Least algorithmic subspecialty?

[jbing222]

---

Members don't see this ad.

I would be really interested to hear people's thoughts on which subspecialty in internal medicine is the least algorithmic in nature. Although I understand the importance of algorithms in patient care, during the course of my training I have found implementing these algorithms to be the least satisfying aspect of medicine. I enjoy critical thinking, navigating shades of grey (nothing to do with that book!), and outside the box thinking...I guess some would call it the "art" side of medicine.

Any thoughts on which subspecialities would fit that description best? Thanks!

 

[gutonc]

I would be really interested to hear people's thoughts on which subspecialty in internal medicine is the least algorithmic in nature. Although I understand the importance of algorithms in patient care, during the course of my training I have found implementing these algorithms to be the least satisfying aspect of medicine. I enjoy critical thinking, navigating shades of grey (nothing to do with that book!), and outside the box thinking...I guess some would call it the "art" side of medicine.

Any thoughts on which subspecialities would fit that description best? Thanks!

Cash only concierge medicine. Seriously.

Look, we all want to think outside the box, move in and out of the known and unknown, think critically and practice the art of medicine.

You know what else we all like to do? Get paid and not get sued. And these days, that means following guidelines, algorithms and other decision making tools. This is sort of a good thing because (most...obviously not all) of these guidelines and algorithms are data-driven.

That said, I think every subspecialty (including primary care) has plenty of opportunities to do what you're talking about. The thing is, you have to do the "standard of care" first and if/when that doesn't work, you get to go off the reservation and "think critically" or "be a cowboy." The reason you haven't experienced this part of it much to this stage in your training is because you have to learn to walk before you can fly. As a med student (and frankly as a resident and a fellow too), your job is to learn the "right way" to do things. Once you've mastered that, then you can go crazy. I'm a first year attending and, while I have lots of ideas about ways that I might be able to treat my patients, once they go off the guidelines, even when I have data on my side, it scares the hell out of me.

Patience grasshopper.

 

[jbing222]

Thanks for the response, and I definitely get what you're saying about the importance of algorithms in medicine. I guess what I'm asking then is if there is a field that has fewer or maybe less well defined algorithms from which managment decisions are made? The only simplistic example I can give is in infectious diseases with antibiotic selection - there a good handful of times it seems where there is not one set regimen that is indicated for a given patient. I've seen more than one attending use their own gestalt in these cases as opposed to some pre-defined guideline.

 

[gutonc]

Thanks for the response, and I definitely get what you're saying about the importance of algorithms in medicine. I guess what I'm asking then is if there is a field that has fewer or maybe less well defined algorithms from which managment decisions are made? The only simplistic example I can give is in infectious diseases with antibiotic selection - there a good handful of times it seems where there is not one set regimen that is indicated for a given patient. I've seen more than one attending use their own gestalt in these cases as opposed to some pre-defined guideline.

Yes. This is the case in virtually every specialty. If the guidelines say "give one of these 5 things for that problem" you can choose the one you like.

As an example in my specialty, adjuvant chemotherapy and endocrine therapy for ER+ breast cancer. I have data that say there are 4 endocrine drugs that work and at least 4 different chemo regimens with generally equivalent efficacy. The art and judgement comes in deciding which to use when.

In the case of endocrine therapy, I give them the one their insurance pays for. There's economics at work too.

 

[Bostonredsox]

Yes. This is the case in virtually every specialty. If the guidelines say "give one of these 5 things for that problem" you can choose the one you like.

As an example in my specialty, adjuvant chemotherapy and endocrine therapy for ER+ breast cancer. I have data that say there are 4 endocrine drugs that work and at least 4 different chemo regimens with generally equivalent efficacy. The art and judgement comes in deciding which to use when.

In the case of endocrine therapy, I give them the one their insurance pays for. There's economics at work too.

Isnt the NNT for tamoxifen in those pts like 30? Are the numbers for arimidex or any of the other options any more effective?

 

[gutonc]

Isnt the NNT for tamoxifen in those pts like 30? Are the numbers for arimidex or any of the other options any more effective?

That's for primary prevention or in the setting of DCIS/LCIS. That's not what I'm talking about.

For actual adjuvant therapy the NNT is more in the 3-8 range (depends on the study).

 

[Bostonredsox]

That's for primary prevention or in the setting of DCIS/LCIS. That's not what I'm talking about.

For actual adjuvant therapy the NNT is more in the 3-8 range (depends on the study).

ah ok. I remember reading a paper not too long ago on age >70 females with Breast ca, i think it was prevention of contralateral ca or ipsilateral recurrence, can't remember with, s/p lumpectomy + Rad +/- tamoxifen and the NNT was ridiculously high.

 

[gutonc]

ah ok. I remember reading a paper not too long ago on age >70 females with Breast ca, i think it was prevention of contralateral ca or ipsilateral recurrence, can't remember with, s/p lumpectomy + Rad +/- tamoxifen and the NNT was ridiculously high.

Well now you're getting way off track. And irrelevant to the topic at hand.

 

[pntgrd]

Rhuem. Critical care also isn't so data driven.

 

[PlutoBoy]

Rhuem. Critical care also isn't so data driven.

Rheum was the one I had in mind.

There's a fair amount of algorithms in CC. Early goal directed therapy, DKA, cardiac arrest, ARDS...

 

[Bostonredsox]

Well now you're getting way off track. And irrelevant to the topic at hand.

haha yeah I know I am just bustin your balls.

 

[Bostonredsox]

Rhuem. Critical care also isn't so data driven.

Critical care is substantially algorithm driven.

Rheum would be my choice as well.

 

[gutonc]

Critical care is substantially algorithm driven.

Rheum would be my choice as well.

Rheum isn't algorithm driven because the answer is always "steroids."

 

[PlutoBoy]

Rheum isn't algorithm driven because the answer is always "steroids."

Sometimes is biologicals 😛

 

[Bostonredsox]

Rheum isn't algorithm driven because the answer is always "steroids."

lol. love it. "well its not lupus, but lets try prednisone."

same with derm.... "well I am not sure exactly what type of rash this is, try triamcinolone."

 

[redy]

lol. love it. "well its not lupus, but lets try prednisone."

same with derm.... "well I am not sure exactly what type of rash this is, try triamcinolone."

Unfortunately its true mostly of pulmonary too.

 

[Bostonredsox]

Unfortunately its true mostly of pulmonary too.

yes but pulm slots are mostly pulm/cc slots, and the CC side is as I mentioned, very algorithm driven.