What exactly is a clinical informatics fellowship?

This forum made possible through the generous support of SDN members, donors, and sponsors. Thank you.

doctorendgame

Full Member
Joined
Apr 16, 2025
Messages
15
Reaction score
3
I came across it and it seems to be a fellowship that doesn't focus on patient care, but trains doctors in population health? What exactly does this fellowship open doors for? It seems like it's like an MPH, but tailored for doctors.

From what I can tell, any specialty can pursue it: primary care, pathology, ID, surgery. The training itself is not clinical, but fellows do practice in their residency field so their clinical skills don't get rusty.
 
Last edited:
It's data analysis to.find patterns and optimize care. Probably to develop and monotor AI programs that do this.
How well known is this field in the medical community? It looks like it's more known among IT and software engineers than doctors or adcoms.
 
It's not exactly unknown so much as it is emerging.

It also resists definition: clinical informatics, to some, can mean EMR engineering—learning to apply software engineering/computer science to clinical platforms. To others, clinical informatics means bioinformatics: using everything from blastx to AlphaFold to CRISPR to solve medical challenges. In my own experience, one of my informatics experiences lied outside of biology and neuroscience altogether and was pure organic/synthetic chemistry. I have also seen informatics associated with the humanities through epidemiology and statistics.

I've only ever met one physician, an anesthesiologist MD-PhD in clinical informatics out of a T10. They are highly dissatisfied with their path and argue that the focus is impractical. They have been on the board of some prominent tech companies, but they argue that doctors don't really understand (or care to understand) AI. Most feel that the promises of AI in reducing physician burnout are myopic at best: even if they do work the way early adopters say it will, it will only encourage hospital executives to drive up physician workload and encourage them to make do with less personnel support. He painted a grim picture for the field.

I personally disagree with him in outlook, but can only give a resigned shrug in practice. I've seen doctors express strong ambivalence toward AI, but Dragon used to feel like the sixth circle of Hell and has gotten better; dot-phrases radically improved workflows even if only by a couple clicks; periodic improvement has shown that implementation is ambivalent and then improves over time.

As far as clinical opportunities—for the most part, I don't think it entitles you to practice any specific type of medicine. Most people leverage it to reach into the health tech space...but if that's your goal, there are easier ways than going through medicine.
 
As far as clinical opportunities—for the most part, I don't think it entitles you to practice any specific type of medicine. Most people leverage it to reach into the health tech space...but if that's your goal, there are easier ways than going through medicine.
Eh..... I don't know about that. If tech companies are bringing physician informaticists on board, it must be because the medical specialty is still needed. I've read on AI apps that aid in physician workload and I have a friend who is a software engineer for a dental imaging startup, and these companies need clinicians.

I read up on it more since I posted this, and one of the tenets of clinical informatics is bringing about organizational change. Using AI to streamline work definitely falls under it. I also contacted a new physician informaticist that work for these health tech startups and yeah, the work they're doing is bringing about that change.

I think it's more accurate to say clinical informatics help to bridge between medicine, tech/data science. Without it, physicians lack the necessary background to work and synchronize with these fields. Otherwise, why go through all the trouble of establishing an entire specialty? So I wouldn't say they leverage it just to work for health tech, but rather they bring something no other employee can bring that is needed to fulfill the aforementioned core value. The ones that work in health tech typically fulfill the role of CMO or Chief Informatics Officer.

You're right and wrong that it doesn't let you practice in any field; it's a fellowship, so you can't apply without finishing residency. It's meant to enhance your practice. Although it is a non-clinical specialty/training, fellows still practice in their residency field as part of their fellowship so their skills don't get rusty. To be board-certified in clinical informatics starting next year, you need to have done a fellowship. Ultimately, it is meant to enhance your practice in whatever you did a residency in if you go the clinical route.
 
Last edited:
Eh..... I don't know about that. If tech companies are bringing physician informaticists on board, it must be because the medical specialty is still needed. I've read on AI apps that aid in physician workload and I have a friend who is a software engineer for a dental imaging startup, and these companies need clinicians.

I read up on it more since I posted this, and one of the tenets of clinical informatics is bringing about organizational change. Using AI to streamline work definitely falls under it. I also contacted a new physician informaticist that work for these health tech startups and yeah, the work they're doing is bringing about that change.

