2018 NRMP match for pathology

[Doormat]

Looks like another down year for pathology in the 2018 NRMP match. Data is available on the NRMP website. 601 pathology residency spots were offered nationwide. Of those, 237 filled with US seniors (MDs and DOs), which translates into a 36.6% fill rate. I think last year the fill rate was 35%. Again, these statistics are the lowest among all major specialties. 20 programs did not fill. Over 1 in 20 pathology residency spots did not fill at all.

I'd like to spin this is a positive way. Not unlike dermatology, which has tightly limited the number of residency spots every year, the very low interest in pathology among US grads is essentially reducing the number of trainees eligible to enter the job market every year.

My friend in Anesthesia said that the huge drop in applicants in the mid 1990s has contributed to a significant and ongoing shortage of Anesthesiologists, which benefits that field right now.

Maybe in 10 years there won't be enough pathologists to signout GI cases for $27 for a global bill, with the rest as kickback money to the referring gastroenterologist.

---

Members don't see this ad.

 

[dukeresident]

Looks like another down year for pathology in the 2018 NRMP match. Data is available on the NRMP website. 601 pathology residency spots were offered nationwide. Of those, 237 filled with US seniors (MDs and DOs), which translates into a 36.6% fill rate. I think last year the fill rate was 35%. Again, these statistics are the lowest among all major specialties. 20 programs did not fill. Over 1 in 20 pathology residency spots did not fill at all.

I'd like to spin this is a positive way. Not unlike dermatology, which has tightly limited the number of residency spots every year, the very low interest in pathology among US grads is essentially reducing the number of trainees eligible to enter the job market every year.

My friend in Anesthesia said that the huge drop in applicants in the mid 1990s has contributed to a significant and ongoing shortage of Anesthesiologists, which benefits that field right now.

Maybe in 10 years there won't be enough pathologists to signout GI cases for $27 for a global bill, with the rest as kickback money to the referring gastroenterologist.

Assuming that AI automation hasn’t reduced the number of pathologists needed to handle the work flow. Unlike some others, I see both pathology and radiology both threatened by AI. I expect the numbers needed to drop significantly.

 

[WEBB PINKERTON]

Gas is suffering from mid-level encroachment. With nursing degrees now the equivalent of biology degrees in the eyes of CMS, it sounds as if nurses may be spreading into lab medicine. Will a nursing midlevel take turf from pathologists?

 

[midi12]

Not too familiar with the US matching system - how many of the unfilled spots are typically taken by IMGs after the match? Here in Canada it seems path has actually been on a bit of an upswing over the last 5 years due to an overall increase in CaRMS competitiveness and some cuts in residency spots. After the first round of CaRMS there were only 3 unfilled spots outside Quebec in 2018 vs 15 in 2014. Hopefully this bodes well for the stature of the field here in the future. Seems like many residency spots need to be cut in the US in order to regain any kind of market leverage.

 

[Unty]

Dont worry, there are tons of foreign grads that are more than happy to fill the empty slots.

 

[jupiterianvibe]

35% AMG fill rate = 65% FMG fill rate.

There is no such thing as a pathology residency position that goes unfilled, and why would there be? It would be like giving away free money!

This mess is on the CAP, the ASCP, the academics, and each of us for allowing the status quo to continue without abatement as long as we get ours.

There's no way that anyone can assert that our field is NOT providing substandard care, given the number of FMGs matching to positions.

 

[coroner]

^^^This^^^ Some other warm body will just fill those spots

601 pathology residency spots were offered nationwide.

WTH...It was slightly over 500, circa 2008. It was way too many back then; but, now we need 600 graduating/yr...

Will a nursing midlevel take turf from pathologists?

Nah, they'll see how crappy the job market is and realize our turf's not worth it...haha. Plus, according to you, it's more likely bots will replace us...

 

[path24]

The unfilled spots will go to people that will eventually work for peanuts...hence GI offer. And everyone gets to compete with that. No jobs. Massive oversupply. Lots of competition for jobs/work.

"We have filled our open Pathologist position. We did select an individual that responded to the ad posted on your service site. In addition, your prompt turnaround in posting the position was paramount in expediting the candidate search process. The quality of the candidates who responded were topnotch." L. Tatum, Bowling Green Associated Pathologists, Kentucky, 13 March 2018

"Advertising only on your site provided almost more candidates than I could deal with and we were able to select from many highly qualified individuals. I will certainly recommend using you guys to anyone looking in the future!" R. Forsythe, M.D., Boulder Valley Pathology, Colorado, 1 February 2018

Another proud match year for academics, CAP, ASCP, Quest, Labcorp...etc.

 

[Thrombus]

IDIOT ACADEMICS HAVE TANKED THE FIELD OF PATHOLOGY TURNING US INTO THE INDENTURED SERVANTS OF BIG HOSPITAL, BIG CORPORATE, BIG ACADEMIA.

FLEE PATHOLOGY NOW!!!!!!

 

[LADoc00]

I love the Doomsayer posts, even after all these years on SDN they still warm my heart.

 

[dermpathdoc]

Dont worry, there are tons of foreign grads that are more than happy to fill the empty slots.

