Pathology Informatics?

[wild0ne]

---

Members don't see this ad.

I was wondering if anyone has any insight into this emerging field? I've looked at the Association for Pathology Informatics website, and have a general understanding of the scope of the field.

However, if anyone on this board has any experience (fellow colleagues, anecdotes, etc.) about the field, I'd love to hear about it.

Are the job prospects for this field favorable? Is it needed in private practice? Would my anatomic skills go to waste if I chose this path?

Any insight would be appreciated!

 

[Thrombus]

I was wondering if anyone has any insight into this emerging field? I've looked at the Association for Pathology Informatics website, and have a general understanding of the scope of the field.

However, if anyone on this board has any experience (fellow colleagues, anecdotes, etc.) about the field, I'd love to hear about it.

Are the job prospects for this field favorable? Is it needed in private practice? Would my anatomic skills go to waste if I chose this path?

Any insight would be appreciated!

Seems like a great field -- but do you need an MD and residency training?

Seems like a bachelors/masters in computer science would make you much more useful.

 

[earmuff]

Seems like a great field -- but do you need an MD and residency training?

Seems like a bachelors/masters in computer science would make you much more useful.

This sort of resistance to developing the role of pathologists is what will make you obsolete.

Good riddance.

 

[ScotchBonnet]

This sort of resistance to developing the role of pathologists is what will make you obsolete.

Good riddance.

That may have been a little harsh, but I agree with you completely. As our chairman says, we pathologists must make ourselves useful to our colleagues. We typically do this as consultants that render tissue diagnoses, but we should be on the lookout for new opportunities to make ourselves indispensible.

Also for the OP, I met one informatics guy who said that you do not, as a pathologist, need to know the ins-and-outs of computer architecture, etc. You job is to bring medical knowlegde, understanding of laboratory workflow, how physicians want to see and use data, etc. to the development of new applications. (That's not verbatium but gives the gist of our conversation.)

 

[pathstudent]

Seems like a great field -- but do you need an MD and residency training?

Seems like a bachelors/masters in computer science would make you much more useful.

Are you serious?

In a few years it will probably be a board certified ABP subspecialty like molecular. Hopefully ABP will be quick to jump on that so they don't mess things up like they did with Molecular Pathology which they stalled and almost lost it to the genetics people.

 

[earmuff]

That may have been a little harsh

Yeah, you're right... it's just that some of these posts are like nails on a chalk board.

All I hear is blah blah blah... can't find work blah blah blah

Then, someone asks a legitimate question about a developing field of pathology, and this is the response they get...

The irony is palpable. The lack of insight is astonishing.

This is your ass, that is a hole in the ground.

 

[yaah]

People who are really good at informatics are tremendous assets to departments and even private groups. But it takes more than doing a fellowship, IMHO, to become good at informatics. Pathology as a specialty seems to be doing a pretty poor job of emphasizing informatics as a key to the future. There are some people who are on top of things, but a lot of them are so hard to understand that regular joes like us just end up getting confused.

 

[isaishere]

Are you serious?

In a few years it will probably be a board certified ABP subspecialty like molecular. Hopefully ABP will be quick to jump on that so they don't mess things up like they did with Molecular Pathology which they stalled and almost lost it to the genetics people.

Hey, I think this is possible too...... agree that present and future pathologists need to be more analytical, visionary....
we can't turn a blind eye to these changes..

I wonder, which programs are more willing to move forward, to bet on these areas

 

[Thrombus]

This sort of resistance to developing the role of pathologists is what will make you obsolete.

Good riddance.

Whats next? A fellowship in "billing"? How about one in "transcription"? Or one in "ER Path"? Do you show resistance to pathologists in these roles?

Good grief its one man's opinion. It seems like some people think pathologists don't have enough to learn already! I have never had a problem getting IT guys to help solve problems. I never have had the thought cross my mind that an IT trained pathologist sure would be nice!

 

[earmuff]

IT is not informatics.

👎

 

[earmuff]

People who are really good at informatics are tremendous assets to departments and even private groups. But it takes more than doing a fellowship, IMHO, to become good at informatics. Pathology as a specialty seems to be doing a pretty poor job of emphasizing informatics as a key to the future. There are some people who are on top of things, but a lot of them are so hard to understand that regular joes like us just end up getting confused.

agreed. It's difficult stuff, but it will integrate pathologists and make them central to patient care. Precisely what we need.