I think it's more accurate to say clinical informatics help to bridge between medicine, tech/data science. Without it, physicians lack the necessary background to work and synchronize with these fields. Otherwise, why go through all the trouble of establishing an entire specialty? So I wouldn't say they leverage it just to work for health tech, but rather they bring something no other employee can bring that is needed to fulfill the aforementioned core value. The ones that work in health tech typically fulfill the role of CMO or Chief Informatics Officer.

You're right and wrong that it doesn't let you practice in any field; it's a fellowship, so you can't apply without finishing residency. It's meant to enhance your practice. Although it is a non-clinical specialty/training, fellows still practice in their residency field as part of their fellowship so their skills don't get rusty. To be board-certified in clinical informatics starting next year, you need to have done a fellowship. Ultimately, it is meant to enhance your practice in whatever you did a residency in if you go the clinical route.
OK... so... what do you think they do? I've never "informaticized" a patient. It really has very little tangible clinical value as a clinical specialty, because for all intents and purposes, they are not clinicians (at least if they're doing informatics work 100% of the time). The most clinical value I can think of is maybe as a clinical geneticist? Maybe pathology? I think it could really evolve into personal genomics/molecular medicine sort of practice...but even that feels far away. Ultimately, anything in informatics must first become a medical device/formal intervention, phased trials, etc etc—and by the time it does, any other physician can implement it, it's not like informatics has a corner on innovation. In fact, you might butt up against the translational science MD-PhDs, or even the implementation science grads out of schools like Dartmouth. There are lots of people trying to do new things in medicine.

I did a lot of informatics work over the years. I got a graduate certificate in AI at undergrad. Emotional modeling at NASA, machine learning/AI from scratch using Python in college, research in neuronal modeling using AI... It's all very interesting, but I don't see it as a destination, more so a vehicle to blend a clinical specialty with. I think LLMs will absolutely redefine the way psychiatry is practiced, and that's why I'm interested in the field. That's not a far away thing, the technology is already here and already available. If it weren't, maybe my perspective would be different.

To say you want to become a doctor to be an informaticist is like becoming a lawyer to do social work. It's all very impressive, but the job you're getting at the end is probably not worth what you had to do to get there. The juice isn't worth the squeeze, so to speak. But that's just my opinion. I'm happy that you found something in medicine to get excited about. Good for you.
 
OK... so... what do you think they do? I've never "informaticized" a patient. It really has very little tangible clinical value as a clinical specialty, because for all intents and purposes, they are not clinicians (at least if they're doing informatics work 100% of the time). The most clinical value I can think of is maybe as a clinical geneticist? Maybe pathology? I think it could really evolve into personal genomics/molecular medicine sort of practice...but even that feels far away. Ultimately, anything in informatics must first become a medical device/formal intervention, phased trials, etc etc—and by the time it does, any other physician can implement it, it's not like informatics has a corner on innovation. In fact, you might butt up against the translational science MD-PhDs, or even the implementation science grads out of schools like Dartmouth. There are lots of people trying to do new things in medicine.

I did a lot of informatics work over the years. I got a graduate certificate in AI at undergrad. Emotional modeling at NASA, machine learning/AI from scratch using Python in college, research in neuronal modeling using AI... It's all very interesting, but I don't see it as a destination, more so a vehicle to blend a clinical specialty with. I think LLMs will absolutely redefine the way psychiatry is practiced, and that's why I'm interested in the field. That's not a far away thing, the technology is already here and already available. If it weren't, maybe my perspective would be different.
At the patient care level, it helps to improve health equity. Some ethnicities of people are more susceptible to certain conditions than others. But medicine has historically been Eurocentric, so a condition that is taught to be rare is only rare among White people, but can be common in Asian people. We all know black people are more susceptible to vitamin D deficiency and anyone with a biology background can explain why. Clinical informatics helps you identify these trends. That's just one way I see clinical informatics enhancing patient care. So it's quite valuable if you plan to advocate for minority patients and health equity.

Another way I see it being valuable is helping to diagnose autism in girls. A key trait of autism if obsession with certain topics, but obsessing with celebrities is a common thing for teenage girls, so autism often gets masked by typical girl behavior. Clinical informatics can help with this issue too.

I'll ask some physician informaticists but something tells me they can give a thesis defense on why it has clinical value.

I think you are underestimating its value in patient care by a wide margin because again, if they went through the trouble of establishing it, it has to have value to patient care.

