Pharma & Healthcare 8/26/2014 @ 1:23PM
Remember that Boston Globe story from July that questioned the safety and efficacy of electronic health records? If you haven’t seen it, I recommend that you check it out, but take a grain of salt with it, since the story neglects to say that the savings and quality gains promised with the $27 billion Meaningful Use EHR incentive program aren’t supposed to be realized until Stage 3, which won’t start before 2017.
One of the people quoted in that article was University of Pennsylvania sociologist Ross Koppel, someone who has long been critical of what he sees as complacency among major health IT vendors. (Essential reading on Koppel is his 2005 report in the Journal of the American Medical Association demonstrating 22 different ways that computerized physician order entry systems can cause or exacerbate medication errors.)
I had a chance to talk to Koppel recently for an article in another publication that should appear shortly after Labor Day. I left out a long discussion we had about a public hearing held while federal officials were in the early stages of development for Meaningful Use Stage 2 requirements, but I’m going to share that here.
Koppel recalled an April 2011 hearing on EHR usability in front of the Health IT Policy Committee‘s workgroup on certification and adoption, a workgroup that included several vendor executives. He was on a panel with University of Maryland computer scientist Ben Shneiderman, an expert in user interfaces.
Shneiderman scoffed at testimony that suggested common EHR user interfaces would be a bad idea. According to a transcript of the meeting, “My greatest concern is the [closed-mindedness] of this industry to discussion.”
Of particular concern is the lack of standardization of terminology, something that remains a problem today. Shneiderman related the story about a physician who got a report saying that a patient had Congolese hemorrhagic fever (sometimes called Crimean-Congo hemorrhagic fever). But the patient didn’t really have that deadly disease. The diagnosis had been improperly coded by someone who didn’t recognize the abbreviation “CHF,” which also stands for congestive heart failure, a common, chronic, treatable condition.
“The change in values, in metrics, and the lack of terminology standardization are really problematic issues,” Shneiderman said. “We’re dealing with medical reconciliation, the names of drugs, the spelling of them and so on. There‘s so much variability because everybody wants their independence. The nature of computerization in almost every field is to come together to make some common ideas, some concepts, workflows and terminology so that people can work together, so that learning times are shortened, so that collaboration that was so important here is facilitated. ”
According to Koppel, vendor representatives essentially argued that usability was not measurable because it was too theoretical.
Shneiderman criticized panel member Carl Dvorak, president and COO of EHR vendor Epic Systems Corp. “
Carl Dvorak, he celebrates the diversity and the product differentiation of competing products, but they’re differentiating on the wrong metrics. If they differentiated on the support of creative diagnostics, I’d be very impressed. But to differentiate by the color of the fonts and last name/first name versus first name/last name and the 35 ways that systolic and diastolic pressure are shown is not creativity. It’s not freedom. It’s just chaos,” Shneiderman said.
Vendors voluntarily report findings on adverse events to the Food and Drug Administration’s Manufacturer and User Facility Device Experience (MAUDE) reporting system. While the reports are made public, they are made anonymously, which frustrates Koppel. “It’s completely voluntary, but completely secret,” Koppel said.
For their part, vendors said insurance companies were partly to blame because they make it hard to access key cost data.
” For example, we heard complaints [at the hearing] about too many clicks to order a mammogram. That may have been bad software design, but it may have been that in that particular case, the payer was putting up a hurdle in front of ordering the mammogram by asking for additional data with an overt intention of discouraging the use of what might have been judged to be an inappropriate test,” Dr. David McCallie, vice president of medical informatics at Cerner CERN -0.43% Corp., said at that 2011 hearing.
“We tried to design an e-prescribing system a few years ago, and one of my goals to the team was to surface to actual cost to the patient out of pocket of the medicines so that the physician could give the patient choices between the various—once a day is going to cost you more than two times a day. We couldn’t get that data; it’s hidden from us. The payers wouldn’t let us get it, so we had to build cumbersomeness into the user interface that, in fact, reflected external constraints,” McCallie said.
There has been more talk lately of price transparency, and more collaboration between payers and providers on making cost data more accessible, but the three-way blame game among payers, providers and vendors continues to this day. Among the losers are taxpayers who are ponying up billions for health IT that so far has been more helpful than harmful, but hardly ideal. But the real losers here are patients, whose safety continues to be at risk.