 

[LoaLoa]

Whats next? A fellowship in "billing"? How about one in "transcription"? Or one in "ER Path"? Do you show resistance to pathologists in these roles?

Good grief its one man's opinion. It seems like some people think pathologists don't have enough to learn already! I have never had a problem getting IT guys to help solve problems. I never have had the thought cross my mind that an IT trained pathologist sure would be nice!

I think you're misunderstanding what is meant by "pathology informatics". Your comment refers to "IT guys" who, based on your comment, must do things like work a computer support desk. "I can't open this document in Word. My monitor has this wierd line on it. How do I install WinZip on my laptop."

Pathology informatics would be more along the lines of "How can I get all the data from our laboratory information system into the hospital's medical record system? How can I best analyze this aggregate genetic array data? How can I create a decision support system for HPV test ordering" In other words, larger questions that require medical knowledge to answer.

This is currently a very fertile ground for innovation. Several informatics fellowships exist, but there is currently no solid framework for how someone becomes "trained" in pathology informatics. Most of us have learned our informatics skills on our own, but as time goes by, I would not be surprised to see training become more formalized. A boarded suspecialty certificate, as someone alluded to above, would be perfectly appropriate, in my opinion.

 

[pathstudent]

Exactly informatics is not I.T. or computer science. It is an ever expanding immensely important area of medicine. Pathology is in the driver's seat with taking control of informatics. If we drop the ball and another specialty takes it over it will further marginalize pathology. We need to move beyond the H&E slide. It is 200 years old.

I think you're misunderstanding what is meant by "pathology informatics". Your comment refers to "IT guys" who, based on your comment, must do things like work a computer support desk. "I can't open this document in Word. My monitor has this wierd line on it. How do I install WinZip on my laptop."

Pathology informatics would be more along the lines of "How can I get all the data from our laboratory information system into the hospital's medical record system? How can I best analyze this aggregate genetic array data? How can I create a decision support system for HPV test ordering" In other words, larger questions that require medical knowledge to answer.

This is currently a very fertile ground for innovation. Several informatics fellowships exist, but there is currently no solid framework for how someone becomes "trained" in pathology informatics. Most of us have learned our informatics skills on our own, but as time goes by, I would not be surprised to see training become more formalized. A boarded suspecialty certificate, as someone alluded to above, would be perfectly appropriate, in my opinion.

 

[wild0ne]

Whats next? A fellowship in "billing"? How about one in "transcription"? Or one in "ER Path"? Do you show resistance to pathologists in these roles?

Good grief its one man's opinion. It seems like some people think pathologists don't have enough to learn already! I have never had a problem getting IT guys to help solve problems. I never have had the thought cross my mind that an IT trained pathologist sure would be nice!

Ah, you gotta love Luddites.😍 I'm not sure if you're being facetious. If not, this is a good example of the inertia in medicine that contributes to its lack of relative technical progress.

Take EMRs for example. Most people that spend time on floors would agree that current EMRs do a very good job of storing information. If you need to look up vitals, labs, or imaging, it's now very easy to do it. However, if you have 10 minutes to understand the clinical course of a person that's been in the MICU for 20 days, it's still tedious. The problem is that these EMRs were developed by your "IT guys" with very little understanding how clinicians actually use the software.

Similar concept in pathology. Informatics training does not focus developing the technology (e.g., writing code, building architecture, debugging, etc.). Rather, you're a bridge between pathologists/clinicians and the engineers. It's your job to choose (help develop, or even develop if you have the skills) the appropriate technology to make the workflow more efficient.

It's not surprising that you have never felt the need for an "IT trained pathologist". I think many healthcare workers don't realize how much of their job can be simplified by software. It's only a matter of time before microscopes are replaced by computers. And yes, this has been said many times before. However, the infrastructure and cost of digital slides has been sufficiently prohibitive. As memory becomes cheaper, software more advanced, and bandwidth more available, we'll likely see this change.

I'm just wondering if folks in practice (academic or private) are finding a budding need for pathologists with formal training. My argument above is just my opinion. It would be helpful to have some real-world stories/experiences.

