Robots in Medicine

eberry writes "The Cincinnati Children's Hospital Medical Center will use a robot to mix intravenous medications and prepare its syringes. The robot, about the size of three refrigerators strapped together, can fill 300 syringes an hour, each with a custom dose and a bar-code label routing it to a particular patient. The robot should reduce the potential for errors and improve patient safety. The robot still needs further approval by the Ohio State Board of Pharmacy, but that should come within a month. It should be noted that five Cincinnati hospitals already use computerized pill-dispensing systems." On the other hand, reader Bobbert sends in a cautionary note: "'A group of German patients has filed a lawsuit against financially beleaguered Integrated Surgical Systems Inc., alleging that the Davis company' Robodoc surgical robot is defective and dangerous, according to a company filing with the Securities and Exchange Commission.' So now with robotic surgery, both the doctor and the robot can liable for damages. Next thing you know, telecoms will be liable for medical malpractice if the network connections fail during remote robotic surgery."

Comments

Human Error and Logic

[fembots]

Will these robots "sense" possibles error in the prescription though? For instance if the doctor entered the incorrect dose, an experienced nurse might just be able to pick it up, but a robot will just do as told.

It reminds me a tail strike incident where the pilot entered the incorrect weight and the system didn't pick it up. The incident report stated that the weight/speed combination should not have been allowed by the system at all, but nobody wrote that checking code at the beginning.

Re:Human Error and Logic

[dcarey]

Will these robots "sense" possibles error in the prescription though?

I'm sorry, your question does not compute. Shall we play a game of chess?

Re:Human Error and Logic

[Anonymous Coward]

I wonder what operating system they use in this
thing. Image how failsafe it might be if it actually runs any version of Windoze.

Re:Human Error and Logic

[Nurseman]

For instance if the doctor entered the incorrect dose

I've worked with these types of machines for years, they WILL pick up these kinds of errors, but they will also give alot of false positves. Many times the doctors will order more than the maximum dose, in emergencies, in cases where the person is very sick etc. The machine will not dispense "more" than it is programed to. In these instances, I just opened it with a key, and took what I needed. Drove supervisors crazy :-). The nice things is they pick up on interactions that me, the nurse, or the MD may not even know about.

Re:Human Error and Logic

[drmike0099]

All prescriptions in a hospital are reviewed by a pharmacist before being entered into the computer system. If the hospital has computerized physician order entry, they additionally go through checking as the order is placed. Nurses still take the drug in their hand and review it before they administer it. Humans still review everything.

This replaces the very error-prone menial task of filling up vials with the appropriate dose and concentration of medicines. Assuming the system works as intended, there is absolutely nothing being lost here, only gained.

Re:Human Error and Logic

[underpar]

It would help to the extent that a calculator prevents errors in math. If the input is wrong or if the answer is 'low battery' a person would have to be there to figure it out.

Re:Human Error and Logic

[xenocide2]

In theory, pharmacists are indespensable human elements that go over these prescriptions, ensuring that bad combinations of drugs aren't administered (some combinations of drugs have a habit of reacting in the blood stream and forming a precipitate that clogs veins and arteries, among other things). In theory, doctors are supposed to be superhumans as well, and not prescribe these combinations.

In reality, there are no super humans. Its not something the medical profession enjoys admitting. New studies of drug interactions come out regularly, and few can really keep up with the pace. If you were to test a pharmacist and a robot during a month long study, I'd expect that either the robot wins, or the pharmacist winds up being extra dilligent on behalf of the study and ties it for perfection.

You act like its impossible to program in failsafes, like nobody knows exactly how much is too much, let alone poor helpless software engineers. Certainly, lives are put at risk in both avionics and medical computing. In this case, however, one of the core duties is to check exactly for these things, which places extra emphasis on an already important task.

Re:Human Error and Logic

[MotherErich]

Anyway you look at it - it all comes down to human error. A doctor, nurse, programmer. The most rational way to look at it is what's going to help us save the most lives. There's always going to be human error, we can't prevent that.

Think about where the 'robot' is getting the prescriptions - what if someone puts the wrong medication in the wrong storage area of the machine? (I presume the machine's got a number of different med's to deal with) It'd be the same if the nurse somehow grabbed the wrong bottle from the shelf.

Bottom line: faster + more efficient = more lives saved
Unfortunately the equation gets a lot more complex when you factor in fear, doubt, and lawsuits.

There's always going to be someone to say, "if only it wasn't for that damn machine!"
But then there's always going to be someone to say, "if only it wasn't for . . . the countless other things that can go wrong in life."

-brother bummer & daddy downer

Re:Human Error and Logic

[nizo]

I would be more worried that the hospital will hire some untrained minimum wage slave to refill the robot instead of a more expensive nurse. Somehow I doubt the robot would know it is dispensing medication that has been loaded in the wrong place.

Re:Human Error and Logic

[word munger]

My father has worked for years in a robotic prescription-filling department. He visually checks each pill bottle the robot fills against a computer database photo showing what that particular pill is supposed to look like. His department only handles refills: a human has to fill the initial prescription. He says there are far fewer errors in this system than the old manual system, and it eliminates many sources of human error. Of course, a human still has to fill all the hoppers with the right pills, so there still is the potential for error there, especially with many pills looking quite similar. Then the error might be repeated over several hundred prescriptions, instead of just one.

Re:Human Error and Logic

[puck01]

I'm a doctor that works at this very hospital. The order entry system in place already screens most of the dosage errors and much more. After that a pharmacist and a nurse will also make sure the orders make sense. I'm not sure it would be necessary for this machine to carry out that specific task.

Re:Human Error and Logic

[puck01]

If the hospital has computerized physician order entry, they additionally go through checking as the order is placed.

This particular hospital does have a computer order entry system that was built largly in house specifically for pediatrics. As a doctor at CCHMC, I can say its pretty good for the most part. The screening is, if anything, is too sensitive, but overall it is very effective at catching my errors. Fortunately, the environment there is such that feedback is very much welcomed and taken very seriously so it continues to improve all the time.

Re:Human Error and Logic

[Anonymous Coward]

The nice things is they pick up on interactions that me, the nurse, or the MD may not even know about.

There are even more false positives with interactions than with doses (I am a pharmacist with 8 years experience and I can only think of a couple of instances where above maximum doses are clinically appropriate - 'really sick' isn't one) The problm with interactions is that they are much more 'fuzzy' and unpredictable and not understood by the software engineers or their 'experts' - a tip for
any developers DO NOT ASK DOCTORS they don't understand pharmacokinetics and therefore they do not understand most drug interactions.

Re:Human Error and Logic

[StarsAreAlsoFire]

Hear hear.

When I was about 12 I got sick. My mother took me to the doctor, and he prescribed some meds for me. We went to the pharmacy and picked them up. I was out of it, and waited in the car.

Doctor said to start right away -- Not out of urgency, it was just some nasty fever that I probably shouldn't have been medicated for, but at anyrate... So I popped a pill upon getting them.

