Wednesday, April 23, 2014

Will automation affect surgeons' skills?

Although it has been known for over two years, news outlets are again reporting that automation is degrading pilots' skills. Links are here and here. I blogged about this back then as part of a comparison of pilots to surgeons. My point was that surgeons did not have autopilots to rely on in the operating room.

This new report has prompted some to wonder whether robotic surgery will lead to deterioration of surgeons' skills.

In my opinion, that is not likely at this time because the robot is not really doing the surgery by itself. It is simply a tool that helps the surgeon and is under the surgeon's complete control at all times (except when it runs amok).

However, ever since the advent of laparoscopic surgery over 20 years ago and its popularity for many of the common procedures surgeons do, there has been concern that surgeons may eventually lose proficiency for open procedures. And a number of other open operations have been done less frequently due to alternate ways of treating patients such as non-operative or interventional radiologic techniques.

Here are some examples from the ACGME resident log data for the academic years 1999-2000 and 2011-2012.
We are approaching the critical lower limit for open gallbladder surgery expertise especially when you consider that only the most difficult cholecystectomies will be done as open cases from now on.

What will happen in 20 years when few surgeons will have sufficient skill to do a very inflamed open gallbladder?

Does anyone really believe that a surgeon can confidently remove an enlarged spleen having done fewer than 2 such cases during training?

This is a bigger problem and far more pressing than the possibility that automation will render human surgeons obsolete.

There's another issue too, which is the predicted shortage of general surgeons in the near future. How are more surgeons going to be trained if there are not enough open cases to train the current number of graduating residents, of which there were 1092 in 2012?

Has anyone else thought about these questions?


msquaredb said...

"This is a bigger problem and far more pressing than the possibility that automation will render human surgeons obsolete."

When do you anticipate this happening? At what point will robots be able to "think" through an entire surgery? Even if they can perform the correct steps during a simple case, what happens if a simple case goes bad? I have a hard time believing that a robot would be able to recognize issues deviating from normal and be able to respond and adapt quickly enough. I think significant technological leaps must be made before this is feasible.

However, I could see how surgeons could become nearly extinct before those advances are made. If robots can just get to the point where they can do the simple surgeries but not adjust to complications, then hospitals may find they only need to keep a few surgeons who are on standby to handle a= complications that the robot cant.

Skeptical Scalpel said...

MS, thanks for commenting. I don't see robots operating independently any time soon. I agree that a human surgeon would have to be immediately available in case of a complication or system crash.

artiger said...

I think about these questions often, and then thank God that I finished residency in 1997.

Bryan H. said...

Are 4 splenectomies really that much better than 2? I would expect surgeons to have done hundreds if not thousands of such cases during training. What are they doing for all those years instead?

Justin said...

I don't see this 1999-to-2011 comparison as a reflection of automation's effect on skills. Not to be dismissive, but shouldn't this comparison take into account the treatment histories for these "rare" operations? Trauma splenectomy is not considered necessary in as many cases as it was in 1999. While it's true that there will be less resident experience with these tough cases, won't this evolve like everything else into a subspecialty? Maybe a more pressing problem is whether 100% of general surgeons should be held to the same standards of autonomy in tough open choles that they are in bread-and-butter cases. If my fellowship director won't even think I can do a simple operation independently after I finish my 5 years, what makes anyone so surprised that I won't be independently capable of an operation that is necessary in less than 5% of the cases I'll see? I think it's pointless to complain that other management strategies are making us worse surgeons when they make our patients less likely to need the riskier treatment we once provided better. History and progress are forcing us trade in some of our skills for better patient care overall. In the meantime, maybe a relative few of us should actively maintain those skills while the rest of us work to keep ourselves relevant.

Anonymous said...

I am not a surgeon, but I think the specialty with the least worries about automation should be Surgery. There is not even a prototype for a robot lancing a boil yet.

Visual recognition without prior (human) sorting is a big hurdle. Surgery will be performed by humans for the foreseeable future.