 

[raider]

Thrombus dude don't worry, the "little kiddies" are bedazzled by the academic "bull****" they are fed everyday, let them step in the real world "if they have guts to get out of the academic pit" and they will find out much to their chagrin that the pathologist who they ardently listened to ranting about "the revolutionary field of pathology informatics" was a grade A loser who could never cut it out as a "diagnostician" (which pathologists really are) and decided to make it as an "IT" guy. This is exactly the same bull**** that goes on in academic departments under the name of teaching pathologists "manager skills" or a "molecular genetic fellowship" for that matter. All you need to know in the "real world" is that the "tissue" is yours then you can hire (for really low wages) the IT guy, the manager and the pHD and direct them (as a general directs his forces) on what you want them to do to get to "your goal" . Bottom line is he who controls the "tissue" gets to control everyone else and this is where our efforts should be directed viz. being the clinician who controls tissue diagnosis.

By the way for doubters, I get to whip the pHD person(who runs the molecular lab), the IT person ( who handles all the problems with cerner,copath or whatever) and the pathology manager and since they know where there is salary is coming from (the revenue I generate by making diagnoses on the very valuable H&E) they act as a good slave should .

My advice, leave all this bull**** to bull****ters, concentrate on being an "EXCELLENT DIAGNOSTICIAN" because that is your real job which no one else but you can do. If you do want to learn something useful, learn the business aspects e.g. how to bid for a hospital contract, how to generate revenue, how to setp up a lab etc (aka how to maximize profit) Do something which only you can do because of your MD degree and diagnostician training.

And NEVER EVER BASH THE H& E; IT IS CHEAP AND GIVES MORE INFO THAN THAT FANCY 4000 DOLLAR MULTIGENE TEST (WHICH THE UNSCRUPULOUS BIOTECH COMPANY HAS SOLD AS THE GREATEST THING SINCE THE WHEEL TO THE "CLUELESS CLINICIAN"). BUT TO GET THAT INFO YOU NEED TO BE A STELLAR DIAGNOSTICIAN. SO SPEND YOUR EFFORTS IN BECOMING THAT DIAGNOSTICIAN. THE H& E IS NOT GOING ANYWHERE. SOON THE CLINICIANS WILL BE IN AWE OF YOUR DIAGNOSTIC SKILLS AND YOU DICTATE WHAT NEEDS TO BE DONE TO GET THE DISEASE DIAGNOSED.

 

[KeratinPearls]

Thrombus dude don't worry, the "little kiddies" are bedazzled by the academic "bull****" they are fed everyday, let them step in the real world "if they have guts to get out of the academic pit" and they will find out much to their chagrin that the pathologist who they ardently listened to ranting about "the revolutionary field of pathology informatics" was a grade A looser who could never cut it out as a "diagnostician" (which pathologists really are) and decided to make it as an "IT" guy. This is exactly the same bull**** that goes on in academic departments under the name of teaching pathologists "manager skills" or a "molecular genetic fellowship" for that matter. All you need to know in the "real world" is that the "tissue" is yours then you can hire (for really low wages) the IT guy, the manager and the pHD and direct them (as a general directs his forces) on what you want them to do to get to "your goal" . Bottom line is he who controls the "tissue" gets to control everyone else and this is where our efforts should be directed viz. being the clinician who controls tissue diagnosis.

By the way for doubters, I get to whip the pHD person(who runs the molecular lab), the IT person ( who handles all the problems with cerner,copath or whatever) and the pathology manager and since they know where there is salary is coming from (the revenue I generate by making diagnoses on the very valuable H&E) they act as a good slave should .

My advice, leave all this bull**** to bull****ters, concentrate on being an "EXCELLENT DIAGNOSTICIAN" because that is your real job which no one else but you can do. If you do want to learn something useful, learn the business aspects e.g. how to bid for a hospital contract, how to generate revenue, how to setp up a lab etc (aka how to maximize profit) Do something which only you can do because of your MD degree and diagnostician training.

Totally agree.

 

[Sulfinator]

. . . YOU NEED TO BE A STELLAR DIAGNOSTICIAN. SO SPEND YOUR EFFORTS IN BECOMING THAT DIAGNOSTICIAN. THE H& E IS NOT GOING ANYWHERE.

I think raider is right about that. Those seem like words of wisdom to me. I just wish they weren't in all caps. 😉

 

[LoaLoa]

. . . the "little kiddies" are bedazzled by the academic "bull****" . . ."grade A looser who could never cut it out as a "diagnostician" (which pathologists really are)" . . ."I get to whip the pHD person(who runs the molecular lab), the IT person ( who handles all the problems with cerner,copath or whatever) and the pathology manager and since they know where there is salary is coming from (the revenue I generate by making diagnoses on the very valuable H&E) they act as a good slave should ."

