportunity to nip it in its infancy, as long as we don’t voluntarily give up ground. It starts with reading CXR under supervision (like a resident), then moves to all radiographs, CT and other advanced imaging. Eventually they will be claiming they are competent to read independently... “look significant errors were caught by the supervising radiologist only 0.1% of cases....” etc. etc. This can only happen if enabled by radiologists agreeing to sign off these studies in the first place.
I think a PA or NP could read CXR perfectly well with some training and some oversight.
Yeah right. A lot of subtleties in reading CXRs, which is one of the more difficult studies to consistently read accurately and well. Easier to read a chest CT.
- Joined
- Jan 25, 2007
- Messages
- 1,077
- Reaction score
- 1,962
Yeah right. A lot of subtleties in reading CXRs, which is one of the more difficult studies to consistently read accurately and well. Easier to read a chest CT.
We can agree to disagree. I have practiced at PP and academic centers and the subtleties that us radiologists like to sit around and mentally masturbate about have near 0 clinical significance.
- Joined
- Oct 3, 2014
- Messages
- 390
- Reaction score
- 467
This is an interesting development in radiology -- I work with some of these radiology assistants (RAs). I think alarmism is kind of misguided, because it doesn't really understand what these people do and/or what radiology is like in 2020.
General radiology departments (x-rays, fluoro) barely make any money and most are money losers. You can thank obscenely low reimbursements for that (in addition to other factors). Reading radiographs takes a year to learn, a decade to learn how to read well, a lifetime to master. Most rads don't even learn to read them well because it's a long run for a short slide -- no bang for the buck. Radiographs often are a useful diagnostic tool if ordered appropriately, but no rad could sustain him or herself on radiograph reimbursements. They have to be subsidized by their colleagues. Add to this problem that a tremendous number of radiographs are pumped out every day, and a rad is just trying to get through that massive list as fast and efficiently as he or she can.
Enter the radiology assistant (not an NP... they would probably be too expensive and those credentials are unnecessary). They prelim reports similarly to a resident. The attending then agrees and signs off or changes the report and then signs it. The list is now much more manageable and instead of a gigantic time sink it's now only a big time sink. Everybody wins. The RA has a job, the rad can spend more time doing other more difficult work. It's especially valuable in academics because it used to be the residents' job to plow through all those radiographs, most of which had no learning potential. Routine scut work. Now they are free to pick up as many radiographs as they like and then move to the advanced stuff whenever they want. Residents love it.
The arguments against RAs are variations of the "thin edge of the wedge" argument -- RAs will put rads out of jobs. I see this as pretty unlikely. Radiologists are working at their max right now. Radiographs will not pay a basic MD salary so I'm not sure how an RA can put a rad out of job. A rad who reads mostly radiographs has to be subsidized by his or her colleagues. Volume is not going to decrease so rads will just read more and more cross-sectional imaging and fewer and fewer radiographs (directly). Another argument is that they will take their skills elsewhere to other departments (ortho, or something)... again, radiographs are not a money maker. Other departments are probably not going to jump at this as fast as people think.
I think RAs are less like CRNPs than like sonographers. Some of them can prelim as well. Would anyone in their right mind like to back to MDs scanning every patient? Can the sonographer go work for another department? Sure. Is there a low barrier in learning how to scan? Yes. Is it easy to scan well and keep yourself out of trouble? No. Does the quality of the sonographer make a difference? Depends.
Having worked with these RAs, they run a range. Some are awesome. Some are middling. A few are not so hot. They're pretty static for the most part. In that sense they seem different than sonographers which often get better with time. They're not like a resident reading up at night, following their cases. They do just enough to get by and get through the list. Also, unlike a sonographer, RAs don't tend to do everything. They work in just one area: chest, MSK, etc. Learning all of radiography would be a real challenge, but it could be done with training, I suppose.
There's an idea that they will somehow take over US, CT, MRI. US sonographers prelimming is def a possibility, but the stakes are a little higher on average I think, the job harder, the economic pressure less acute, and really they already kind of do this in many places. RAs in CT and MRI will likely not happen unless there is a major major change in something (like reimbursements fall to zero or something). These are just too hard and the RAs would take years of training even in just one body area. Even then they would probably be more of a liability than an asset. Some cross-sectional imaging will also just not tolerate a non rad -- like stroke CTs. By the time someone would want to get RAs in CT and MRI, AI will have obviated them in radiography, probably.
tl;dr: Radiographs are valuable diagnostic exams and an intellectual challenge, but if you're going into rads to read radiographs, you're not going to make ends meet. RAs are a way to stop subsidizing radiologists to read radiographs.
