Study Finds Robot Surgeons Are Actually Slower and More Expensive (theregister.co.uk)

"Robot-assisted surgery costs more time and money than traditional methods, but isn't more effective, for certain types of operations," reports the Register, in an article shared by schwit1: In a study of almost 24,000 laparoscopic surgeries just published in The Journal of American Medicine, researchers from Stanford University School of Medicine analyzed data from 416 hospitals around the U.S. from 2003 to 2015. Robotic assistance provides 3D-visualization, a broader range of motion for instruments, and better ergonomics for physicians, according to the study. While it has advantages in scenarios where a high-degree of precision is required or where improved outcomes have been demonstrated (like radical prostatectomy), it appears to be a waste of resources for the two operations examined... But the patient outcomes were more or less the same.

A thematically-related economic study presented by the National Bureau for Economic Research on Monday suggests that while AI and machine learning have received substantial investment over the past five years and have been widely touted as a transformative technologies, "there is little sign that they have yet affected aggregate productivity statistics... The simplest possibility is that the optimism about the potential technologies is misplaced and unfounded," muse Erik Brynjolfsson and Daniel Rock (MIT), Chad Syverson (University of Chicago) in the paper.
But instead the paper's author suggest that fully realizing the benefits of AI "will require effort and entrepreneurship to develop the needed complements, and adaptability at the individual, organizational, and societal levels to undertake the associated restructuring."

Comments

So?

[mark-t]

There was a time when what were once called horseless carriages was slower than using a horse,.

Give it time.

Re:So?

[Hognoxious]

The study says even less than that. It's more like some horseless carriages were slower than some horses, on an odd-numbered Wednesday if there's an R in the month and the wind's blowing from the east.

Re:So?

[mark-t]

There's no such thing as unrestricted human performance because human performance is always limited by what is actually physiologically possible in our biological framework. AI is restricted only by what is physically possible, which is a far broader domain.

Way to miss the point

[Anonymous Coward]

Wow, these people completely missed the point of these procedure techniques. They are not intended to make surgery more efficient or cheaper. They are intended to make them SAFER. They have a much lower impact on the human body and produce much less shock. They have shorter recovery time. They produce drastically less scarring.

Amazing how wildly out-of-touch technologists can be that they consider speed and cost to be more important than the health and well-being of the patient.

Re:Way to miss the point

[sjames]

Nobody did that here. They looked at 3 cases: Open surgery, laparoscopic surgery, and robot assisted laparoscopic surgery. They found that laparoscopic beat open in patient outcomes (so it's a good thing). They further found that adding the robot did nothing to improve outcomes but did cost more (a bad thing) and take longer (also bad and potentially a safety risk).

An exception, called out in the summary, is radical prostatectomy where the robot did improve outcomes.

Re:Way to miss the point

[Kjella]

They found that laparoscopic beat open in patient outcomes (so it's a good thing). They further found that adding the robot did nothing to improve outcomes but did cost more (a bad thing) and take longer (also bad and potentially a safety risk).

Unfortunately when you make statistics like this it's never the same doctor operating on the same patient. For example, it's quite possible that younger doctors are prevalent when it comes to robot-assisted surgery while older and more experienced doctors who had to learn to do it on their own continue to do so and that the actual difference is due to differences in experience not technique. It's also possible that those who do robot-assisted surgery can do it with less training than the unassisted ones without being properly reflected in direct costs. If you don't correct for the selection bias then sometimes you get the result that the most experienced doctors have the worst outcomes - because they treat the most difficult cases. It's not always that more data is better data, sometimes case studies where you go really in-depth and make really sure you have apples-to-apples comparisons are more valuable.

Re:Way to miss the point

[laird]

How is "But the patient outcomes were more or less the same" a statement that all we care about is dollars and cents. If the more expensive equipment didn't yield better medical outcomes, it's just wasting the patient's money (or the insurance coverage). The US already has bloated healthcare costs, because people like the latest, greatest stuff. But if the medical outcomes are the same, we're all better off doing things at lower cost, because that leaves more resources for other things of greater medical value.

And these weren't robots replacing surgeons, they were robots supporting surgeons.

Design you fools

[gurps_npc]

In other words, human doctors have made absolutely sure that the surgeries they design are easily done by the surgeon doing the planning, rather than saying "Hey, I'm not sure if I can avoid killing you, but what the hell, let's give it a shot."

No freaking da.

Sample size

[The Cynical Critic]

I hope that I'm not the only one puzzled with the number of surgical procedures being evaluated here. A single procedure done carried out using 3 different methods is hardly going to produce enough data to make any kind of assessment as how the different methods compare. Sure, you can talk about how they compare for this particular procedure, but even at that we're talking about footnote-level importance here.

Hell, I'd even go far as to argue that the real story here is how something as insignificant as this was given this much attention and how badly it's very specific conclusions were over-extrapolated to make a headline as eye-catching as possible.

Re:Sample size

[rhsanborn]

That's actually the point. We aren't trying to determine if robots are good or bad in general. It's irrelevant. The researchers are trying to determine if robot assisted surgery is better than non-robot assisted surgery for specific surgeries. This is how doctors determine the best course of action to use in a specific situation, and whether insurance companies will cover extra costs for those particular actions. It looks like 2 particular surgeries show no benefit of robots and that doctors probably shouldn't generally use those robots for those procedures, and insurance companies shouldn't pay anything extra for the use of robots in those procedures. We should hope they're working on studying the efficacy of robotic assistance in more surgical procedures.