While I agree with the point of your post (don't forget about the fundamentals), I've gotta ask why you need to be such a dick about it.

 

[earmuff]

Bottom line is he who controls the "tissue" gets to control everyone else and this is where our efforts should be directed viz. being the clinician who controls tissue diagnosis.

It's not about controlling "tissue" as much as it's about controlling and improving testing and reporting.

Big pictures are tough, I know. Chicken little has trouble seeing anything but the big blue sky.

 

[pathstudent]

google University of Pitt informatics for more information. They are the leaders of informatics. er.

 

[yaah]

You guys do realize that informatics expertise can increase your profit margin and grow your business in private practice, right? Discounting it for anyone outside of academics is something you do at your own peril. If your group has someone who is good at informatics, system integration, speaking the language, etc, your group's bargaining power and ability to profit will benefit. It is all well and good to employ competent individuals (like managers) who also have many of these skills, but that isn't everything. But hey, good luck with that attitude! Short-sighted and narrow thinking have been working out great for businesses for decades. Just don't come crying here when your group starts losing business and influence.

By the way, for reference, here's a definition:

Loser (noun): Someone who spells the word "loser" as "looser."

 

[Patho2009]

los·er<SCRIPT>play_w2("L0254800")</SCRIPT><OBJECT style="MARGIN: 1px" codeBase="http://fpdownload.macromedia.com/pub/shockwave/cabs/flash/swflash.cab#version=6,0,0,0" classid=clsid27cdb6e-ae6d-11cf-96b8-444553540000 width=13 height=21>
























<embed src="http://img.tfd.com/m/sound.swf" FlashVars="sound_src=http://img.tfd.com/hm/mp3/L0254800.mp3" menu="false" width="13" height="21" wmode="transparent" type="application/x-shockwave-flash" pluginspage="http://www.macromedia.com/go/getflashplayer"></OBJECT> (l

z

r) </P>n. 1. a. One that fails to win: the losers of the game.
b. One who takes loss in a specified way: a graceful loser; a poor loser.

2. a. One that fails consistently, especially a person with bad luck or poor skills: "losers at home seeking wealth and glory in undeveloped countries" (Arthur M. Schlesinger, Jr.)
b. One that is bad in quality: That book is a real loser.



loose<SCRIPT>play_w2("L0248400")</SCRIPT><OBJECT style="MARGIN: 1px" codeBase="http://fpdownload.macromedia.com/pub/shockwave/cabs/flash/swflash.cab#version=6,0,0,0" classid=clsid27cdb6e-ae6d-11cf-96b8-444553540000 width=13 height=21>
























<embed src="http://img.tfd.com/m/sound.swf" FlashVars="sound_src=http://img.tfd.com/hm/mp3/L0248400.mp3" menu="false" width="13" height="21" wmode="transparent" type="application/x-shockwave-flash" pluginspage="http://www.macromedia.com/go/getflashplayer"></OBJECT> (l

s) adj. loos·er, loos·est 1. Not fastened, restrained, or contained: loose bricks.
2. Not taut, fixed, or rigid: a loose anchor line; a loose chair leg.
3. Free from confinement or imprisonment; unfettered: criminals loose in the neighborhood; dogs that are loose on the streets.
4. Not tight-fitting or tightly fitted: loose shoes.
5. Not bound, bundled, stapled, or gathered together: loose papers.
6. Not compact or dense in arrangement or structure: loose gravel.
7. Lacking a sense of restraint or responsibility; idle: loose talk.
8. Not formal; relaxed: a loose atmosphere at the club.
9. Lacking conventional moral restraint in sexual behavior.
10. Not literal or exact: a loose translation.
11. Characterized by a free movement of fluids in the body: a loose cough; loose bowels.

adv. In a loose manner.

v. loosed, loos·ing, loos·es
v.tr. 1. To let loose; release: loosed the dogs.
2. To make loose; undo: loosed his belt.
3. To cast loose; detach: hikers loosing their packs at camp.
4. To let fly; discharge: loosed an arrow.
5. To release pressure or obligation from; absolve: loosed her from the responsibility.
6. To make less strict; relax: a leader's strong authority that was loosed by easy times.

v.intr. 1. To become loose.
2. To discharge a missile; fire.