General radiology departments (x-rays, fluoro) barely make any money and most are money losers. You can thank obscenely low reimbursements for that (in addition to other factors). Reading radiographs takes a year to learn, a decade to learn how to read well, a lifetime to master. Most rads don't even learn to read them well because it's a long run for a short slide -- no bang for the buck. Radiographs often are a useful diagnostic tool if ordered appropriately, but no rad could sustain him or herself on radiograph reimbursements. They have to be subsidized by their colleagues. Add to this problem that a tremendous number of radiographs are pumped out every day, and a rad is just trying to get through that massive list as fast and efficiently as he or she can.
Enter the radiology assistant (not an NP... they would probably be too expensive and those credentials are unnecessary). They prelim reports similarly to a resident. The attending then agrees and signs off or changes the report and then signs it. The list is now much more manageable and instead of a gigantic time sink it's now only a big time sink. Everybody wins. The RA has a job, the rad can spend more time doing other more difficult work. It's especially valuable in academics because it used to be the residents' job to plow through all those radiographs, most of which had no learning potential. Routine scut work. Now they are free to pick up as many radiographs as they like and then move to the advanced stuff whenever they want. Residents love it.
The arguments against RAs are variations of the "thin edge of the wedge" argument -- RAs will put rads out of jobs. I see this as pretty unlikely. Radiologists are working at their max right now. Radiographs will not pay a basic MD salary so I'm not sure how an RA can put a rad out of job. A rad who reads mostly radiographs has to be subsidized by his or her colleagues. Volume is not going to decrease so rads will just read more and more cross-sectional imaging and fewer and fewer radiographs (directly). Another argument is that they will take their skills elsewhere to other departments (ortho, or something)... again, radiographs are not a money maker. Other departments are probably not going to jump at this as fast as people think.
I think RAs are less like CRNPs than like sonographers. Some of them can prelim as well. Would anyone in their right mind like to back to MDs scanning every patient? Can the sonographer go work for another department? Sure. Is there a low barrier in learning how to scan? Yes. Is it easy to scan well and keep yourself out of trouble? No. Does the quality of the sonographer make a difference? Depends.
Having worked with these RAs, they run a range. Some are awesome. Some are middling. A few are not so hot. They're pretty static for the most part. In that sense they seem different than sonographers which often get better with time. They're not like a resident reading up at night, following their cases. They do just enough to get by and get through the list. Also, unlike a sonographer, RAs don't tend to do everything. They work in just one area: chest, MSK, etc. Learning all of radiography would be a real challenge, but it could be done with training, I suppose.
There's an idea that they will somehow take over US, CT, MRI. US sonographers prelimming is def a possibility, but the stakes are a little higher on average I think, the job harder, the economic pressure less acute, and really they already kind of do this in many places. RAs in CT and MRI will likely not happen unless there is a major major change in something (like reimbursements fall to zero or something). These are just too hard and the RAs would take years of training even in just one body area. Even then they would probably be more of a liability than an asset. Some cross-sectional imaging will also just not tolerate a non rad -- like stroke CTs. By the time someone would want to get RAs in CT and MRI, AI will have obviated them in radiography, probably.
tl;dr: Radiographs are valuable diagnostic exams and an intellectual challenge, but if you're going into rads to read radiographs, you're not going to make ends meet. RAs are a way to stop subsidizing radiologists to read radiographs.
Last edited:
Interesting discussion, thanks for the replies. As one of you mentioned, it may be perfectly acceptable to train midlevels to “specialize” in interpreting one particular area of imaging, in much the way they do with central line teams, etc. With the implementation of this sort of industrial production model in healthcare what is to stop hospital administrators and educators from seizing the opportunity to produce “CT specialist” NP/PAs, or CXRs etc until every procedure and modality is parceled out to a less expensive midlevel who is an “expert” in that particular area, rather than have a radiologist doing everything? I could see this happening with surgery as well. Even surgeons say you can train a monkey to do simple surgeries, so what’s to stop this from happening?
