What specialties can be outsourced or mostly automated requiring less number of doctors in next 15-20-25 years?
e.g. Radiology?? Psychiatry? or any other?
e.g. Radiology?? Psychiatry? or any other?
What specialties can be outsourced or mostly automated requiring less number of doctors in next 15-20-25 years?
- Thread starter khichadi
- Start date
- Joined
- Dec 23, 2015
- Messages
- 1,071
- Reaction score
- 2,169
....psychiatry????
- Joined
- Sep 17, 2019
- Messages
- 219
- Reaction score
- 916
Yeah haven't you heard? They just pushed out Lithium vending machines in all major cities.
- Joined
- Dec 4, 2016
- Messages
- 366
- Reaction score
- 641
I second this. Loss of practice territory to other physicians (see the turf war for valve replacement in cards vs ct surg), to mid levels (see FM/EM), etc.
Sure, one can imagine a world where competent AI and robots replace even the most specialized proceduralists. However, getting to that point requires a metric boatload of time, energy, and money. Odds are good that you will be nearing the end of your career, if not literally deceased, by the time this becomes a real threat.
Honestly it’s hard for me to imagine any world in which AI completely replaces any specialist, even imaging-based fields like radiology. The most competent AI imaginable can still get bugs or break down, and the only people who will be trained enough to know when that happens and how to fix it will be medical specialists. At worst physicians may move to a supervising role but I don’t see how AI can exclude human involvement entirely.
- Joined
- Jul 23, 2018
- Messages
- 3,179
- Reaction score
- 6,254
Well I don’t see ER being AI.... mid levels - maybe . But not AI. Radiology , pathology - I see those
- Joined
- Jun 18, 2015
- Messages
- 391
- Reaction score
- 845
The only place AI will replace entire specialties is in the minds of sales reps and hospital administrators.
Last edited:
- Joined
- Dec 28, 2010
- Messages
- 5,050
- Reaction score
- 6,082
Outsource/AI replacement is something is a common pre-med/start of medical school concern. They view medicine as regurgitation of facts because that's what they are exposed to. Couple that with neuroticism and you get a bunch of threads about machines replacing doctors. Once you start clinical medicine, you recognize how unrealistic this is.
In terms of mid levels, I don't see them as an existential threat to any field. The world is going to get more litigatory and Americans always best and hence want a doctor, not an NP/PA.
In terms of mid levels, I don't see them as an existential threat to any field. The world is going to get more litigatory and Americans always best and hence want a doctor, not an NP/PA.
Last edited:
If midlevels can regurg facts and get independent practice, why can’t AI? I’d trust AI more than midlevels to practice independently tbh
- Joined
- Dec 28, 2010
- Messages
- 5,050
- Reaction score
- 6,082
Because AI can't walk, talk, and get sued.
On a more serious note, medicine is never black/white. The condition of a patient's heart can not be reduced to some printed vector lines representing current. An EKG is just one data point physicians use in their assessment. When you go into the room, what does the patient look like, how do they feel, how did your conversation go. The ability for humans to assess that amongst each other is what we're evolved to do.
When a patient gets discharged, the discharge summary is forced to include a set of discrete conditions the patient had, but those are in essence a set of best guesses based on clinical reasoning. That would screw AI up. AI is dependent on accurate/discrete input. You could also suggest that with AI you could simply feed it limitless variables and let the AI machine churn (ruining everyone's clinical research projects). Even with the rules, there would be no way anyone would let AI make medical decisions as humans have infinite idiosyncrasies that will throw the AI off. I'm not saying that AI won't become more useful, but it's going to be utilized by doctors.
---EDIT
Regarding the mid level thing, you can downvote me as much as you want but I don't think the NP/PA thing is an issue SDN, Reddit makes it seem to be. Tons of posts are just obnoxious. Additionally, medical students perceive PA/NPs to be more autonomous than they truly are with their limited perspective. Add that to the fact that they're constantly bumping elbows with them on rotations to make the sentiment worse and I'm not surprised why most have a negative sentiment. Finally, there are a TON of crazy NP/PAs who attempt autonomous, illegal practice and there are serious issues regarding their training so there is plenty of anecdotes to fuel the fire. At the end of the day, there is labor to be done that physicians are not doing (rounding, perioperative management). Physicians have grown to rely on these midlevels to perform these tasks for them. I don't see that as an existential threat to healthcare either but I do agree the field needs to keep a close eye on overstep.