Idiom: on the loose 1. At large; free.
2. Acting in an uninhibited fashion

 

[lipomas]

Oh man, the lack of respect for informatics and the ignorance on this thread is incredible. By all means get a great IT guy or gal. But don't confuse that with informatics. And don't just assume that it is all going to work out for you if you are a "good diagnostician." But sure, hey, all of us who think it's important are truly "loosers."

What we really need is someone who can teach and explain informatics to people who are curious but don't have a math PhD background. There is this guy at Yale (Sinard or Simard?) who wrote a book about informatics which was the closest anyone has come to making it tangible. But unfortunately what it focused on was the IT part of it and when it came to relating it to pathology and business successs it trailed off. I read it and understood networks a little better, but the network guys take care of the networks. Being well versed in it is nice though. And every other time I have gone to an informatics lecture I feel like I may as well be listening to a quantum mechanics lecture. In Russian. And I don't speak Russian.

 

[Mags74]

I was wondering if anyone has any insight into this emerging field? I've looked at the Association for Pathology Informatics website, and have a general understanding of the scope of the field.

However, if anyone on this board has any experience (fellow colleagues, anecdotes, etc.) about the field, I'd love to hear about it.

Are the job prospects for this field favorable? Is it needed in private practice? Would my anatomic skills go to waste if I chose this path?

Any insight would be appreciated!

i too am seeking to specialize in pathology informatics -- it's good to hear from others who are interested in going into it as well. i do have a phd in cancer biology with an emphasis on molecular epidemiology, and have a fuzzy vision of how to combine that training with a career in pathology.

any feedback regarding residency/fellowship training programs that are good for someone who wants to go into the informatics end of path? i saw the post about the pitt program -- many thanks.

i've also wondered whether my lack of background in computer science will be a hinderance in this career track.

 

[Freak]

For path informatics, MGH and Yale also look like they have tracks.

There is a conference every year (APIII) which might be of value ... next year it is in Boston. Path Informatics is not simply limited to LIMS (CP), but has some potential in AP as well (albeit imaging is a much harder problem than number crunching).

 

[TMZ2007]

I'm currently rotating on informatics (which includes some biostats in my program). Is there anything I should be learning (besides the stats) that is relevant for AP/CP board exams?

 

[BrainPathology]

I think a part of the job problem has been explained to me by this thread. I will be looking for a job in 2011. I agree with Raider, Thrombus and their allies that the jobs they want (a job in 1960) are nearly impossible to find these days.

 

[BrainPathology]

I have to apologize, I actually came to this thread to offer something at least partially relevant. Some of the responses were just so breathtaking in their ignorance and irony that my first response came first - not that I would take it back. I don't think it's possible to be too hard on our colleagues who ignore or dismiss this field.

There are few enough people who go through med school and then pathology residency who want to do informatics training that I think it would be extremely valuable. In any lab there are probably hundreds of thousands if not millions of dollars waiting to be saved by a competent pathologist with informatics training and some good ideas. I've seen some of these types of changes happen at my medical school and the figures given for the savings were substantial. Couple that with the (irrelevant to most on this board I understand but possibly interesting to a minority of doctors here) improvements to actual patient care that this field can offer, and you can craft yourself a satisfying and idespensible career.

I had a colleague leave my residency to move to a program closer to a computer-based business he was starting. The last I heard he HAD a fellowship in this area. If the OP would like to private message me I'll do what I can to put you in touch with them. I believe one of the neuropathologist attendings at UCLA is also very much interested in informatics (from memory at AANP conferences - I don't remember his name) and an inquiry there might be a place to start as well.

 

[yaah]

I'm currently rotating on informatics (which includes some biostats in my program). Is there anything I should be learning (besides the stats) that is relevant for AP/CP board exams?

There is very little on the boards on informatics. I would learn a little bit about operating systems, information systems, general things like that. If I remember right you may get occasional questions about RAM and ROM and different types of interfaces. These questions unfortunately are very good at suggesting to people that all you need to know about informatics is what the IT guy knows. Brainpathology's comments on the matter are more important, but it is also hard to test that stuff.