About an hour after we got home, phone rings. Turns out is the senior pharmacist in a state of controlled panic --- Tells my mom to NOT give me any more of the drug, asks how much I've had, and gets our address. He then IMMEDIATLY drives out 20 miles to deliver the CORRECT medication. I'd been given some heart medication or another. This was a MAJOR pharmacy, doing more business then than the typical Walgreens does now.

Moral of the story? Well, my whole family now checks every drug we buy at the store with the drug interaction guides -- I use the 'net, the parents use some HUGE book. And I know that this would be much less likely to happen today; BECAUSE OF computers. However! If it DID happen -- e.g. hadn't been a drug that was so obviously NOT aimed at a 12 year old, having a human on staff that is TRAINED could very well save the day.

To me, it is like adding yet another layer of protection. Just because you bought a car alarm doesn't mean you stop locking the doors of your car. Or just because you have installed that happy parachute system on your small airplane doesn't mean you are going to stop checking to see if your buddy put in enough gas. Layers. We are adding layers.

In other news...

[Anonymous Coward]

The governments of Vancover, Canada and Amsterdam, Netherlands have placed orders of 10 of these machines each presumably to placed on street corners.

Re:In other news...

[TheGavster]

You know, putting powerful painkillers in vending machines could mean the end of income taxes, if done properly. Not to mention taking out the bottom levels of society as well ...

Re:In other news...

[raider_red]

Also, Keith Richards has had one installed in the drawing room of his English country home.

Re:In other news...

[Anonymous Coward]

Also, Keith Richards has had one installed in the drawing room of his English country home.

To much envy from Jimmy Page, who's announced that he's definitely getting one too

Just someone else to get sued

[drsmack1]

I hate frivilous lawsuits, but at least with a human doing the filling of drugs there is some common sense that can be a fail-safe. With a machine all it takes is a bug to have 300 vials of poison dealt to unsuspecting patients. Won't there still need to be human oversight?

Re:Just someone else to get sued

[tygerstripes]

To the best of my knowledge, it's human oversight that causes most drug administration accidents...

Re:Just someone else to get sued

[Anonymous Coward]

Maybe we need Robot Lawyers for the robot pill-dispensers.

Re:Just someone else to get sued

[dcarey]

but at least with a human doing the filling of drugs there is some common sense that can be a fail-safe. With a machine all it takes is a bug to have 300 vials of poison dealt to unsuspecting patients. Won't there still need to be human oversight?

Kinda frightening isn't it? For comparison:

Robot Bartender. Error = client is drunk.

Robot Pharmacist. Error = client is dead.

Re:Just someone else to get sued

[jeffy124]

I dont think a computer bug is likely to be the cause of vials of poison being accidently sent to patients. Human error might strike first in the form of someone refilling the machine with the incorrect drugs. Instead of replacing a canister of Drug A with a new canister of A, an orderly accidently hooks up Drug G. Result could be a dead patient who needed Drug A and got G instead.

As an analogy - stores like Loews have a machine that mixes paint to the customer's desired color. The machine starts with a base color (eg, pure white) and adds various amounts of red, blue, yellow, black, etc to the can to get the desired color. A similar error here would be replacing an empty blue canister with a red canister. The error likely wont be noticed until someone gets a can of paint that required the blue mix. At least this would be easy to track down when the store employee goes to check the paint color before selling the paint.

Re:Just someone else to get sued

[Anonymous Coward]

With a machine all it takes is a bug

This system is not so complex that bugs cannot be eradicated. If someone builds a system that takes source A and source B and mixes them into spout 13 in such a way that it is too complex to maintain, they need to start over.

In fact, the system probably consists of 500 plastic tubes hooked up to the medicine components with 500 little pumps that squirt out a measured dosage into the mixer, which spins a little then dumps its contents into a syringe (if the mixer wasn't already the syringe itself), which then drops into a labeller and rolled out a chute in the front of the machine. Flush the mixer between each batch.

The drug mixing part would basically be a lookup table (ie, for drug xyz we need 3ml of #234, 1cc of #13, and 1cc of saline). Making sure the lookup table was correct for every possible drug would be about as fun as proofreading a phonebook, but possible.

The labelling part would be harder, especially at the speed they claim to be able to dispense. Making sure the labeller was always in sync with the mixer, and guaranteeing the labelling was done in time for the next syringe to be labelled.

Re:Just someone else to get sued

[drinkypoo]

And filling the machine with the wrong drugs could be easily checked with bar code or rfid. Next objection?

Hacker practical joke

[Anonymous Coward]

All you need is a hacker with a sense of humor and that circumcison can turn into a sex change.

Re:Hacker practical joke

[Anonymous Coward]

Interesting... would the patient not notice?

Puts the I in IRobot

[Anonymous Coward]

The I stand for Injector++.

Robotic anal probing?

[Anonymous Coward]

When can we look forward to the day when robots do the bulk of anal probing and diagnostic screening for colonoscopies? As a proctologist, I would gladly hand some these types of jobs over to robots.

Robots in the hospital

[Icarus1919]

Until the system is fixed so doctors and nurses don't have a constant case of jet lag from being up for different shifts every day, introducing new ways to prevent careless errors is the best way to save lives.

Re:Robots in the hospital

[LostSinner]

My girlfriend actually works at Children's here in Cincinnati. I just emailed her about them making it into a /. article and haven't had a chance to talk to her about what she knows, but I thought your post was quite perceptive.

The schedules that nurses and doctors work at hospitals are just bizarre to me as a guy with a regular schedule. I don't work a job where people's lives are in my hands on a regular basis, but I have a regular schedule. It's not odd to see my girlfriend work 7AM-3PM on monday, 11:30PM tuesday to 7AM wednesday and then 3PM to 11:30PM wednesday and then a twelve hour shift on the weekend. She's always worn and stressed out when she comes home. It seems to me that these kinds of professionals in particular could benefit from having a regular schedule. I'll never understand it.

Re:Robots in the hospital

[puck01]

As a doctor, I agree that providing us with more human-like work hours would overall be beneficial to all involved. This would certainly decrease the number of errors and improve patient care. However, not all of the errors I or other doctors make (which are usually picked up by a pharmacist, nurse or other) are because of lack of sleep. Often time its a simple oversight or calculation error. There are so many potential variables that can influence the dose and schedule of a drug or treatment, that any human will have a difficult time getting it right 100% of the time, even in the best circumstances.

so while I agree, we need to improve the work hours, its only one of many problems. Fixing the way orders are screened and implemented is certainly among a long list of things that need to be improved to reduce these errors.

Poor analogy...

[rednip]

This analogy is unfair...

' So now with robotic surgery, both the doctor and the robot can liable for damages. Next thing you know, telecoms will be liable for medical malpractice if the network connections fail during remote robotic surgery."

When you build a product, there is (at least) an implied warranty that it is fit for a specific use. A surgical robot, *should* be able to conduct an operation. We aren't talking an apples and oranges thing here. I think the auther is trying to place a back end comment about tort reform.Now tell me again why we need tort reform...

oh, yea, Malpractice is up 25% in 10 years (but medical costs have risen much higher...).

