CT Scan "Reset Error" Gives 206 Patients Radiation Overdose

jeffb (2.718) writes "As the LA Times reports, 206 patients receiving CT scans at Cedar Sinai hospital received up to eight times the X-ray exposure doctors intended. (The FDA alert gives details about the doses involved.) A misunderstanding over an 'embedded default setting' appears to have led to the error, which occurred when the hospital 'began using a new protocol for a specialized type of scan used to diagnose strokes. Doctors believed it would provide them more useful data to analyze disruptions in the flow of blood to brain tissue.' Human-computer interaction classes from the late 1980s onward have pounded home the lesson of the Therac-25, the usability issues of which led to multiple deaths. Will we ever learn enough to make these errors truly uncommittable?"

Comments

Not the engineers fault

[PhasmatisApparatus]

Requiring that doctors RTFM is the first step.

Re:Not the engineers fault

[betterunixthanunix]

The machine's software should not be capable of triggering the release of that much radiation; any change in the radiation levels should require some kind of hardware interaction. Even an idiot who did not RTFM should not be able to cause harm with the machine.

Re:Not the engineers fault

[smitty777]

Couldn't disagree more. Unfortunately, enforcing training and reading manuals would probably have little effect. In my 10+ years doing usability for missile systems, you have to build in the mechanisms to keep the users from doing bad things. Even if you force the user to read the *entire manual* before each use, people still have bad days, hangovers, fights with significant others. It has to be designed in.

Re:Not the engineers fault

[Greyfox]

My machine would irradiate the operators by default and would require that a obscure button sequence be pushed in order to irradiate the patient instead. That way the idiot who didn't RTFM would end up dying of radiation poisoning, not the patient. Eventually the survivors who DID RTFM would breed and pass on their proclivity to RTFM. Really it's for the good of the entire human race, if you think about it...

Re:Not the engineers fault

[BigDukeSix]

It's not quite that simple. The CT scanner is set up with a distinct scanning protocol for whatever part of the body you're imaging. If you're trying to get a detailed image of the bones of the pelvis you have to use more power than if you're imaging the lungs. The scan is further individualized by patient size. Given that infants and very large people are imaged on the same scanner, the software has to vary radiation dose over a reasonably wide range, and it's a different setting for every scan.

Re:Not the engineers fault

[RobertLTux]

I see you are about to fry this patient like an egg (doseage set for multiples of normal protocol)
would you like me to

1 reset the machine to standard defaults
2 book you a flight to africa
3 call your lawyer now
4 forge the documents to show %person% did the treatment

or
You Are about to kill this patient [cancel] or [allow]

Re:Not the engineers fault

[lysergic.acid]

Thank god you're not responsible for the design of complex, life-critical systems, like those commonly found on passenger jets, in nuclear power plants, in high-speed rail systems, etc. All of those systems incorporate fail-safe measures so that if something were to go wrong (like an operator losing control) the system would fallback on a safe state.

Sure, in an ideal world, every operator of a life-critical system would have total understanding of that system, know the value of every system setting at all times, never forget, never be tired, and have an IQ of 200. In the real world, operators are often overworked, susceptible to distractions, minimally qualified, and sometimes under-trained or even improperly trained. Even the most experienced and well-trained veteran airline pilots can lose focus and make deadly mistakes (which is why Cockpit Resource Management has been a major area of research in aviation psychology). You can base your system design on ideal conditions, or you can base it off of real-world conditions; either way, it's going to be operating in the later.

You also seem to be missing the main purpose of mechanization and automation, which is to simplify a task or make it easier to perform. When you buy a cappuccino machine, you don't want to understand the details of how it operates or be asked for input every step of the process to make a cup of coffee. Eliminating/minimizing the human factor in a particular process is another major advantage of automation. It provides more consistent results and helps to minimize human error. All of this helps to reduce the learning curve and skill level required to perform a task, which confers economic benefits. However, not every well-designed system can necessarily be operated by unskilled personnel—nor would you want a high school drop out to be operating most life-critical systems. Nonetheless, you still want mechanization/automation to simplify the task in these cases. That's because some tasks are so inherently complex and mentally demanding that, without automation, it simply can't be performed.

Flying a passenger jet is a perfect example of this. Even with all the sophisticated automation (including autopilot) on a modern airliner, it still takes a full cockpit crew (not to mention support personnel on the ground) to safely fly & land the plane. With all of the complex duties that airline pilots need to perform simultaneously, they don't have the time to monitor the status of every system component or manually adjust every actuator on the plane to control its flight surfaces. It may take 50 different mechanical actions to retract the landing gear on a plane, but why clutter the cockpit interface with 50 items when a single switch or button will do? Likewise, doctors and nurses are already required to undergo extensive medical training; they don't need to have to learn how to mechanically calibrate a CO2 laser or calculate the spectrum of an X-ray machine based on the anode material of its emitter and the voltage passed through it. Medical personnel should mainly be trained in medicine and only need to learn how to operate a particular medical device, not how to troubleshoot it or read its blueprints.

A simple and streamlined interface is much less distracting and more intuitive than a field of buttons and dials for a thousand different minute settings and system readings. Even with the utmost simplification, most industrial machinery and complex systems are still overwhelmingly difficult to operate by an untrained person. It's never just a single "magic button" for the operator to press. A nuclear power plant might take hundreds of different readings from multiple sensors and summarize it with a single status message or indicator light on a controller's console, but that message/light would likely be sitting next to a dozen other status indicators that each take hundreds of other readings. And although a complex process like lowering the reactor temperature might be simplified down to a single "magic button," the c

Default setting...

[courteaudotbiz]

The default setting for an equipment that can be lethal should be "Emit zero radiation". Then for each exposure, set the level of radiation you intend to use. This way, you ALWAYS KNOW the level of radiation the equipment will emit.

Better investigate "Hey, we got no picture" than "Hey, we got pictures, but everyone dies after that..."

Didn't RTFA.

HULK MAD!

[BumbaCLot]

Even under normal circumstances, the procedure requires more radiation than most other types of CT scans, said David Brenner, director of radiological research at Columbia University Medical Center in New York.

Anyone else read this as David Banner?

Re:Maybe testing it afterwards?

[RDW]

'How hard would it have been to stick a dosimeter in the machine after the change and run it though a test'

Supposedly the actual dose would have been displayed on the machine's screen (I wonder how prominently?):

http://www.latimes.com/news/local/la-me-cedars-sinai14-2009oct14,0,5065886.story

'"It's in your face on the screen," said Dr. Donald Rucker, chief medical officer for Siemens, a manufacturer of CT scanners.'

'CT technicians are trained to monitor dose levels, and some hospitals conduct checks before every scan..."There are other places where the techs might be operating more as button-pushers," said Dr. Geoffrey Rubin, a professor of radiology at Stanford University. "The user becomes a little blind to these numbers."'

Comment removed

[account_deleted]

Comment removed based on user account deletion

Re:It's About Automation

[antifoidulus]

The advantages of simplified training are not just beneficial on an economic scale. While its unfortunate that this error killed people, think of how many more people would die if complex training was required to use these types of machines. Ultimately, it would lead to fewer operators and thus less access to the machine, which ostensibly helps save lives.