But some IMG's that are US graduates don't know how to call to ask about open positions? Just trying to help out an acquaintance

 

[coroner]

But some IMG's that are US graduates don't know how to call to ask about open positions? Just trying to help out an acquaintance

??? An IMG by definition is someone who is not a US graduate.

 

[tensoranalysis]

Assuming that AI automation hasn’t reduced the number of pathologists needed to handle the work flow. Unlike some others, I see both pathology and radiology both threatened by AI. I expect the numbers needed to drop significantly.

One of the biggest hurdles for AI in pathology and radiology are large, curated data sets for every disease that you want it to be able to assess. We are talking hundreds to thousands of each case to encompass biological variability. Then you will likely need years to validate its performance in a clinical setting - so probably a clinical trial. We don't even know how AI will perform when given so many classifiers as options - most of the studies you are looking at are testing AI with binary classifiers (cancer vs not cancer). The metrics frequently used, ROC curves, don't really make sense when comparing performance to a human pathologist. What pathologist do you know that can vary their sensitivity or specificity?

AI will likely be integrated into our workflow in the future to make our lives easier and improve our diagnostic capabilities, but like gene therapy, stem cell therapy, and targeted cancer therapy - we have a way with overestimating what technology is ultimately capable of, especially when that technology is just in its infancy!

 

[Violetta80]

Sad to read comments like this...I am IMG and will follow my passion with pathology (luckily I am very confident my future program seems to know and appreciate that). My country is accepting IMGs in a different way ( and coming from an economically very strong, european country I can say that I really came for very specific reasons) - I would hope you guys keep things like this in mind.

 

[LADoc00]

Sad to read comments like this...I am IMG and will follow my passion with pathology (luckily I am very confident my future program seems to know and appreciate that). My country is accepting IMGs in a different way ( and coming from an economically very strong, european country I can say that I really came for very specific reasons) - I would hope you guys keep things like this in mind.

Some of my favorite residents were IMGs. One incident though sticks out. I had gone out drinking hard one night and got to work around 7am the next day near incoherent via taxi and fairly certain I was gonna get booted from the program. I think I was even crying..I was that far gone. An IMG, an incredible woman from India, actually baby-fed me scrambled eggs and coffee until I came around. I may not even be here posting if it were not for IMGs. That's a true story folks.

Sheesh..I got all teary eyed remembering that so I have to hide in my office for the next 15 min...

 

[Pathbusiness]

Some of my favorite residents were IMGs. One incident though sticks out. I had gone out drinking hard one night and got to work around 7am the next day near incoherent via taxi and fairly certain I was gonna get booted from the program. I think I was even crying..I was that far gone. An IMG, an incredible woman from India, actually baby-fed me scrambled eggs and coffee until I came around. I may not even be here posting if it were not for IMGs. That's a true story folks.

Sheesh..I got all teary eyed remembering that so I have to hide in my office for the next 15 min...

You do make sense sometimes.

 

[dukeresident]

One of the biggest hurdles for AI in pathology and radiology are large, curated data sets for every disease that you want it to be able to assess. We are talking hundreds to thousands of each case to encompass biological variability. Then you will likely need years to validate its performance in a clinical setting - so probably a clinical trial. We don't even know how AI will perform when given so many classifiers as options - most of the studies you are looking at are testing AI with binary classifiers (cancer vs not cancer). The metrics frequently used, ROC curves, don't really make sense when comparing performance to a human pathologist. What pathologist do you know that can vary their sensitivity or specificity?

AI will likely be integrated into our workflow in the future to make our lives easier and improve our diagnostic capabilities, but like gene therapy, stem cell therapy, and targeted cancer therapy - we have a way with overestimating what technology is ultimately capable of, especially when that technology is just in its infancy!

I doubt that the large datasets will be necessary to be better and faster than human pathologists. AI is not standing still.

 

[tensoranalysis]

I doubt that the large datasets will be necessary to be better and faster than human pathologists. AI is not standing still.

Yes it is not, and in the future large data sets may not be necessary, but they are a requirement now. I guess what we are now arguing is 'if AI ever be able to do ...'

sure why not? We have been able to make a lot happen with enough time and effort. With the tech's current trajectory, assuming consistent advances, we are decades away from having an AI that can arrive at a precise diagnosis in a space of multiple classifiers in a clinical setting. There are so many hurdles/problems and these just take time to solve - likely with compromises to ultimate performance, but yea i'm not saying it won't happen.

 

[Pathbusiness]

"Microscope will be obsolete" is something I have been hearing for two decades. No one can scan a slide faster than cheaper than a pathologist. All other things will be tools to use.

 

[Violetta80]

Some of my favorite residents were IMGs. One incident though sticks out. I had gone out drinking hard one night and got to work around 7am the next day near incoherent via taxi and fairly certain I was gonna get booted from the program. I think I was even crying..I was that far gone. An IMG, an incredible woman from India, actually baby-fed me scrambled eggs and coffee until I came around. I may not even be here posting if it were not for IMGs. That's a true story folks.

Sheesh..I got all teary eyed remembering that so I have to hide in my office for the next 15 min...

heart-breaking story Now I am super satisfied.