A more likely scenario is that non-surgeons make the incisions, insert ports, and the robotic arms are controlled from a remote location by a surgeon.

Totally nonmedical, but here is a fascinating look how robots can now milk dairy cows with benefits for everyone involved - including the cows.

Vamsi Aribindi said...


How is there simultaneously a shortage of general surgeons, and yet not enough cases per surgical resident? From my understanding, surgical case volumes as a whole have been steadily declining, and may further decline as medical therapies solve problems previously solved via surgery. (as PPIs effectively ended vagotamies, who's to say that other procedures won't become obsolete?)

Have case volumes stabilized? Will we truly see a shortage of surgeons, or just a geographical maldistribution due to surgeon preferences in living?

Vamsi Aribindi

Skeptical Scalpel said...

Thanks for all the comments.

I wasn't trying to blame automation for the declining case numbers in residency training. I thought I said that rather than worrying about automation, we should figure out how to get more cases for residents to do.

Yes, the management of many problems has changed. That's a huge issue and it is responsible for much of the decline in numbers.

This is why I don't see how expanding existing surgical programs and creating new ones will do anything but worsen the problem.

As I have written elsewhere, It is time to start tracking people into subspecialties sooner. Every chest surgeon doesn't need to do 5 Whipples during residency.

artiger said...

How about shortening medical school and lengthening residency? Three years of school, a year or year and a half of internship, with selection of specialty near the end of internship?

Skeptical Scalpel said...

I don't see how med school can be shorter. There's too much to learn. I agree that choosing a specialty later would be a good idea, but it will never happen. The specialties have no incentive to change the system.

artiger said...

I'm not sure I agree (about not being able to shorten med school). A lot of year 1 and some of year 2 is stuff I haven't used since, well, year 1 and 2. How many of us have said something like "I didn't really learn anything useful until I got into residency".

My comment is partly tongue in cheek. Partly.

Anonymous said...

One issue with early sub-specialization is that with exception of cardiac, plastics, and maybe vascular, other subspecialists don't have enough business without also practicing general surgery (outside of academic centers).

Ksatriya821 said...

So, just so I understand this correctly- Skeptical Scalpel & co are lamenting the fact that today's surgery residents perform less of certain traditional open procedures, solely because of the fact those traditional open procedures aren't performed nearly as much anymore (dinosaurs do go extinct after all), and this is somewhat an indication that today's residents are less trained because they don't perform as many archaic procedures? Do you understand my confusion?

I'm a plastic surgery resident. I suppose I could lament the fact that my training will include absolutely zero delayed tubed flaps, or I could focus on the free flap techniques that replaced them and not care about what the previous generation thinks

Skeptical Scalpel said...

Anon, I agree.Some subspecialists done do enough procedures and have to fall back on GS.

Ksatriya821, you are making an incorrect assumption. Laparoscopic cholecystectomies do not eliminate the need for open cholecystectomies. Open choles are just not done as frequently, but they are done when the case simply can't be done laparoscopically. Open choles still need to be done and are often the most difficult due to inflammation. That makes them even harder to do.

Anonymous said...

What about bariatrics? I saw an item where several of them were shipping over to a couple countries in the Middle East because you couldn't get enough of them over there and the demand was outpacing the supply more so than America.

Skeptical Scalpel said...

I'm not certain that I understand your comment. Are you saying that bariatric surgeons are going to the Middle East?

It's not necessarily a benign environment in which to practice. Here's what happened to a South African MD who was tried for manslaughter over the death of a patient,

RuggerMD said...

Well, when all these new surgeons get into a bind or a complication they can always call me...the Acute Care Surgeon.
At least that is what is happening now.
Difference is we are making the hospital pay for our services in a big way. So far they are glad to have us do all the complex stuff. What I see happening though is the general surgeons case loads are now filled with hernias, gallbladders, breast and bariatric surgery.

Skeptical Scalpel said...

Rugger, you might be interested in reading a comment from April 27 on another post of mine.

Post a Comment