Last edited:
- Joined
- Dec 7, 2016
- Messages
- 1,247
- Reaction score
- 2,038
Sure hasn't slowed things down too much so far...
I say we automate the patients too. Start simple by automating diabetes and hypertension, but eventually work towards automating cancer and even schizophrenia as the technology advances. No more no shows and medication non-compliance. These human patients are too inefficient!
Last edited:
- Joined
- Nov 18, 2016
- Messages
- 271
- Reaction score
- 681
Eric Topol (a top ten most cited physician) wrote a book called "Deep Medicine" that talks about this. To use radiology as an example, AI will be a disruptive force, but is it a replacing force? Not necessarily. Just as how PACS, MRI, and teleradiology were labeled as "radiology-destroyers" very much like AI is today, introduction of these technologies didn't diminish the need of the radiologist, it merely altered and improved the roles of the radiologists.
First let's talk about why it's hard to automate radiology, the field unanimously declared to be the "first one to go" in the era of AI:
- A person receives healthcare from so many different providers over a lifetime. AI works best if data is comprehensive and complete, it's really hard to get complete data on a single patient because of this "fractured" system of US healthcare.
- Currently it's still hard to have AI that works inter-institutionally. Humans are able to use one example to extrapolate to multiple other similar instances. While AI works best when it uses multiple similar examples to extrapolate the results of one scan. This means that at its core, AI is brittle. If it sees new data presented in a different way from a new institution, it gets confused and spits out the wrong diagnosis.
- Radiologist when looking at a scan will look (unconsciously) for various abnormalities like rib fractures, Ca deposits, heart enlargement, fluid collections. Currently the advancement of clinical AI has been narrow, only ruling in or ruling out one disease. However, interestingly, Stanford's ChexNext AI recently reported being able to detect the 14 most common pathologies on chest X rays. But still, this is a baby step to creating an AI that completely replaces the radiologist.
- George Moore, Google VP of Healthcare, "There literally have to be thousands of algorithms to even come close to replicating what a radiologist can do on a given day"
So what will radiology look like in the future?
- Triage of important cases by AI. Urgent cases can be read faster by radiologist which improves patient outcomes via early intervention
- Serve as "gatekeepers" who review the necessity of ordered scans before asking the tech to perform the scan. Unnecessary scans are currently a huge financial waste in the US. The tradition of "passing off" a patient to the radiologist is costing too much money. Not only for financial reasons, but having too many unnecessary tests bring up a high probability of false positives. Take prostate cancer. For every 1000 patients scanned, you will save ~1 from prostate cancer related death. HOWEVER, a man is 40-80 times more likely to get a false positive and get unnecessary radiation therapy, than he is being saved from the detection of a true positive prostate cancer (citation below). Moreover, AI can reduce the chances of false positives, again saving more money and saving the patient from unnecessary stress. Furthermore, this is an opportunity for dialogue and teaching from the radiologist to the EM doc/PCP/etc. This luxury is only available when the repetitive and boring parts of radiology become automated with AI.
- AI is still able to maintain high accuracy with low res images. This means that patients only now have to sit for 10 minutes instead of 60 for their MRI. This also means we can reduce the radiation dose for CT scans. And lower costs too!
- More patient contact. Like the subspecialty of breast imaging, there is a lot of patient contact. And that usually provides unique insights not obvious when you are just staring at a scan. Moreover, you can discuss results with the patient directly, especially if you can explain how AI contributed to the healthcare of the patient (as some patients might be anxious having their care done by AI). The future radiologist must be competent in the fundamentals of AI, particularly in the skill of appraising a model and identifying its nuanced strengths and weaknesses.
So I think to some extent radiology will get automated, but this doesn't mean that the job of the radiology is under attack. It's just going to change and improve over time, just as it has always been changing and improving. It'll make the life of a radiologist much easier!
Pinsky, P. F., Prorok, P. C., & Kramer, B. S. (2017). Prostate Cancer Screening-A Perspective on the Current State of the Evidence. The New England journal of medicine, 376(13), 1285-1289.
Edit: The specialties of dermatology, ophthalmology, pathology, oncology, cardiology, psychiatry, and even surgery are specialties that see the most growth in AI, and thus to a large extent, automation. But again, it's just all variations on the same theme. Not replaced, but improved.