To say you want to become a doctor to be an informaticist is like becoming a lawyer to do social work. It's all very impressive, but the job you're getting at the end is probably not worth what you had to do to get there. The juice isn't worth the squeeze, so to speak. But that's just my opinion. I'm happy that you found something in medicine to get excited about. Good for you.
Again, it's not a standalone specialty; it's a fellowship. You can only apply for a clinical informatics fellowship after residency. As a clinician, it is meant to enhance your clinical practice.

At the end of the day, patient care is 90% pattern recognition; this is true whether you're a physician, PA, or NP. Clinical informatics helps you identify when a patient might be outside your usual pattern recognition standards.
 
Last edited:
At the patient care level, it helps to improve health equity. Some ethnicities of people are more susceptible to certain conditions than others. But medicine has historically been Eurocentric, so a condition that is taught to be rare is only rare among White people, but can be common in Asian people. We all know black people are more susceptible to vitamin D deficiency and anyone with a biology background can explain why. Clinical informatics helps you identify these trends. That's just one way I see clinical informatics enhancing patient care.
That's epidemiology/population medicine—you can do that in public health/preventive medicine. Those guys are also not generally patient-facing and tend to work overwhelmingly in academic medicine or policy.

You're not going to casually walk into work, enter the first exam room, hack your way into the CDC mainframe, dump some data into SPSS, visualize in Tableau, perform remote robot surgery across time zones on a break through your Vision Pro, and singlehandedly eradicate tuberculosis in Namibia by lunch.

That said, your points are valid, but any of those research foci could easily be inherited by just about any physician from the vantage point of their specialty... so I'm not really sure what case you're making about what informaticists actually do that is distinct from any other physician. Dermatologists can investigate complexion differences vs. vitamin D deficiency in alopecia areata; internists can investigate complexion differences vs. vitamin D deficiency in osteoporosis, or something. And even then, I don't see how the fellowship training would better prepare you to do that research.

Even if I gave you enough suspension of disbelief to agree informaticists have a corner on abstracting data. OK, and? What are you going to do, just generate problems you cannot resolve with that data ad infinitum? Are you really contributing to health equity, or just finding distinctions in health status across demographic lines?

I'm really not trying to come at you sideways here, I just think you're not really clear on what they actually do. But based on your interests, it sounds like you might really want to go for a dual MD/MS in artificial intelligence...a couple schools have them, you can write a thesis, invent something...the opportunities are endless.
 
Another way I see it being valuable is helping to diagnose autism in girls. A key trait of autism if obsession with certain topics, but obsessing with celebrities is a common thing for teenage girls, so autism often gets masked by typical girl behavior. Clinical informatics can help with this issue too.
No, that's genetics. I worked medical genetics and swabbed kids and their parents all the time. You put the swab in a tube and send it away; you get the results back a week or two later. No truth serums, data wizardry, or crystal balls necessary.

Crisis averted.
 
You brought up a good point. I got an answer back from a physician informaticist and they told me the biggest differences comes from one key point: access to PHI. Public health officials lack the free access to PHI physicians have, allowing them to design protocols to reduce readmissions and propose changes to patient care to improve outcome. That said, that does sound like any physician can theoretically do, so my guess is the fellowship better equips you for doing this kind of thing. That's just my guess. Anyway, good talk!

I'm not interested in what you said per se, this field just piqued my interest because I never heard of it before.
 
I briefly considered a fellowship in clinical informatics after completing my Internal Medicine and Genetics and Genomics training. However, I already completed a Masters in clinical research and the program I looked into was heavily focused on EHR-based efforts, which wasn't my cup of tea. Kudos to those interested in this training—I do think there's a specific need at IT-physician interfaces, it just wasn't for me.

No, that's genetics. I worked medical genetics and swabbed kids and their parents all the time. You put the swab in a tube and send it away; you get the results back a week or two later. No truth serums, data wizardry, or crystal balls necessary.

Crisis averted.
I think you're incorrectly assuming that because collecting samples is simple and reports are generated quickly, there isn't extensive 'data wizardry' involved in genetic testing. I can order a colonoscopy or an MRA, but I won't fool myself into thinking that means either procedure is simple to perform or interpret.

Genetic testing is exceptionally complex. In fact, there's so much to this that goes into testing that there's two separate fellowships that cover the process from start-to-finish. First is the clinical medical geneticist and team, but second and perhaps more interesting to the OP is the LGG-trained doctors who oversee, run, and interpret the genetic tests while creating the final reports

If you have interest in clinical informatics - why not at least explore Laboratory Genetics and Genomics (LGG) fellowship pathways. If you are interested in AI but perhaps less excited about EHR-based efforts, this might strike a balance for you. Directors of clinical genetics labs (hospital/private/for-profit companies) are now required to be LGG-fellowship trained. These are the 'data wizards' who are always behind-the-scenes. Training is open to MDs, DOs, and PhDs - it's a well-defined and pragmatic training program and could be of interest to you.
 
