The industry is moving, albeit very slowly, into digitized medical records. Already we order our X-rays and CAT scans on DVDs to carry to the next doctor. The inch thick (or more) medical folder is being replaced by the before mentioned laptop.
This is progress, and long over due. When I was working on my Master’s in Comp. Sci. at the University of Denver, one of my favorite and most interesting classes was an Information Systems series taught by a professor from the Daniels School of Business. This man had been CEO and CIO of a dozen companies and he enlightened me to the value of an “end-to-end” information system.
That was my motivation during a project I worked on at InfoPrint Solutions my last four years there. We completely rebuilt our IS system using Oracle and Siebel to create a system that could track our products from design to build to sale to installation to service to end of life. The power of being able to track the progress of our products and slice and dice views from different perspectives is a powerful tool used by world-class companies such as Wal-Mart and Apple. This really gives the business executive decision support.
Thinking of my doctor manually graphing my blood tests or Linda’s Dad’s doctor tracking kidney function with a hand drawn chart made me want to cry out, “where’s your information systems support?” I had that very discussion with my own personal care physician and he told me that the current program he was using could not import or export to Microsoft Excel! I told him that was like buying a car that didn’t match the nozzle on the pump at the gasoline station. I mean, what is wrong with these people? Don’t they understand computer interoperation!!
Of course, there are many, many issues specific to computerizing medical records. There is federal legislation designed to guard patient privacy. However, like many if not most government regulations, HIPAA (Health Insurance Portability and Accountability Act of 1996) is often an obstacle to efficiency while providing about as much protection of our privacy as our fence along the southern border provides to illegal immigration. (Noticed that in all conversations with doctors and assistants the first thing they ask is your birthday – that’s to securely identify you and to protect you identity. I suppose it helps, but I’m not really sure how.
And in an area that I’m particularly interested in, the risks of technology, medical software is often sited first as some of the baddest examples of system failures. An X-ray therapy machine designed in Canada is famous for the many patients it killed before all the software bugs were removed.
Now we’re moving toward networked and shared, integrated and interoperating, end-to-end, highly secure and depended upon (even if not necessarily dependable) medical information systems. There have certainly been some great success stories in the risks arena regarding medical software advances. For example, using bar codes to dispense medications in hospitals has greatly reduced human error and literally saved lives while, at the same time, improving efficiency and lowering costs. Those are the stories we love to hear.
But, as any expert on risks will explain, it is the human factor and the attitudes and psychological aspects of fully integrated systems that are often the forgotten requirement in system design. Here’s a little cautionary tale that should serve to warn us of the brave new world waiting for us in 1984 – err, I mean 2014 – just a little error there folks, nothing to worry about!
This story comes from British Columbia and shows a weak side of Canada’s single payer health care system which, naturally, includes an end-to-end information system. It is really a funny story, which is why I’ve preserved the original quotes from the characters. But it is a cautionary tale too. The failure isn’t exactly technical – really more of a human error story. Enjoy.
A man in Prince George, British Columbia, Canada was forced to convince his doctor and his girlfriend that he doesn’t have cancer after a mixup with the province’s pharmacy system confuse him with his dog. I turned out that Rick Gillingham does not have cancer, but his dog, Cooper, was taking the medication phenobarbital for canine epilepsy.
When Gillingham went to the local university hospital for a simple painkiller, the doctor started asking him questions about his cancer. “I told him, ‘I don’t have cancer,’ and he kept telling me not to be coy, that nobody was within earshot, so it was all right to talk about it, and he needed to know,” said Gillingham.
Gillingham’s girlfriend overheard the conversation from the waiting room, prompting her to storm in and demand to know why she was kept out of the loop, too. “She was saying things like, ‘They are professional, they don’t make these kind of mistakes.’ And I really didn’t know what to tell her. I was at a loss for words.”
As Gillingham attempted to convince his girlfriend that he wasn’t sick, the doctor finally revealed the source of the confusion. “As they were arguing the physician piped up and said, ‘Well if you’re not the one taking phenobarbital, who is?’ As soon as he said that the light went on for her. It was for the dog, not for me. I didn’t even remember the name of the dog’s stuff, but she did and it all clicked.”
Cooper’s veterinarian had prescribed phenobarbital to ease the animal’s canine epilepsy. But when the dog’s medication was entered into the province-wide PharmaNet system, there was nothing to indicate that Gillingham and his dog were not the same person – or even the same species. Adding to the confusion, the vet’s name is identical to a well-known cancer specialist.
“It would probably not have been harmful (if Gillingham took the dog’s drug), but if someone else had this happen, it could be life-threatening,” said his girlfriend. “There should be some way to knowing the difference at a glance between human medications and animal meds. This is scary stuff.”
Now, here are the lessons I learn from this somewhat humorous tale. 1) The doctor trusted the system despite protests. 2) The doctor said the Gillingham and he could not be overheard, but they were. 3) The girlfriend trusted the system. (Or maybe she just didn’t trust her boyfriend!) 4) There was apparently no way to distinguish Gillingham and Cooper. 5) There are two very different kinds of prescribers with the same name.
Now, from a lessons learned perspective, how would you change the requirements and specifications of the PharmaNet system to prevent a reoccurrence of this circumstance? Or would you change the training for the doctor? Or “training” for the patient? What about the design of the facilities and the location of the waiting room. (Have you noticed the little “stand behind this line until called” designs of medical intake facilities?)
Remember, we’re talking about “end-to-end” systems. Everything is part of the system including the staff, the patients, their significant others, even the building. And you thought it was all about C++ programming and compilers! That is the point I learned: end-to-end includes it all, especially the people.
I am a big, big, big fan of technology. I worked with it my entire adult life. But I know how susceptible technology is to error. We used to joke that computers have no common sense. Sadly, neither do many humans. Meanwhile, check that prescription you just took. Especially if it is giving you an urge to scratch behind your ear and chase cars.
No comments:
Post a Comment