First let's talk about why it's hard to automate radiology, the field unanimously declared to be the "first one to go" in the era of AI:
- A person receives healthcare from so many different providers over a lifetime. AI works best if data is comprehensive and complete, it's really hard to get complete data on a single patient because of this "fractured" system of US healthcare.
- Currently it's still hard to have AI that works inter-institutionally. Humans are able to use one example to extrapolate to multiple other similar instances. While AI works best when it uses multiple similar examples to extrapolate the results of one scan. This means that at its core, AI is brittle. If it sees new data presented in a different way from a new institution, it gets confused and spits out the wrong diagnosis.
- Radiologist when looking at a scan will look (unconsciously) for various abnormalities like rib fractures, Ca deposits, heart enlargement, fluid collections. Currently the advancement of clinical AI has been narrow, only ruling in or ruling out one disease. However, interestingly, Stanford's ChexNext AI recently reported being able to detect the 14 most common pathologies on chest X rays. But still, this is a baby step to creating an AI that completely replaces the radiologist.
- George Moore, Google VP of Healthcare, "There literally have to be thousands of algorithms to even come close to replicating what a radiologist can do on a given day"
So what will radiology look like in the future?
- Triage of important cases by AI. Urgent cases can be read faster by radiologist which improves patient outcomes via early intervention
- Serve as "gatekeepers" who review the necessity of ordered scans before asking the tech to perform the scan. Unnecessary scans are currently a huge financial waste in the US. The tradition of "passing off" a patient to the radiologist is costing too much money. Not only for financial reasons, but having too many unnecessary tests bring up a high probability of false positives. Take prostate cancer. For every 1000 patients scanned, you will save ~1 from prostate cancer related death. HOWEVER, a man is 40-80 times more likely to get a false positive and get unnecessary radiation therapy, than he is being saved from the detection of a true positive prostate cancer (citation below). Moreover, AI can reduce the chances of false positives, again saving more money and saving the patient from unnecessary stress. Furthermore, this is an opportunity for dialogue and teaching from the radiologist to the EM doc/PCP/etc. This luxury is only available when the repetitive and boring parts of radiology become automated with AI.
- AI is still able to maintain high accuracy with low res images. This means that patients only now have to sit for 10 minutes instead of 60 for their MRI. This also means we can reduce the radiation dose for CT scans. And lower costs too!
- More patient contact. Like the subspecialty of breast imaging, there is a lot of patient contact. And that usually provides unique insights not obvious when you are just staring at a scan. Moreover, you can discuss results with the patient directly, especially if you can explain how AI contributed to the healthcare of the patient (as some patients might be anxious having their care done by AI). The future radiologist must be competent in the fundamentals of AI, particularly in the skill of appraising a model and identifying its nuanced strengths and weaknesses.
So I think to some extent radiology will get automated, but this doesn't mean that the job of the radiology is under attack. It's just going to change and improve over time, just as it has always been changing and improving. It'll make the life of a radiologist much easier!
Pinsky, P. F., Prorok, P. C., & Kramer, B. S. (2017). Prostate Cancer Screening-A Perspective on the Current State of the Evidence. The New England journal of medicine, 376(13), 1285-1289.
Edit: The specialties of dermatology, ophthalmology, pathology, oncology, cardiology, psychiatry, and even surgery are specialties that see the most growth in AI, and thus to a large extent, automation. But again, it's just all variations on the same theme. Not replaced, but improved.
Last edited:
- Joined
- May 13, 2016
- Messages
- 1,666
- Reaction score
- 2,593
This has become very apparent with each note I write as a M3
- Joined
- Dec 7, 2016
- Messages
- 769
- Reaction score
- 863
primary care specialties will be worse for physicians, because midlevels armed with AI will become pretty acceptable for Americans in a system that costs far too much for the shrinking middle class. And for the destitute... well, midlevels also pass on lower costs to taxpayers.
I'd also be nervous about pathology and *maybe* radiology - not because AI will obviate the need for physicians, but because even a moderate reduction in demand could lead to very significant market impacts in these fields. Probably safer for those who remain in academic medicine.
I'd also be nervous about pathology and *maybe* radiology - not because AI will obviate the need for physicians, but because even a moderate reduction in demand could lead to very significant market impacts in these fields. Probably safer for those who remain in academic medicine.