I briefly considered a fellowship in clinical informatics after completing my Internal Medicine and Genetics and Genomics training. However, I already completed a Masters in clinical research and the program I looked into was heavily focused on EHR-based efforts, which wasn't my cup of tea. Kudos to those interested in this training—I do think there's a specific need at IT-physician interfaces, it just wasn't for me.


I think you're incorrectly assuming that because collecting samples is simple and reports are generated quickly, there isn't extensive 'data wizardry' involved in genetic testing. I can order a colonoscopy or an MRA, but I won't fool myself into thinking that means either procedure is simple to perform or interpret.

Genetic testing is exceptionally complex. In fact, there's so much to this that goes into testing that there's two separate fellowships that cover the process from start-to-finish. First is the clinical medical geneticist and team, but second and perhaps more interesting to the OP is the LGG-trained doctors who oversee, run, and interpret the genetic tests while creating the final reports

If you have interest in clinical informatics - why not at least explore Laboratory Genetics and Genomics (LGG) fellowship pathways. If you are interested in AI but perhaps less excited about EHR-based efforts, this might strike a balance for you. Directors of clinical genetics labs (hospital/private/for-profit companies) are now required to be LGG-fellowship trained. These are the 'data wizards' who are always behind-the-scenes. Training is open to MDs, DOs, and PhDs - it's a well-defined and pragmatic training program and could be of interest to you.

My responses were geared toward OP's perception that a CI-trained physician could be occupying a hybrid patient-facing, bench, tech, and policy role—serving as a geneticist, informaticist, translational research scientist, implementation manager, and public health physician all at the same time, which is absurd. I hope my response is clear that there is no one specialty that supersedes all others and does not respect the clear boundaries between professionals, especially in a neoliberal bureaucracy like medicine where roles are defined increasingly conservatively.

The subtext, in my opinion, was a latent anxiety around the increasing specialization of physicians broadly, and a student's wish to be protected from an anticipated future where generalist care will become increasingly devalued. They are looking ahead in hopes of building a monopoly around their niche in practice.

Of course, in a specialty that is ostensibly not "medicine" the way we would normally describe it, that creates questions like "OK...what illness do you suspect you will cure with your superuser Epic skills?" More broadly, one might ask why someone would want to become a physician at all if their ideal job involves virtually all admin work and zero patient care. A lot of these fellowships are open to PhDs, which begs the question as to why someone interested in such a field would choose not to go down the path of least resistance through the academic route, which is a cheaper and more common vehicle to doing that work.

Despite the technical/practical answers to OP's questions, I think those fundamental existential questions undergirding OP's curiosity are actually really important for professionals to discuss. We are feeling pressure to look forward to subspecialization because job security/relative compensation really feels that fragile right now—within a profession that actually could not be any more essential to sustained human life and flourishing.

That is a real problem, and it's not just morale. It's all about the promise medicine makes to students: endure the brutal training, and you'll never have to endure again. We now live in a world where physicians regularly boo-hoo cry on the internet about how poor and depressed they are. And they're real people, not supermodels in scrubs. No longer haunted by Grey's Anatomy applicants, the profession broadly needs to combat the "medical school is a cult" ideology. I don't necessarily agree, but it's out there and people are listening.

We, as applicants, expect compelling reasons not to believe that...and lacking a response manifests in thousands of neurotic individuals hoping to be all things to all people to minimize the possibility of having a negative experience.

Anyway, I'm into psychiatry. :laugh:
 
Last edited:
There have always been folks who get into medicine for all the right reasons and discover that the things have changed. Perhaps it is burn-out, or a scientific breakthrough that has greatly reduced the need for your specialty (polio vaccines, TB and psych meds greatly changed the need for iron lungs, sanitariums, and frontal lobotomies), or a newly acquired disability that precludes continuing with your chosen field. In such cases, the knowledge, skills and abilities acquired in medical training can be applied to bioinformatics, or implementation science, or population health policy. In other cases, it could be the desire for a career in academic medicine that marries clinical care with bioinformatics /clinical informatics and the desire to get additional training to take on that secondary role.
 
Top