This was hashed out in the subspecialty encroachment thread but I’ll add my 2 cents here;
In your opinion, what inherently prevents neural models operating on raw input from capturing this type of interaction when the data is available?
Honestly, I think the biggest obstacle to an AI is in the information gathering step. There are so many aspects to human interaction that we subconsciously consider, I don't see AI taking it _all_ into account any time soon. I can't tell you how many times in clinic patients don't even know what to ask or bring up in terms of coordinating their own care and choosing between treatment options. I don't think a standardized questionaire, form or survey would be able to get at it all.
amazing article-I remember reading when it first was published. Anyway, I think it’s obvious one of the huge aspects of oncology is human interaction so to minimize that piece would be foolish
This is going down an incredibly stupid rabbit hole. Re-read back you yourself what you wrote there
And what @whoknows2012 is pointing out is that, the oncologist's value add is not just in the knowledge/expertise but ALSO in the psychosocial aspects of cancer, treatment and supportive care (what you so derisively dismiss as "hand holding"),
Hah-from someone with actual experience talking to the pcp of my patients, good luck with that one. Also this has become a mind numbingly stupid conversation. What’s your angle? That all types of medicine are apt for AI encroachment? Or that oncology in particular is. I think we can all agree as @bobsmith said above it’s hard to know what people will accept in 20 years, but like, is oncology MORE at risk than say GI, cards, rad onc etc etc??
GI and cards do procedures... so probably. Is onc more at risk than any other mostly cognitive specialty? Probably not.
Well, the "talking" isn't just reciting a list of side effects and saying something like, "OK see you in 3 months after you've finished all your adjuvant therapy," particularly since people react to therapies differently and adjustments need to be made.
vmghealth.com
I shouldn’t have said gi and cards, I meant other non procedural specialties, but then again does it matter? Procedural specialties will be undone by mid level encroachment
Like other fields as more practices do this their ability to leverage and strong arm their way in to a community and kill private practices will grow. Physicians need to stop selling to them and stop working for them but it won’t happen.It seems the the consensus is that AI isn't a threat from current oncologists. However, what about private equity? Do you find yourselves employed by private equity/VC groups?
Like has this consolidation affected you all?
Oncology on the Rise: Private Equity Investment in Cancer Care
Written by: Michael Miner and Vincent M. Kickirillo Dry powder in the private capital markets has been on the rise since 2012 and surpassed the $2 trillion mark for the first time in 2018.[1] With so much capital to deploy, and the rise of asset values since the Great Recession, private equity...
Do you foresee a future where you oversee like 20 NP's, pushed on you by PE groups, and just sign off on all their notes?
The "problem" is that PE is behind the curve on this. It's literally the only area in the business of medicine where academia has been on the forefront, primarily due to 340b pricing which is like manna to oncology groups. Between universities buying up community practices and slapping their name on the door, and things like McKesson buying USOncology, the margin for PE to make fat cash in oncology is pretty small.
I dont know they make it work on the inpatient side where all they get is E/M billing and nothing else. I can imagine leveraging size/midlevels into a lower outpatient overhead and turning that profitable as well to the detriment of patient care.
As a non oncologist looking from the outside...my 2 cents, I think the getting "systemic therapy" part is the biggest piece to protecting your field as it currently stands.
Sure, I agree that if there is an easy cure for all cancers, then oncologists will no longer be needed =D
Where does this ad nauseam paranoia about MLP and AI encroachment come from ?
Its not happening now, if you think otherwise then you need to lay off Elon Musk's Kool-Aid. If it happens in 30yrs then so what? People are ingenious and make themselves relevant one way or the other.
Those interested in the dynamics of the labour market and how the threat of automation never realized in developed world job market, please read the Economist last week print.
Riding high in a workers’ world
A jobs rebound, shifting politics and technological change could bring a golden age for labour in rich countrieswww.economist.com
I didn't say they couldn't do it. Just that they're late to the party.
Great article.
www.upi.com
...is this trolling?Therapeutic cancer vaccine shows promise against multiple tumor types - UPI.com
A vaccine that uses genetics to teach a person's immune system how to precisely target the cancer has proven safe and feasible in early clinical trials, according to results of a new study.
Not at all. Just incase, as an oncologist, you're not aware that cancer vaccines are a thing. Pun intended.
...and poor reading comprehension.
I'm not sure how a new treatment option (note that the "vaccine" isn't for preventing the development of cancer, but treating an established one...basically CAR-T therapy) is a threat to oncologists.