Going back to the original formula of 7 Up, I see.
Not today, but what about in 15-20-25 years? Lot changed in last 25 years.
true, but hopefully I will have saved enough to retire by that time. I plan on having a FIRE mentality early career.
- Joined
- Dec 28, 2010
- Messages
- 5,050
- Reaction score
- 6,082
When you graduate (or maybe even now), there are tons of Facebook groups you'll be able to join. Thousands of physicians share that sentiment.
- Joined
- Jul 14, 2018
- Messages
- 237
- Reaction score
- 269
People hate speaking to automated phone operators-- anything with face-to-face value isn't going away unless people entirely stop existing as physical entities and instead transition into the virtual ether; which again, people hate speaking to automated phone operators. The potential integration of AI is overhyped and has a hard stop at the hard problem of consciousness.
- Joined
- Oct 27, 2015
- Messages
- 745
- Reaction score
- 862
There is an artificially low number of physicians in this country due to AMA lobbying. It doesn’t do our patients or our lifestyles any favors. Does help the wallet though (questionably).
Last edited:
Im sure your lifestyle is going to be great when you finish residency and can no longer find a job at a salary equivalent to current incomes.
- Joined
- Nov 18, 2016
- Messages
- 271
- Reaction score
- 681
I thought the same until I looked more into this topic. Here's something interesting, in psychiatry, chatbots like Wysa, Lantern, Joyable, Woebot, and X2AI which mimicked cognitive behavioral therapy (CBT) showed equal or even greater performance than traditional CBT with a psychiatrist. Though it won't replace psychiatrists, it can be a good transitionary alternative for a significant number of underserved individuals (up to 85% don't get psych care in low/middle income countries). Though there are still advancements to be made, I wouldn't be so confident in writing off AI solely because people crave human encounters. AI can help with particularly the self management of psychiatric illnesses. Though I do agree, the full integration of AI to replace doctors is overhyped.
Firth, J., Torous, J., Nicholas, J., Carney, R., Pratap, A., Rosenbaum, S., & Sarris, J. (2017). The efficacy of smartphone‐based mental health interventions for depressive symptoms: a meta‐analysis of randomized controlled trials. World Psychiatry, 16(3), 287-298.
- Joined
- Oct 27, 2015
- Messages
- 745
- Reaction score
- 862
Honestly I do not understand this. I’d happily take a more normal salary for highly educated, technical professions like 150k (rather than 300+) if it means the system works better as a whole and if I don’t have to work as hard. Debt sucks but PART of the high cost of healthcare and justification for high cost of med school is how much physicians are paid, so you’d expect that debt load would be lower too.
- Joined
- Aug 4, 2014
- Messages
- 5,974
- Reaction score
- 12,068
What percentage of healthcare costs are the physicians salary?
Edit: also, the high cost of medical school is not due to justification based on future earnings, it’s because it costs money to train physicians (and to some degree likely due to administrative costs). You can look at historical cost of medical school in the US vs physician income and inflation. Physician income has always been high, especially relative to the general population, while the cost for medical school has not been so.
Why Has the Cost of Medical School Skyrocketed?
www.savvypremed.com
How Much Does It Actually Cost to Go to Medical School?
It’s no secret medical school is getting more expensive. So how are the future doctors of covering this expanding price tag? Here’s a look.
www.sofi.com
Last edited:
- Joined
- Aug 7, 2019
- Messages
- 608
- Reaction score
- 910
The cost of healthcare in the US attributed to physician income is something like 8% so I dont think that's an accurate statement.
Because engineers can make 100k out of college with the right background. By going your route I will have wasted 10 years of my life in medical training just to make 50k more in salary than a college graduate
Not to mention the fact that Congress "lowering debt" will affect everyone, not just doctors. So effectively I could have been an engineer with zero debt and 100k salary at 22 years of age versus a doctor with 150k salary at 35 and 200k debt instead of 250k debt.
When you run the numbers its crazy how bad a deal becoming a doctor will be if the far left gets their plans enacted.
- Joined
- Jul 11, 2013
- Messages
- 2,043
- Reaction score
- 4,741
Meanwhile, administrative bloat, hospital systems, and insurance companies are 90%+. What's the easiest thing to get rid of in this list, hmmm???