Re:Poor analogy...

[sessamoid]

oh, yea, Malpractice is up 25% in 10 years (but medical costs have risen much higher...).

Nice try. In Florida where I used to practice, our malpractice insurance premiums increased 50% in the last two years. Medical costs aren't rising at anything like that fast. Sure we make a comfortable living, but nobody likes to be paid a lot less for a lot more work. Keep that up for a few years, and it really starts to get old.

Eep! Imagine the barcode scenario...

[tygerstripes]

Boop...

Boop...

Brrz!
"Benzadrine. Price check on Benzadrine."

*shudders*

There's still a level of human interaction

[bwcarty]

From what I read, the robots don't administer the injections.

Having been through chemo, I know that the first thing the nurse did each time was show me each of the syringes that were to be injected into my IV. Each was labelled with the medicine name and dosage.

I never saw the syringes being filled, but since I'm still alive, I trust that there's some degree of verification before I even saw the bag that contained all my chemo meds. For all I know, a robot could've mixed the meds, and I'd be none the wiser.

Re:There's still a level of human interaction

[Nurseman]

Having been through chemo, I know that the first thing the nurse did each time was show me each of the syringes that were to be injected into my IV. Each was labelled with the medicine name and dosage.

Its called the "Five Rights" it is how you are supposed to verify the patient : Right Drug ?
Right Dose ?
Right Route ?
Right Patient ?
Right Time ?

I dont care if it's Tylenol, the nurse should ask you this each time he/she gives you anything.

Re:There's still a level of human interaction

[GigsVT]

Kinda hard if you are giving it to someone already half-asleep on vicoden or whatnot.

Re:There's still a level of human interaction

[Anonymous Coward]

And now the six right- write documentation.

As an aside, one place I worked at had a system similar to the one in the article with bar codes PO/and some IV medications right down to the patient level. While it did significantly reduce the amount drug errors related to the wrong drug being given (and also to the wrong person... for a time), it seriously increased another type of drug error: drugs given at the wrong time, which ultimately lead to poor drug interactions.

The increased overhead of steps necessary to dispense a drug wasn't accounted for on the floor, nor the fact that a patient might be in X-ray longer than the allotted window to give the drug (the drug schedules were synched with the Pixas. If you were outside the time window, the drug would not be dispensed. You had to call down to pharmacy, which took even more time and threw the other drug schedules out of wack until someone in pharmacy updated the patient data, which often conflicted with new orders received, ad nauseam...).

Ultimately this led to nurses circumventing the system entirely (keeping bar code tags in their pocket, etc.) and keeping a separate handwritten account (even more overhead) of actual drugs dispensed and with times to be passed on to the next shift. Oh, and spending the next hour or so imputing the actual drug record into the computer.

I noted a similar system was never implemented in the ER.

While I am all for technological aids to help ensure the safety of my work, I get the sense of being dumped on by immature technologies, or those technologies being dumbed down to the level of the most incompetent nurse (i.e.-inflexible).

This does not bode well, and only changes where the mistakes will occur.

Remote robotic surgery...

[Anonymous Coward]

I hope the PunkBuster folks are hard at work on an update, or I see nothing but problems ahead for this technology.

Prescribing errors.

[Tackhead]

> I hate frivilous lawsuits, but at least with a human doing the filling of drugs there is some common sense that can be a fail-safe. With a machine all it takes is a bug to have 300 vials of poison dealt to unsuspecting patients. Won't there still need to be human oversight?

Take a look at your doctor's handwriting the next time you get a prescription. If you can't figure it out, your pharmacist probably can't either.

Human oversight is having sufficient presence of mind to ask your doctor "What drug am I being prescribed? At what dosage? In what form?", remembering the answer, and comparing what your doc told you with what's on your prescription... and with your pharmacist gives you after reading your prescription. In at least one recent study, around 6% of prescriptions result in errors.

In the absence of that oversight, I'll take my chances with the robot.

Re:Prescribing errors.

[multiplexo]

No kidding, there was recently a case at Virginia Mason Medical Center in Seattle where a woman who was admitted for a routine procedure was killed when she was injected with the chlorhexidine solution used to prep the skin prior to catheterization instead of the contrast dye. So much for human oversight.

Re:Prescribing errors.

[stupidfoo]

I'll take my chances with the robot.

Nice line. It'd make a good sig, me thinks.

Tort reform urgently needed!

[Anonymous Coward]

To bring back sanity to the medical system, we need to keep patients from being able to sue doctors or hospitals for malpractice. We need to make sure that those who are trusted with our health have all the tools necessary to get their difficult job done, so we need to end the insanity that is medical malpractice lawsuits.

Re:Tort reform urgently needed!

[Kobun]

So the doctor who mis-diagnosed 2 of my grandfather's three heart attacks should be totally protected from all consequences of his actions? I think you need to clear up the above statement, otherwise it seems extremely stupid right now.

Re:Tort reform urgently needed!

[YrWrstNtmr]

To bring back sanity to the medical system, we need to keep patients from being able to sue doctors or hospitals for malpractice. We need to make sure that those who are trusted with our health have all the tools necessary to get their difficult job done, so we need to end the insanity that is medical malpractice lawsuits.

And before we can get to that point, we need to insure that doctors don't actually perform malpractice

Re:Tort reform urgently needed!

[Anonymous Coward]

Absolutely. It is because of liberals like you that the costs of medicine are going up much faster than what people can afford. Doctors are not gods, they make mistakes. Patients need to understand that doctors can not and should not be sued by every tom, dick and harry that wants to make a quick buck off their parents heart attack or bout of cancer.

Re:Tort reform urgently needed!

[Duhavid]

Doctors are not Gods, that is true.

Doctors are human beings, capable of all the range of human deed and misdeed that everyone else is subject to.

Should they be given carte blanche in their actions? No consequences?

I would agree that it is a bit too easy to sue, and that there are some people who take advantage of that. I dont think that eliminating the ability to take a doctor to court is the answer. That would put us in the position of the doctor being able to do anything with impuny.

Where in the parent ( to your post ) do you see someone looking to make a quick buck?

In every other industry, if you make a mistake in your product or service that harms people, you are liable for that harm. In what way is this different?

Re:Tort reform urgently needed!

[stimpy1306]

Oh please...

in the example you cite

1,000,000 procedures annually:

20 cases - Wrong Patient, Wrong Site - Surgical

103 cases - Incorrect Procedure - Invasive

90 cases - Retained Foreign Body

incidence of error .00021%
( i.e. the probability that an error might occur )

or

99.99979 of procedures had no errors

that's a low incidence by any standard.

Tell me when did doctors stand up and pronounce that they were perfect ?

In fact that's higher performance then almost every system you come in contact with every day .

To put this in perspective your incidence of auto accident in the US with injury is 1.68%...8000 times more likely !

Should doctors and hospitals try to avoid mistakes ..YES

Should doctors who make too many mistakes face consequences ..YES

BUT your comment just shows common unrealistic perspective about reality.