Sent from my iPhone using SDN
Sent from my iPhone using SDN
Poor patient care. Many clinically important pieces of information present subtly. I've seen myeloma and lymphoma called on CXRs based on subtle findings many times by one of the experienced older rads, something that most of us and the younger attendings would miss because modern training emphasizes cross-sectional imaging. But money is the bottom line in healthcare and unfortunately, plain films don't make much money. It's also true that those sort of important findings might come 1/100-1/200 CXRs, so even though a miss may have a devastating outcome for those particular patients, the hospital admin doesn't care.
Overall there is still minimal midlevel penetration into diagnostic radiology, unlike general medicine (FM/IM/ED) and anesthesia where they are practicing completely autonomously in some cases. I doubt there will be any true penetration into the mainstay of modern radiology (i.e. cross-sectional imaging) mainly because medical subspecialists and surgeons already make complaints whenever a nonsubspecialized rads makes reads on their requested advanced imaging. A big part of radiology's clinical value is being able to pick up the phone and help answer questions from the doc on the other end and participating in tumor boards, something which takes a lot of medical knowledge and experience. Another reason why there will not be significant penetration in midlevel reads is because of how much midlevel penetration there is already into the ED and IM wards. There is a lot more dependence on radiology reads and subspecialty consults now than in the past because the clinicians running these departments have less experience. A single MD running a team of PAs in the emergency department doesn't have time to look over the imaging him/herself and relies heavily on a decent radiology read.
I am not worried in particular about midlevels in radiology but do feel that as a whole, physicians need to be more active in ensuring patients receive adequate care in all specialties. Right now things are held together because medical subspecialists, surgeons, radiologists, and pathologists do not have significant penetration by midlevels so complex procedures and the 'final say' are still mostly in the hands of docs. I would be worried for patients if any of these fields start to have significant midlevel penetration.
Overall there is still minimal midlevel penetration into diagnostic radiology, unlike general medicine (FM/IM/ED) and anesthesia where they are practicing completely autonomously in some cases. I doubt there will be any true penetration into the mainstay of modern radiology (i.e. cross-sectional imaging) mainly because medical subspecialists and surgeons already make complaints whenever a nonsubspecialized rads makes reads on their requested advanced imaging. A big part of radiology's clinical value is being able to pick up the phone and help answer questions from the doc on the other end and participating in tumor boards, something which takes a lot of medical knowledge and experience. Another reason why there will not be significant penetration in midlevel reads is because of how much midlevel penetration there is already into the ED and IM wards. There is a lot more dependence on radiology reads and subspecialty consults now than in the past because the clinicians running these departments have less experience. A single MD running a team of PAs in the emergency department doesn't have time to look over the imaging him/herself and relies heavily on a decent radiology read.
I am not worried in particular about midlevels in radiology but do feel that as a whole, physicians need to be more active in ensuring patients receive adequate care in all specialties. Right now things are held together because medical subspecialists, surgeons, radiologists, and pathologists do not have significant penetration by midlevels so complex procedures and the 'final say' are still mostly in the hands of docs. I would be worried for patients if any of these fields start to have significant midlevel penetration.
Last edited:
- Joined
- Oct 3, 2014
- Messages
- 390
- Reaction score
- 467
Interesting discussion, thanks for the replies. As one of you mentioned, it may be perfectly acceptable to train midlevels to “specialize” in interpreting one particular area of imaging, in much the way they do with central line teams, etc. With the implementation of this sort of industrial production model in healthcare what is to stop hospital administrators and educators from seizing the opportunity to produce “CT specialist” NP/PAs, or CXRs etc until every procedure and modality is parceled out to a less expensive midlevel who is an “expert” in that particular area, rather than have a radiologist doing everything? I could see this happening with surgery as well. Even surgeons say you can train a monkey to do simple surgeries, so what’s to stop this from happening?
Sent from my iPhone using SDN
In theory you could do this. In practice, it's not quite as easy as that since there really isn't a large supply of capable people. It's not a manual skill, this takes some level of thinking and the more complex, the more necessary it is to have the right person.
I think the PICC team is a good analogy. No IR is losing sleep because PICC teams are active. Could you train a PICC nurse or an IR tech to do the mechanical component of TACE? I guess... in theory. Not more than a few would really be capable, though. Bad thing will happen -- not that admins ever care about that argument.
D
deleted1139628
Im so sorry can you explain in diffrent words ?I've just started seeing radiology NP CXR reads with a sign off by a supervising radiologist. I'm an M2 at a large regional center in the Midwest who is potentially interested in radiology, in part because people on this board and others have said the field is immune to midlevel takeover. Anyone else seeing this?
Similar threads
- Replies
- 41
- Views
- 4K
