Medical Equipment Design

Not exactly the first instance.

http://en.wikipedia.org/wiki/Therac-25

Who is John Galt is exactly right. Who do we blame for the way the world is. Redfred is exactly right, the more safe we design things the more creative the idiots get. I dont know if Ayn Rand got it right in Atlas Shrugged, but what would happen if all the smart dedicated people just up and quit. I am more of a Dagny Taggart, I will stay at my post trying to save the world we are in as long as I can.

The only true solution I can see is education, too bad so many Americans who can have it merely neglect it and are satisfied living the life of Trailer Trash. Perhaps we do need some patients over-radiated to clean up the gene pool, just make sure they are offering free CT scans to the right people.

Drew

the distinction between an idiot and a genius is

the a genius has boundaries

Some excellent discussion here:

http://cr4.globalspec.com/thread/44809

milo

Add a knob which requires setting the dosage before each scan. If dosage is not set, the machine won't work.

Bill

They'll stick a tab off a pop can or a piece of tape to defeat that. Done it myself on conveyor belt interlocks...

milo

The problem is that no matter how idiot proof you make a device, there will be some idiot who will figure out how to defeat it, and if you automate the system too much, then you breed complacency into the operator because they come to rely on the machine making all the decisions for them and they stop checking to make sure it really is doing what they think it should be doing.

Is there a solution? I'm not sure there is. if you bury all the complexity an interlocks into the system then the operator will either be cut out of the loop completely or will not understand when there is a problem. If you leave it to a skilled operator to ensure everything is running correctly, then you are relying on the human to perform flawlessly every time.

Isn't this what the real issue is in the story posted?

I think you've clobbered it!

milo

My idea is that when the scan is done, the machine automatically sets dosage to 0. The knob I mentioned is connected to a many spoked wheel used to pulse an LED/photodetector setup. The machine increases the dosage depending on how many pulses it receives.

Have to talk to the wife about this. She is the medical imaging specialist in the family.

Bill

and what happens when the connector to the photodiode/transistor is faulty and it generates a bunch of noise as you turn the knob generating a whole bunch of extra pulses? or a pullup resistor fails open and the shift register counting the pulses outputs the wrong number?

Therein lies the danger my friend. Nothing is or can be fool proof, and the more things you throw into the mix, the more things there are to go wrong. In fact the mathematical relationship is exponential. two failure points quadruple the likelihood of failure, three increase the likelihood by a factor of nine. and that is assuming that the failures don't interact and cascade. Many times failures are cascade failures where several minor failures, each in of themselves not terribly serious, interact and combine in unexpected ways resulting in a massive cascade failure. There have been several famous (or infamous) failures of this sort, I can think of three in the space launch business off the top of my head that resulted in total loss of the crew and/or spacecraft, and one that came mighty close.

Yes! and for each one of us trying to make things foolproof, there are at least two people out there making more fools.

My wife says that there is no way that their CAT scans can generate 20 times the proper level of X-rays... but as we know, Murphy lives.

Excellent example of "progress" your wife might find interesting. I installed a digital x-ray system for my dentist that uses a CCD sensor with a image intensifier/sinctillator array on top of it. the power output from the x-ray head (and therefore the patient dosage) was reduced by a factor of 10.

My doctor's office installed a "digital x-ray system" that used re-usable phosphor plates that were scanned and then reset. the dosage had to be DOUBLED to get a usable image compared to conventional film. Yes it saves storage space and cost, but at the expense of doubling the radiation dosage patients receive. So much for 'first do no harm".

You have to watch the medical system. There is a lot of this kind of stuff going on.

take drug research:

In one study conducted by the FDA itself, it was discovered that one in every five doctors investigated--doctors researching the effects of new drugs--had invented the data they reported and pocketed the fees. These are not unusual of isolated cases. John Braithwaite, a criminologist at the Australian Institute of Criminology (and also former Commissioner of Trade Practices in Australia), states: "The problem is that most fraud in clinical trials is unlikely to even be detected. Most cases which do come to public attention only do so because of extraordinary carlessness by the criminal physician."

According to Dr. Judith Jones, former Director of the Division of drug eExperience at the FDA, if a research facility obtains results that do not demonstrate the safety or effectiveness of a drug, it is not uncommon for the drug company to bury the report and continue testing elsewhere until they find a facility that tives them the results they want. Unfavorable reports are rearely published, and clinicians are pressured into keepin quiet about them.

very sad but true