Is a mistake malpractice ? Even if that mistake can be fixed ? Even if that mistake has little or no effect on the outcome ? Is a surgeon or hospital that has a .00021% error rate guilty of malpractice ?

Do you want to pay a medical bill where 60% of the bill is coverage for insurance ? Even if the doctor in question has never had case of malpractice?

Medicine at the most basic level involves TWO people ..thus the expectation of 100% certainty is absolutely unobtainable and unrealistic . Furthermore the cost of this level of reliability is more than anyone but the richest of the rich can afford.

No one want to have a mistake made on them , everyone wants medicine to be perfect but outside of the "bill gates" of the world no one can afford it .

Robots in medicine can help lower the incidence of mistakes . however mistakes will still exist and a mistake does not equal malpractice.

If we want to improve the medical system them we all need to have a realistic view of the possibilities.

Re:Tort reform urgently needed!

[rednip]

Doctors should feel lucky, a bus drivers kills his passengers because he is too tired to drive, gets charged with homicide and thrown in JAIL. Doctors who have been up for 24 hours and prescribe medicine which kills a person, runs the risk of getting sued, damn that quite a racket.

Re:Tort reform urgently needed!

[YrWrstNtmr]

Never said they proclaimed to be 'perfect'. And yes, that is a VERY low incidence of (reported) problems.

However...completely removing the possibility of doctors being held responsible for their actions is not only wrong, but insane. "Wrong site, wrong patient" is exactly what malpractice is supposed to be for.

I agree that it is far too easy, nowadays. Let's change it, not eliminate it.

Re:Tort reform urgently needed!

[stimpy1306]

Ok 20 cases out of 1,000,000

Thats .00002 % incidence of "wrong site/wrong pt" not exactly an epidemic , nor do we know that any of those cases involved the same doctor or even the same hospital.

While i agree that at face value " wrong site/ wrong pt " should be malpractice let's not jump to conclusions.

While "wrong pt" is harder ( but not impossible) to justify , i can think of quite a few reasons why one might operate on the wrong site.

After all the problem with assessing malpractice is that it always involves the use of 100% diagnosticly accurate HINDSIGHT.

The benifit of 100% diagnostic accuracy is something that doctors almost never have.

Completely removing doctors from responsibilty is something many propose..."lets model england or canada" they say ... doctors in countries with socialized medcine have vast immunity.

" we need a nationwide HMO" ...until people realize that suing a HMO or it's doctors is almost impossible.

As someone who has attended and testified in many medical malpractice trials i can say that the current method for dealing with medical mistakes and malpractice is dreadfully broken.

This issue i was trying to make with the earlier post was one of context . The common person hears "medical mistake" and leaps to "malpractice". That very leap is the first thing we need to change.

I would like to think that the slashdot crowd is above such base assumptions. While this may be a silly position, my last shred of optimism tells me that the best place for rational change to begin is the smarter then average crowd here.

Patients, not patents.

[AltGrendel]

A group of German patients has filed a lawsuit...

Whew, for a second there I thought the EU patent system had screwed up.

Re:Patients, not patents.

[Hiro Antagonist]

No, see, that's Germany. Only in Soviet Russia did patents sue you.

Telecom liability

[GoofyBoy]

"Next thing you know, telecoms will be liable for medical malpractice if the network connections fail during remote robotic surgery."

Telecoms usually have a clause for any business loss due network disruptions. I think that would apply here.

Re:Telecom liability

[Doc Ruby]

Then the network architect who designed that unreliable network into the critical path, with lives on the line, is guilty of malfeasance.

Re:Telecom liability

[GoofyBoy]

Lives are on the line today with telecom networks, this robotics situation wouldn't be anything new.

911 calls is one example.

Re:Telecom liability

[Doc Ruby]

Yes, as the 911/VoIP hearing we held last year in our Technology Committee of the NYC City Council explored. And as our 4-hour 911 outage in NYC last Fall demonstrated, when someone died of a heart attack while waiting too long for an ambulance to be dispatched. The network was the nexus of liability, but it looks like the City will be held liable, rather than Verizon, in the family's lawsuit. Because it's easier to get the City to pay quickly than to squeeze money from Verizon, not because of justice. In the robotics case, with NYC and Verizon out of the equation, the robotics provider would be liable, unless someone else in the critical path overrode their disclaimers of liability.

Fear is part of the problem

[paranode]

People just don't like to trust machines. Some of this is for good reason, but all faults in machines lead back to human error. If humans incorrectly filled, say for example, 200 prescriptions a year and ended up killing 10 people it would be bad and maybe some people would get sued and some folks would lose their licenses. If a machine made one mistake in the course of years that resulted in a death, we'd have everyone up in arms talking about how this could have been prevented and that we're letting people die at the hands of evil machines and then we'd have a battery of laws passed against machines. Unfortunately this sense of losing control takes over people and fear kicks in, even if the machine is 100 times more accurate than a human at the same task.

Re:Fear is part of the problem

[Hyecee]

This kind of fear has more to do with a "bandwagon" mentality combined with a kind of herding instinct than it does with people not trusting machines. I'll agree that there is a general phobia of anything new, but would argue that the herding-bandwagon mentality is more responsible for the kind of reaction you envision. The same reaction is evident all the time toward specific incidents that are over-hyped out of proportion. One innocent shark mistakes a swimmer for a seal and all of the sudden the local government is issuing rewards for shark skins. The only key is that something has to happen to the victim that they themselves could not do anything about. No eating the pretty (and poisonous) mushroom, no bungee jumping accidents. You have to have been walking along a cliff and have it cave in under you to make sure it gets fenced off to the rest of the world.

Re:Fear is part of the problem

[pete-classic]

Part of the fear is rational and part is irrational.

I basically agree with your post. But you seem to have left out both error-multiplying nature of machines and their general inablilty to detect errors.

-Peter

Re:Fear is part of the problem

[Hyecee]

Or I could just not see your whole last sentence.

Well, Instead of a rebuttal, apparently I just agree with you.

Re:Fear is part of the problem

[Anonymous Coward]

The area I reside in has a LOT of railroad tracks crossing the road.

Most of these crossings are "blind" because of trees, curves, etc...

I've never once in my life checked to be _sure_ that there was no train coming when the warning lights weren't flashing.

And I've never thought about this fact until now.

How many times have I trusted my very life to what must amount to nothing more than a simple relay circuit?!

I say bring on the bot. :)

Re:Fear is part of the problem

[Chris Burke]

There is a good reason to be leary of machines: automated fuckups.

The blessing and curse of machines is that they can do the same thing a thousand times over precisely and very quickly. If it's the right thing, they're great. If it's the wrong thing, you're really screwed. To manipulate your example, if 100 people stupidly forwarded a virus-laden email to all their friends, that would be bad. When someone invents an email client that can do the forwarding automatically without the human doing anything, you get "I Love You". Or imagine a nurse filling a dozen syringes with the wrong dosage versus a machine filling several cabinets full.