 

[path19]

This certainly doesn't apply to all of informatics per se, but is 100% true. The 1 and only person I have ever known to do an informatics fellowship (at a very well respected institution) is now employed as a PA. (And they are AMG for what it's worth)

 

[2121115]

The 1 and only person I have ever known to do an informatics fellowship (at a very well respected institution) is now employed as a PA. (And they are AMG for what it's worth)

OMG. I would be hammering down the door to any other residency that would take me at that point. Wow.

 

[wild0ne]

For path informatics, MGH and Yale also look like they have tracks.

There is a conference every year (APIII) which might be of value ... next year it is in Boston. Path Informatics is not simply limited to LIMS (CP), but has some potential in AP as well (albeit imaging is a much harder problem than number crunching).

Yeah, I'm planning on attending that next year (program permitting). It'll be exciting to hear what some of the leaders have to say regarding this field. I'm also gearing up towards an elective during my 3rd or 4th year.

If anyone happened to read the 1/20 issue of CAP Today, there was an article from Dr. Henricks (from Cleveland Clinic), arguing some of the obstacles in implementing digital pathology. It's an interesting read, mainly because it gives an opposing view. I wasn't convincingly swayed, however. As BrainPathology already mentioned, there is a significant amount of money to be saved (and made) with proper implementation of digital pathology. As soon as groups accept this (and no longer feel threatened by the change), we'll see more adoption. To the naysayers (i.e., those that "control" the tissue), I hope you keep your minds open. Our field has enough short-sightedness.

 

[path24]

This certainly doesn't apply to all of informatics per se, but is 100% true. The 1 and only person I have ever known to do an informatics fellowship (at a very well respected institution) is now employed as a PA. (And they are AMG for what it's worth)

Not surpised. Pathology is the only field I have ever heard of Doctors practicing as PAs.....pathetic.

Just more proof of an active chronic poor job market.

 

[BrainPathology]

Yeah, I'm planning on attending that next year (program permitting). It'll be exciting to hear what some of the leaders have to say regarding this field. I'm also gearing up towards an elective during my 3rd or 4th year.

If anyone happened to read the 1/20 issue of CAP Today, there was an article from Dr. Henricks (from Cleveland Clinic), arguing some of the obstacles in implementing digital pathology. It's an interesting read, mainly because it gives an opposing view. I wasn't convincingly swayed, however. As BrainPathology already mentioned, there is a significant amount of money to be saved (and made) with proper implementation of digital pathology. As soon as groups accept this (and no longer feel threatened by the change), we'll see more adoption. To the naysayers (i.e., those that "control" the tissue), I hope you keep your minds open. Our field has enough short-sightedness.

I'm not sure I believe digital pathology would dominate a career in informatics. I was more thinking of things that integrate the process of receving, processing and tracking to reduce errors and reduce turnaround time and also cut down on unnecessary tests. As long as you still have to make a glass slide that then gets digitized I don't see digital pathology taking over that much. No matter how cheap it gets it's still an extra step added to a process that works fine right now. It has it's great uses in archiving slides that demonstrate things for places like tumor boards or second opinions sometimes. And I don't think these have been fully taken advantage of. But right now, if there is a glass slide somewhere you might as well have the pathologist read them first.

 

[pathstudent]

I'm not sure I believe digital pathology would dominate a career in informatics. I was more thinking of things that integrate the process of receving, processing and tracking to reduce errors and reduce turnaround time and also cut down on unnecessary tests. As long as you still have to make a glass slide that then gets digitized I don't see digital pathology taking over that much. No matter how cheap it gets it's still an extra step added to a process that works fine right now. It has it's great uses in archiving slides that demonstrate things for places like tumor boards or second opinions sometimes. And I don't think these have been fully taken advantage of. But right now, if there is a glass slide somewhere you might as well have the pathologist read them first.

Then how are we supposed to sign out cases in our pajamas or while sitting on the beach sipping on tropical drinks? Isn't that what they were promising us?

 

[lipomas]

This certainly doesn't apply to all of informatics per se, but is 100% true. The 1 and only person I have ever known to do an informatics fellowship (at a very well respected institution) is now employed as a PA. (And they are AMG for what it's worth)

From what I have seen, there are two types of people who do informatics fellowships:

1) The people who are really good at it and have some sort of prior experience that relates to it.

2) The people who don't know what else to do so they decide to do an informatics fellowship.

Those in the first category are usually sought after by academic institutions, or they are smart enough to use their expertise to become entrepreneurs or whatever. Those in the second category may as well have not done the fellowship.