- Joined
- Oct 27, 2015
- Messages
- 745
- Reaction score
- 862
Ok the fact that wiping out just $50k of debt is being represented as the “Far left” position says a lot. Outside of pay, there’s a lot of other career perks to being a physician (best job security of any job ever, high autonomy, direct interaction with people who will benefit from your work, more respect than any other profession) so I don’t think that’s a fair comparison at all.
- Joined
- Sep 17, 2019
- Messages
- 219
- Reaction score
- 916
Yeah, the statistics for the increase in the number of physicians compared to the increase in hospital administrators are crazy... yet somehow the docs are the problem?
- Joined
- Oct 27, 2015
- Messages
- 745
- Reaction score
- 862
What percentage of healthcare costs are the physicians salary?
Edit: also, the high cost of medical school is not due to justification based on future earnings, it’s because it costs money to train physicians (and to some degree likely due to administrative costs). You can look at historical cost of medical school in the US vs physician income and inflation. Physician income has always been high, especially relative to the general population, while the cost for medical school has not been so.
Why Has the Cost of Medical School Skyrocketed?
How Much Does It Actually Cost to Go to Medical School?
It’s no secret medical school is getting more expensive. So how are the future doctors of covering this expanding price tag? Here’s a look.www.sofi.com
Agree with all this. Physician pay is only part of the equation. Much bigger fish to fry. That said, I have dug into how my school determines tuition price and it has NOTHING to do with the cost of training a physician. It is a simple formula: 110% of average tuition from an arbitrary ten member list of our peer schools.
- Joined
- Dec 23, 2015
- Messages
- 1,071
- Reaction score
- 2,169
These are sadly debatable.
- Joined
- Jun 17, 2014
- Messages
- 63,099
- Reaction score
- 154,744
SDN has way too much obsession with super high salaries (on top of the usual top tier addiction).
1. Job security will decrease as we increase residency slots. See rad onc
2. High autonomy - what? I don't think even doctors believe they have autonomy.
3. direct interaction with people - could be a positive or negative based on the person
4. respect - lol. The public sentiment regarding physicians is not positive my friend.
All in all, I would trade 2, 3, and 4 for a higher salary any day of the week.
- Joined
- Jul 11, 2013
- Messages
- 2,043
- Reaction score
- 4,741
I agree with all of your trades except for giving away our autonomy. That's just no, lol. That's part of what's killing the profession. Do-nothing admins turning this into a customer service field instead of valuing patient safety.
- Joined
- Jul 21, 2014
- Messages
- 1,318
- Reaction score
- 2,780
On the real though, fast food workers, truck drivers, a lot of modern assembly line jobs, and a dozen other easier targets still haven't even been automated. When/if we reach that point MAYBE I'll start to worry,
- Joined
- Dec 28, 2010
- Messages
- 5,050
- Reaction score
- 6,082
Sidestepping the political implications...
A talented engineer who finds his/her niche early in their career, moves up quickly, and makes financially savvy decisions will likely accumulate a comparable if not larger sum of money than a physician by retirement with a higher quality of life. It's not the rare 1/million people suggest.
There's unfortunately nothing that says we as physicians are entitled to more because we sacrificed our 20s. Going into medicine, I knew there were other fields my friends were doing that may offer a better lifestyle/higher overall earnings, but I honestly had no interest in them, liked biology, and wanted to do what the people I shadowed did.
The good news for physicians is that I imagine we all choose medicine because we enjoyed providing to do something challenging to help vulnerable people and as the pathway moves along, you'll realize you do make differences and there will be several ways to earn money within and outside medicine.
- Joined
- Dec 4, 2016
- Messages
- 366
- Reaction score
- 641
The bolded, louder for the pre-meds in the back.
- Joined
- Jul 14, 2018
- Messages
- 237
- Reaction score
- 269
Big here. We are privileged to spend our 20s learning about the human body and healing. This shift in thought and self-conceptualizing of medicine would help greatly with general medical student mental health. Always thought it was weird to me to view this as a "sacrifice" but that's me, I recognize my bias.
- Joined
- Jul 11, 2013
- Messages
- 2,043
- Reaction score
- 4,741
Not gonna lie, I never understood the
"sacrificing our 20s" thing. I just can't relate. Like what the heck else would I be doing? I don't have a family to take care of or anything. I would most likely not be working in a similarly fulfilling (and well paying) career.