Machines are great, but I am absolutely for having only the highest of standards for their use for medical purposes. The laws should not however be based around banning machines, but around nailing anyone to the wall who sells a faulty machine due to negligence, inadequate testing, or fraud. I'd like to think the mechanism and laws for this are in place, but recently the FDA has been, shall we say, dissapointing.

Brings a whole new meaning to...

[Message Board]

blue screen of death

Re:Brings a whole new meaning to...

[HarveyBirdman]

blue screen of death

Step into the light...

Re:Brings a whole new meaning to...

[quarkscat]

No human endeavor is completely free of error,
and there is no such thing as software without
any coding errors. If the robot pharmacist
makes a mistake filling prescriptions, there
will not be one dead patient, but perhaps 300.

Oops! There goes the medical malpractice
insurance premiums, through the roof, as
well as the demise of the robot manufacturer.

Anyone willing to start a pool? The long term
odds are in favor of a catastrophe.

Interesting Thoughts about Telecoms

[Kobun]

A curious thought struck me about the submitters comments. Telecoms would almost certainly have heavy contractual stipulations about their degree of liability, if they were the providers for remote surgical connectivity service. IANAL, of course, but it would seem to me that there would have to be very specific events described where the telecom would be at fault (their own operator cuts the power/line, for example) while most other occurances (freak storms, non-related contractors) would be excluded. Has anyone had any personal experience with this out there? It would be interesting to hear how they set up a liability structure for such a critical thing. Also, during remote operations is there a qualified in-person backup handy?

Gives a whole new meaning to "Pusher Robot"

[Anonymous Coward]

I don't trust that one with my life. I'm only letting the Shover Robot dose me with the morphine. He'll protect me from the terrible secret of catheterization.

Re:Gives a whole new meaning to "Pusher Robot"

[Anonymous Coward]

Does your house have stairs?

Futurama

[bharatk]

I for one wouldn't want some foul mouthed robot like http://www.gotfuturama.com/Information/Encyc-8-Ben der/ Bender preparing my botox injections.

use your common sense

[painehope]

Next thing you know, telecoms will be liable for medical malpractice if the network connections fail during remote robotic surgery

Yeah, and engineers might be held responsible if the bridges they design and build fail under normal/expected operating conditions.

Oh, that's right, I believe they are.

If the country gets hit with a tactical nuke, I think it's understood that shit happens. If some underpaid joe in Bumfuck, Idaho drives a piece of heavy machinery through the fiber conduit, I expect you to have a near-transparent failover. That's what engineering is about. It's about having the knowledge and experience to design and test well. That's why some people have objections to MCSE or RHCE certs using the word "engineer".

If you're providing the network service for my remote robotic surgery, you goddamn well better have a fault-tolerant re-routable network in place. And an on-call heart surgeon who can be there in minutes. Because if your negligence messes me up, you better believe that myself or my children will pay you a visit personally. We'll have a little chat and it will involve a butane torch and a ball peen hammer. That's a personal message from me to you, mister golden-parachute budget-cutting book-cooking CEO.

Re:use your common sense

[TheGavster]

The problem comes when you can either have a lifesaving proceedure performed by a robot over a fallable link, or go without and die. People get enraged over the failed link, even if without it they had certainty of death instead of X chance of death.

Re:use your common sense

[painehope]

there's just one problem here. You're assuming that it's not doable to have a system that is reliable to 99.999...% - it is. It's not that hard to have redundant routing into a site, across the net, and into the other site.

Shit, if we can keep porn sites up, we can keep a link that someone's life depends on up - in most cases. And you can most definitely have an on-staff physician ready to step in. That's the ultimate fail-over procedure.

And as far as people getting enraged, if you tell me that I have at least a betting chance on a procedure, and the staff do their best, I'm not going to be mad if I die ( not that I would have much to say anyways ). If I'm given a 10% chance to survive an operation, I'll just go home and drink whiskey, eat steak, and shoot heroin until I die. But if I elect to have an operation, and I die because some wanker couldn't design a redundant link, then I fully expect my children to exact a brutal vengeance on whoever is ultimately responsible.

RALP

[Billy the Mountain]

My boss uses a robot for performing Robotic-Assisted Laparoscopic Prostatectomies. I've got an 8 minute video at www.DrSlawin.com so you can see how they place the ports--they screw the robotic manipulators right to the anesthetised patient's abdominal wall. Then the doc sits down at the console and operates away.

BTM

Cool Possible Acronym Name

[Anonymous Coward]

I'm wracking my brain to come up with words for "DEALR"...get it? It's a robot that mixes drugs and prepares syringes so it's a dealer...eh, nevermind.

Oh, the 'R' could be for Robot! :p

Re:Cool Possible Acronym Name

[Duhavid]

Dose Effecting And Labeling Robot.

Oh my!

[HarveyBirdman]

The robot, about the size of three refrigerators strapped together...

Well, I hope it's not for internal use. Can you imagine that thing crawing up your colon?

Oddly, I think some of you could. :-)

Aw, man, here comes another Troll/Offtopic mod. :(

read the fine print

[museumpeace]

...Next thing you know, telecoms will be liable for medical malpractice if the network connections fail during remote robotic surgery."...

I like lawyer bashing as much as the next human but buy just about any grade of service from any telco and you have signed an agreement which, in some obscure paragraph, says more or less "provider will not be held liable for consequential damages that may result from interrupted service". The Disruption of Service clause for Cablevision is typical.

robots suing robots

[hochopepa]

Next thing you know, telecoms will be liable for medical malpractice if the network connections fail during remote robotic surgery. Raises an interesting point. Is it ridiculous to hold an ISP accountable for a reasonable service standard in such a case? Should telecoms be necessarily immune to lawsuit for failing to provide a reasonable level of service for a critical care applications? What if the company boasted 99.999% uptime as part of their sales pitch, but failed to deliver that level, and some of the outages cost innocent patients their lives or wellbeing? The client service agreement the doctor signed to get internet service specifically disclaims any liability for losses the doctor might suffer, so it may be up to the patient to hold the ISP accountable. Or would it be more sensical to make ISP's immune to such liability altogether?

Time to start coding again...

[KennyP]

That way when I go to the hospital and need drugs, I'll get 'em!!!

Kenny P.
Visualize Whirled P.'s

At least don't have these robots perform surgery

[Man in Spandex]

The Pusher and Shover robots must not do what they do best in a hospital! Unless of course the remedy is to shove bread down somebody's throat.

Paging Doctor FUD...

[Shoten]

Next thing you know, telecoms will be liable for medical malpractice if the network connections fail during remote robotic surgery.

Actually, no. Medical malpractice isn't even remotely like what a telecom would be liable for, no matter how badly they screwed up, unless they were actually practicing medicine. What they could be liable for in the above-stated situation is negligence, and frankly I don't have a problem with that. There's nothing exotic about high availablility networking these days.

This scenario also fails to take into account the fact that the link failing wouldn't be the end of the world. It's not like they just wheel the patient into the operating room and leave them there so the robot can go at it, and it's not like the robot will start wildly flailing about with scalpels and other sharp instruments just because it's no longer being told what to do. And lest we forget, the patient whose robot-surgeon has just stopped working is still all set up in an operating room, on an IV with people monitoring their vitals, in the midst of a well-equipped hospital. Not the end of the world at all.

Re:Paging Doctor FUD...

[LiquidCoooled]

A broken connection would be similar to a surgeon with the shits.
It happens. Your right, its not the end of the world.

but because its robots, and because its technology, the infrastructure will take the fall. Just look at the excuses in the counter strike world.
"Poor larry with his 17 bullet wounds through every major organ would be alive today if it wasn't for the high lag."

Re:Paging Doctor FUD...

[Shoten]

You're kidding, right? A guy on CS who blames lag for getting shot up as a legal precedent in a medical malpractice case? Jeez man, even Judge Judy wouldn't go that far!

Re:Paging Doctor FUD...

[LiquidCoooled]

No of course I wouldn't use CS as a precident.
I was merely pointing out the way technology is blamed.
The person responsible for setting up a remote reobotic surgeon had better use connection models that don't depend upon insecure/unstable pathways, or HE will be the only one purely to blame for any cockups.

how long until...

[lysergic.acid]

the rx robot develops a morphine habit and starts skimming off shots of pain killers to feed its own habit?

but seriously, is our nation's medical staff so incompetent/overworked that they can't even load a syringe properly? if so, removing this particular responsibility from their job will only give them more chances to cause potentially fatal blunders in other areas. i've heard so many horror stories about doctors and nurses collapsing patients' veins trying to administer IV medication that I'd almost trust myself more with a syringe than hospital staff. Maybe instead of paying for this $640k robot, they ought to invest more in better training for hospital staff.

Re:how long until...

[OwnedByTwoCats]

Loading a syringe properly is no problem.

RTFA, and see that the pharmacy at this hospital fills 380,000 syringes per year, that's an average of over a thousand syringes per day. That will take a few staff, and coordination about what goes where, and be an incredibly boring job.

Automating repetitive jobs is a way to see that they get done reliably. And you can automate checks that the dosage is within the allowed range of mg/patient kgs.

Of course, there is still the potential that the medicine that the robot thinks it is dispensing in to the syringe isn't the same as the medicine that the robot is actually dispensing into the syringe. The person filling the robot put the right flavors into the right dispensers.

Re:how long until...

[nursedave]

Medical errors have many causes; there are lots of papers in the literature out there, I won't list 'em here. As to your comment about doctors and nurses collapsing patients veins, no. Infiltrations occur, not collapses. An infiltration is where the IV catheter is either not in the vein at all, or more likely, has been in long enough that fluid is able to leak out of the insertion site in the vein and into the tissue. Depending on what that fluid is, you can see either minimal problems (swelling, pain) that go away, or big problems, like tissue necrosis requiring plastic surgery. Pretty shitty deal, for sure. So, you do things to continually reassess the IV site to avoid problems. Where I work, we assess the IV site every hour if it is infusing something which causes tissue necrosis. And if the fluid is running fast enough, you'll see the telltale signs - swelling, redness, complaint of pain. But if the fluid is only running at about 1ml/hr, and the IV is in, but just leaky around the catheter, then fluid can leak out, and cause damage before the problem is evident.

Hospitals spend tons of money in continuing education for medical staff. To keep your license in most states, a certain number of continuing ed. hours are required every year. IV injuries are most of the time unrelated to the 'skills' of a caregiver - they sometimes just happen.

Re:how long until...

[lysergic.acid]

well, maybe it's not common occurrence but it does happen. i have a few friends who have testified to it. and no, what happened to them does not fit the description that you gave. usually it's a doctor or nurse who tries repeatedly to hit the same vein(probably dullying the tip of the needle in the process) and eventually just ends up collapsing the vein from all the trauma. this causes huge track marks to form the next day and severe bruising that covers the entire underside of the forearm. it also causes the vein to disappear.

i can understand junkies fucking themselves up like this, but it's not acceptible when a medical "professional" causes this kind of damage.

Re:how long until...

[nursedave]

Ah, I see! You're damanding perfection. Geez, that would be cool, eh?

I've been starting IVs since 1989. I'm pretty good, too; I currently work with newborns, and can get in most of the time.

However, sometimes, a patient has very difficult veins to hit. The sad irony is, the sicker you are (or at least, the more dehydrated), the harder your veins are to successfully canulate, but the more you need it. Its a bitch. I've seen situations many times, where multiple people, including staff nurses, physicians, and nurses on IV teams (these are nurses in the fortunate hospital that has an IV team whose whole job is responding to pages to start IVs on kids who have proven difficult to establish access on) were simply unable, after multiple tries, to get an IV on a patient. I have then seen, on these same patients, intensivists unable to establish central venous access - basically a surgical procedure where they go into the major vein in the groin or under the collarbone or one of the jugular veins.

Now, establishing venous access is technically very simple - put a needle with a thin plastic sheath over it into a vein, withdraw the needle leaving the sheath, confirm placement by flushing with saline, and secure the whole works. And most of the time, it is this simple. But when you have a patient that several nurses and doctors can not get access on, you must consider the one common denominator: the patient; specifically, the patients veins. Things have been developed to assist in finding veins, such as special transcutaneous lamps and ultrasounds, but it still boils down to one steady handed person trying to get a very thin metal thing into a vein which may or may not be appropriate for use (too thready, too small, not straight enough) but are the best choice available.

As to some of the other things you've said: doubtful needle dulling was a problem. Steel vs. flesh being the battle it is, with steel winning, is irrelevent, as each 'stick' requires a new IV catheter, for infection control reasons. 'Not acceptable when a medical "professional" ... does it..... If a medical professional (no quotes needed, but thanks) does it, then I wonder if you'd prefer a nonmedical person come give a go?

Oh, the units...

[dema]

The robot, about the size of three refrigerators strapped together...

That sounds large, how many Burning Libraries Of Congress is that?

Re:Oh, the units...

[CustomDesigned]

[Three refrigerators strapped together] sounds large, how many Burning Libraries Of Congress is that?

"Refrigerators" in this context is a measure of volume. BLOC is a measure of energy. There is no obvious conversion from energy to a volume unit. e =mc2 doesn't help without specifying a density.

A typical refrigerator is about 60x30x36 cubic inches. The transportation industry is interested in the volume, and in the largest dimension (or the largest dimension of a rectangular box that can contain the object). Assuming the refrigerators are strapped together on their smallest dimension, the resulting machine would be 60x90x36 inches, for a total volume of about 112.5 cubic ft and largest length of 7.5 ft.

Re:Oh, the units...

[Anonymous Coward]

the resulting machine would be 60x90x36 inches

Exactly! But how many Burning Libraries Of Congress is that!?

Robot Lawyer!

[The I Shing]

Who better than to defend the robot surgeon than the RoboMouth 3000, the finest robot lawyer in production today? Why, the RoboMouth 3000 can file motions 63.7 times faster than the fastest human lawyer, and can should "Objection!" at 135db before opposing counsel finishes the offending remark!

Re:Robot Lawyer!

[Rycross]

Or some sort of lawbot perhaps?