It's a blessing to be doing what you feel like you were born to do. But I guess, if you hate med school, then it'll feel like a sacrifice or something akin to that.
Let me reiterate though, I am not starry eyed by any stretch of the imagination. I hate the nonsensical BS that we have to deal with tremendously, but I enjoy the actual medicine aspect of it.
- Joined
- May 13, 2016
- Messages
- 1,666
- Reaction score
- 2,593
honestly we are privileged also in the sense we are given 20-25K a year to live off while we study. Sure we have to pay it back plus interest but I've never been financially worried while in med school and I'm 100% on loans and live alone.
....you are financially blessed to get a loan at 7-10% interest? What am I reading in this thread.
- Joined
- Sep 17, 2019
- Messages
- 219
- Reaction score
- 916
It's considered a sacrifice due to the opportunity costs involved with medical education. Getting the wonderful privilege of medical education comes at the expense of other benefits of working in the 20s. Even if you decided to be a couch potato instead of going to med school, there is an opportunity cost involved. Going to med school now means you don't have the opportunity to be a couch potato. Thus, this would be defined as something you "sacrifice" by going to med school. When people are saying med school is a sacrifice, they are referring to this principle of opportunity cost, and there are dozens of more substantial opportunities lost than not being able to be a couch potato. There is sacrifice in every decision we make, and our decisions are influenced by whether those sacrifices are worth the benefits.
Anyone who believes there is no sacrifice in attending medical school are either zealously blinded by idealism or have no real-life experience
Anyone who believes there is no sacrifice in attending medical school are either zealously blinded by idealism or have no real-life experience
- Joined
- Jul 11, 2013
- Messages
- 2,043
- Reaction score
- 4,741
Do you mind giving examples? I'm honestly curious. Maybe I'm missing something.
Compare the earnings of a primary care doctor under the AOC/far left proposals versus a UPS driver who is aggressively saving in the SP500 or AAPL over the last decade. Then you will have your answer.
- Joined
- Jul 11, 2013
- Messages
- 2,043
- Reaction score
- 4,741
Okay, but both of these scenarios are dealing in extremes. I'm talking about the average case.
- Joined
- Sep 16, 2012
- Messages
- 2,365
- Reaction score
- 3,617
I agree with slowthai the “sacrifice” concept makes no sense. If so, my brother is “sacrificing” his 20’s to get a PhD. My sister sacrificed her 20’s to get teaching credentials and do her internships/apprenticeships/whatever they’re called. Had to bounce between multiple temporary jobs. As a non-read I had a number of jobs prior to med school as well. I had three one one time at one point.
Why is it that we physicians sacrifice our 20’s but no one else does? We’re not the only ones studying our butts off. We’re not the only ones working long hours. Is it a longer path than most? Sort of... 4 years of med school plus 3-5 on average of residency—they’re we’re a fully paid attending immediately. How long do you think an architect has to go to school and then work as an apprentice before they can be the principle of their own firm and be responsible for all the big decisions? Quite a while, unless all they want to do is design garden sheds. Same for so many other professions out there.
If anyone gets to play the sacrifice card, it’s deployed soldiers. They sacrifice. I didn't--I just had to show up to school for another four years, see my friends daily, learn some cool stuff, then do an apprenticeship where others hold my hand most of the time. Most professions don't have a Match process, nor do they have the checks/balances to prevent abuse like we do. New architect apprentice at a big firm? You're not sleeping at all the weeks prior to a big project deadline, and I'd wager most weeks you're well above the 40hrs/week we think everyone who isn't a doctor works. I'm guessing 60-70hrs+/week if you're really trying to get ahead. And then after everything is done, maybe you're looking at a salary of $50-80k as a junior architect.
- Joined
- Sep 17, 2019
- Messages
- 219
- Reaction score
- 916
No one has ever said that other professions don't make sacrifices. I think some of you have a different definition of sacrifice than what I'm referring to. Let's say we sum up a list of positives and negatives involved with going to medical school. You put everything you gain into the pros column (education, satisfaction, career prospects, etc), and you put everything you lose out on in the cons column (going into debt, "losing" your 20s, missing out on sleep, etc.) What I am referring to as sacrifice is anything you want to put into the cons list. Even though the pros outweigh the cons for most of us, there are still things we miss out on. I am guessing when most people talk about the sacrifice of med school, they are referring to what is in the cons list for them personally.