Error code

[Anonymous Coward]

0xDEADBEEF

I don't know why the scientists keep building them

[stratjakt]

This is a deliberate setup.

They eat old people's medicine for fuel. And when they grab you in their metal claws, you can't escape because they're made of metal.

Depression

[the_twisted_pair]

The electron ram stabbed out another searing blaze of light and took
out the appendix.

"How do you think I feel?" said Marvin bitterly.

"Just ran off and left you, did they?" the machine thundered.

"Yes," said Marvin.

"I think I'll shoot down their bloody ceiling as well!" raged the tank.

It took out the ceiling of the theatre.

"That's very impressive," murmured Marvin.

"You ain't seeing nothing yet," promised the machine, "I can take out
this floor too, no trouble!"

It took out the floor, too.

"Hell's bells!" the machine roared as it plummeted fifteen storeys and
smashed itself to bits on the ground below.

"What a depressingly stupid machine," said Marvin and trudged away.

(with apologies to Douglas...)

Sounds like a good use for robots...

[ion_]

Sounds like a good use for robots, as long as they don't hire the Red Robot.

I doubt that this gets state board approval.

[Anonymous Coward]

The state of Ohio pharmacy board might approve this robot but I highly doubt it. The rules for Rx dispensing in Ohio are not very flexible or accomadating when it comes to technology. I doubt they would approve non-human supervised dispensing regardless of the accuracy.

Malfeasance

[Doc Ruby]

If the telco sells the doctor bandwidth for telemedicine, then violates the uptime guarantees, they're liable. Maybe not for malpractice, but neither is "the robot" (really its makers, in another confused accusation). When lives are at stake, tech suppliers are liable for failures they hide to sell unreliable products.

Plenty of these already in use

[Anonymous Coward]

Childrens Medical Center Dallas has been using a system like this for a couple of years now. It is mainly used to pre-fill commonly needed meds like insulin, thereby reducing errors. It's kinda fascinating to watch a "machinegun ammo belt" of syringes run through it in short order. CMC-D also has a medication picker robot arm that vistors can watch through a window on the first floor by the trains.

Law & Order episode comes to mind

[Mr. Cancelled]

Anyone remember the episode wherein a teenage hacker hacks into a medial facility and changes their medical software so that people are given overdoses of insulin?

In the episode, the hacker was a teenager who was under the impression that the medical facility had blinded his father, and made the changes as a form of revenge.

In the real life version, I'm going to guess that we'll have people threatening to do something similar unless they're paid off.

Not that I'm against such changes. I just lost my Grandmother to a similar situation (someone gave her the wrong medicine as near as we can tell at this point), so any technology that can eliminate such errors, or help to reduce them, is welcomed by me and my family. I just think the Law & Order episode illustrates that no automated system's 100% foolproof. We still have to protect them from the script kiddies and such, but this is a huge step towards eliminating human errors, at least.

Re:Law & Order episode comes to mind

[TheGavster]

How hard is it, seriously, to not connect your internal network of life-critical computers to the internet?

Re:Law & Order episode comes to mind

[haakon]

If your manager isn't given the budget to build a completely seperate network, very.....

Reduce errors?

[sakusha]

Are you serious? One coding error could affect HUNDREDS of patients. Does nobody read RISKS Digest anymore? Coding failures in medical equipment has been under continual discussion on RISKS for many years.
All you're doing here is trading one risk for another, a risk that more people are taking on faith, since everyone KNOWS that computers are infallible. Yeah right.

Dangerous hospital robots

[HangingChad]

A group of German patients has filed a lawsuit against financially beleaguered Integrated Surgical Systems Inc., alleging that the Davis company' Robodoc surgical robot is defective and dangerous

Because in a little known incident one of their surgical robots went on a rampage, careening wildly down the hospital corridors wielding a variety of surgical scalpels while shouting, "YOU WANT A PIECE OF ME, MEAT SACK?!"

Re:Dangerous hospital robots

[Anonymous Coward]

the hospital bgu frankfurt, germany uses robodoc for hip operations.

those patients say robodoc works less accurately than a humandoc. those patients have chronic pain in their hips after op.

RoboPhobia

[stkpogo]

Fear of Surgical Robot with 'Made by RonCo' label,
it slices, it dices,,,
http://www.google.com/search?hl=en&q=RoboPhobia&bt nG=Google+Search

Re:Can't resist...

[Anonymous Coward]

I, for one, welcome our new robot pharmacist[ overlord]s.

And why shouldn't they be held libable?

[Anonymous Coward]

>Next thing you know, telecoms will be liable for medical malpractice if the network connections fail >during remote robotic surgery."

If the telecoms are selling these datalinks as being secure and reliable,why shouldn't they
be held accountable for them?

Sounds like you are the classic example of a Snake Oil Salesman.

You peddle something but don't want to be responsible for what you're selling...

QA

[jnjhoot]

I had a temp job working for this company as QA tech a few years ago and I can tell you right now that I would not want to be worked on by a machine that was QA'd by a $12/hr temp. The machine was pretty amazing, what was more amazing was the bunch of incompetents that engineered it.

Error checking in humans

[Toby The Economist]

> Next thing you know, telecoms will be liable for
> medical malpractice if the network connections
> fail during remote robotic surgery

As they bloody well should be!

If they undertake to provide a *guaranteed* connection, which they should be, given that's it SURGURY, if the connection fails, the patient is going to be in serious trouble.

The problem with automation of medical work like this is that it removes a level of error checking - the human who fills in the needles or prescriptions.

Humans pick up a good idea of what's normal and what's not pretty quickly, and do a good job of routine error checking.

Robots don't, not even a bit. And software, as we all know, is not reliable.

So, you get quicker service, and it's cheaper, but you eliminate a level of error checking.

--
Toby

Rise of the machines

[AbsurdProverb]

I certainly hope the Ohio medical machines perform better than the Ohio voting machines. There sure as hell better be a paper trail with this rig. I need to know who who will be recieving the malpractice subpoena.

Do you have any brains?

[andalay]

"Next thing you know, telecoms will be liable for medical malpractice if the network connections fail during remote robotic surgery."

Buddy, if I create a product specifically to do job A, and as a corollary, it can also do B,C,D (unintended uses), the customer cannot blame me if the product fails to work for B,C,D. However, they can blame me if the product is deficient wrt doing job A.

liability ....

[Anonymous Coward]

So how does one reconcile liability for robo-doc when robo-doc screws up becauses of a software error? Remember, in the good Ol' USA software vendors are NOT legally responsible for problems caused by their programs.

simple

[Anonymous Coward]

Track down those people who work on so-called "open source software" and sue THEM instead.

Checks and balances

[booch]

I was under the impression that one of the main functions of a pharmacist is to know most of the various drugs by appearance. That's why drugs come in all different shapes and colors. So the pharmacist doesn't accidentally give you the little blue diamond pill (Viagra) when you're supposed to be getting the big yellow pill, even if the pills are in incorrectly marked containers. (I appreciate the fact that Walgreen's now puts a label on your presicription bottle describing the pill, so you can verify.) Somehow I doubt that automated systems are able to do this yet. Although with injected medicines, I'd supposed they almost all look alike (clear fluid). I'd be willing to bet that the machine has a much lower error rate, but it'd still be nice to have a visual check.

Re:Checks and balances

[drinkypoo]

Computers absolutely can do this, and what's more they can do it more accurately than a human, because they can do it more consistently. While visual processing is definitely a road which has only been started upon, doing things like recognizing a simple shape and detecting a color is relatively trivial and a pretty well-known problem. You could even OCR the letters/numbers on the pills, although shape recognition alone would probably provide sufficient accuracy. As you point out, a visual check is not really useful for injectables. Presumably the machine will be loaded by a human (a human will be involved in the loading somewhere even if they invent a robot that loads the machine, too) and I'm frankly much more worried about that human making a mistake. However, putting bar codes or rfid tags on the bottles or OCR'ing them would essentially ensure that you could not screw that up either.

Veterans Administration is a leader in this area

[ewanrg]

For those of you in the US, you might be pleased to know that the Veterans Administration is a leader in this area. Virtually all of the VA Medical Centers use robot automation for filling pill bottles and preparing Bar Codes - and then use those same Bar Codes on the meds and Bar Codes printed on the patients' wrist band (for inpatient delivery) to make sure that the critical 5 Rights are validated:
Right patient
Right medication
Right route
Right time
and Right dose

The facility in Houston, TX also uses robots designed to ferry prescriptions to the various floors of the medical center. Which is being studied for possible use at other facilities.

Many of the VA centers also work with national Mail Order centers that use "extreme" automation to fill tens of thousands of prescriptions daily when a Veteran or other eligible person needs a refill.

Just another case study to consider if you're looking at the use of robotics and electronic records to improve health care and patient safety.

---

More of my personal stuff here

Similiar to the OS "life / death" clause

[lxt]

It's similar to the standard Windows / Mac OS clause of "This OS is not designed for life / death or mission critical operations"...as if you'd actually want to run Windows in one of those areas :)

I was once replaced by a robot at a hospital...

[Johnny Mozzarella]

Back when I was in college I scored an awesome job working at a hospital. The job paid really good and I worked the graveyard shift so there wasn't much work.

My job was to type up and apply labels to blood, urine and stool specimens when they arrived in the labs and then send them off to the appropriate lab for processing.

There was very little activity so I practiced sleeping sitting up at a computer terminal. Things went well at first but my laziness and irresponsibility started catching up with me. Being a college student I soon discovered the joys of binge drinking. This resulted in me going to work with a hangover on several occasions. As you can imagine hangovers + regular sleep deprivation = diminished quality control.

I was there for about a year before being replaced by a machine that could scan the forms, label the specimens and send them off without yours truly.

So what did I do? I took on a new position as a pizza distribution engineer for a multinational firm named Pizza Hut. Gotta love progress.

AI

[ZeroReality]

The Surgery machine that is mention is a simulated neural network, a learning computer of sorts. Most of the pervious mention errors were programming errors. A neural network is not programmed. It has to learn by practice and observing. A computer will not make mistake due to being tired in a several hour surgery. It even state in the article that it can be more precise. It can also learn to stop. People on the other hand can be stubborn and carry on. The reason it take longer is arm articulation is very difficult to do for a computer. People are born with this knowledge to start with.

The only draw back is the neural network has to have consistent human anatomy. The human body is very uniform for the most part.

As for the medicine dispenser, it can be simply given a set of governing rules. It can be told not to mix to specific compounds because they are lethal, or have drug interaction. Your right it is a bit of a nuisance in extreme emergency case but on a day to day consistent bases it just quicker. If it handing out pain medication and other long term treatment it just plain quicker.

good

[jeif1k]

So now with robotic surgery, both the doctor and the robot can liable for damages. Next thing you know, telecoms will be liable for medical malpractice if the network connections fail during remote robotic surgery.

Yes, and that's as it should be: if you bring a product or service to market and it causes harm because it doesn't work as promised, you should be responsible for the damages.

So, if your robot causes unnecessary harm to patients or if your high-availability comlink goes down too much, then you should have to pay.

A little perspective...

[Chembryl]

.. from someone who programs something similar.

I would say its more 'People just don't trust things they have no knowledge of' rather than machines per se. I certainly wouldn't blindly trust something programmed by myself or any of my colleagues without being able to verify the results independently myself.

In our case the systems we produce analyse the patient's blood or tissue and then provide prescription reports based on the results. The main selling point of our systems (or so I am told) is our ability to reduce or remove doctor/nurse/lab technician's transcription errors from what turn out to be rather complex clinical or medical processes.

Instead of automating the process entirely, our customers generally end up employing a more junior nurse/technician to 'watch' our instruments in case something appears to go wrong. Of course whilst mechincal failures and data corruptions rarely occur, labs and hospitals are wise enough to have their own validation (unit testing, if you will) and calibration methods in place for when they do.

Assumption of Liability

[dos4who]

"telecoms will be liable for medical malpractice if the network connections fail during remote robotic surgery."

This would never fly... It would be akin to a hospital trying to sue the city if an ambulance encounters roadwork delays while trying to transport a patient. Assumption of risk lies with the user of the system.

Do people?

[cosmicg]

The robots most certainly will (probably more often then you'd like), it's the people you have to worry about-- somewhere betwen 40,000 and 100,000 people die every year due to medical errors.
The classic text on this (if an article from 2000 can be considered classic), is To Err Is Human: Building a Safer Health System

Thats not the right fine print

[bluGill]

You are reading the wrong fine print. The fine print on your contract is not the same as what is on their contract. Their contract specifies that at all time where will be at least 2 redundant links in operations that never travel in the same trench. Their contract specifies that their traffic will always have the highest priority for transfer, and has some traverse time (lag) and bandwidth requirements.

They also pay for that level of service.

Synopsis...

[Anonymous Coward]

...you sucked at your job, and were replaced by a machine that did it competently.

Hmm, fairy tale ending.

Futurama in real life

[blauerdrachen]

"Yeah, I got your meds right here - you lumpy bag of mush..." - Bender

Re: Thank God. Bush to the Rescue

[Anonymous Coward]

Thank goodness our president has the foresight to repeal those laws which permit medical malpractice claims to be filed. These are clogging our courts and impeding technological progress and the promise of robotic medicine. We applaud the administration for limiting the need for extensive FDA testing of medical devices before they can be sold.

Our firm is working on a robotic prototype that can double both as a shotgun toting guard in Iraq and as a brain surgeon, and whose third arm is equipped with a Taser capable of incapacitating any lawyer who even thinks of filing a lawsuit. We are also experimenting with a four armed model that can dispense political contributions or oxycontin to political appointees within the administration and congress just as easily as it can help Indefinite Detainees speak freely. Given the current state of our technology, the availability of such dual use technology in a medical malpractice free/battlefield environment is critical to the economic success of our products.

We are encouraged by the president's foresight in expanding the potential markets for all our products.