I grew up reading The Boston Globe.
Mostly, the sports section, but also arts and entertainment, news, and metro. I liked the magazine too. By the time I got to high school, I would read that paper as much as I could every morning.
My interest in the paper led to my interest in journalism and eventually, my career. I owe a lot to The Boston Globe. Without it, I wouldn’t be on the path I am today. Even though I moved to New York City more than four years ago, I still consider myself a Bostonian and that’s why I still read The Globe to this day.
Yet, even the great ones get it wrong. Last year, HCI Editor-in-Chief called out The New York Times, perhaps the greatest paper of them all, for a shoddy report on meaningful use, electronic health records (EHRs), and the entire health IT industry. One year later, I’m afraid I have to do the same thing to The Globe.
This week, I was back in Boston for a friend’s wedding and my mom told me about an electronic medical record (EMR) story that made it to the front page of The Globe. Naturally, I had to take a look.
The hook in the story was a 46-year-old woman who died in 2010 because of a medication error in the EMR. The article paints EMRs as “complex, balky, unwieldy, and error-prone computer systems.” It calls out the government for not requiring mandatory reporting on injuries, deaths, and unsafe conditions from EMRs.
First and foremost, what happened to that woman is tragic and terrible. I feel for her family.
However, the blame seems to be a little misplaced. If you read the article, it appears that her demise came from two orders for insulin – one in an EMR, one involving a paper chart and fax – under the names of different doctors. The lawyer for the woman’s family is blaming the systems for not interacting with each other.
Just so we’re clear, the lawyer says the EMR should be interoperable with a paper chart and that’s why she died? It wasn’t a documentation error? Maybe I’m reading this wrong but it seems odd to blame an EMR for not being able to interact with a paper chart.
EMRs aren’t perfect. Interoperability is a huge issue as are medication errors, adverse events and alarm fatigue. No one would disagree that we have a long way to go in EHR development. But this article takes this as its thesis and shoves it down your throat without any regard for balance.
It uses the results of a study to hammer home its point. The Harvard-affiliated medical malpractice insurer CRICO looked through malpractice claims and found 147 related to EHR-based adverse events, a chunk of which led to patient deaths. Unfortunately, we don’t find out how many claims they filtered to get that number. It also doesn’t include the results of other studies, such as this one posted on the American Medical Association’s website from the Pennsylvania Patient Safety Authority, which revealed EHR-based adverse events are not as dramatic of a problem.
The Pennsylvania Patient Safety Authority studied 1,249 events that occurred over nine years from 2004 to 2013. Of those, they found that 324 could be tracked to the EHR. Of those 324 events, 97 percent did not result in an adverse event. Of the 1,249 events that occurred over nine years, six came resulted in harm as result of the EHR. That’s six too many, but hardly an eye-raising problem.
The research shows that most EHR-related adverse events come from entry mistakes. The article argues that this is because the systems are “complex, time-draining, and frustrating to use.” That may be the case but it kind of conveniently leaves people off the hook for their own mistakes. We’re talking about entering numbers and data into a computer program not flying a plane (which is a comparison used in the article).
The article also ignores the positive work EHRs have had on reducing medical errors, as noted by Sen. Barbara Boxer (D-Calif.) and the Boston Children's Hospital. Or what about Lakeland HealthCare in St. Joseph, Mich., which boasted of decreased medication errors after its EHR implementation? If you look hard enough, there are plenty of examples where EHRs weren’t part of the problem, but the solution. I’d reckon to guess that when it comes to medication errors and EHRs, there’s more positive than negative.
For what it’s worth, I think mandated reporting of adverse events – as promoted by Sen. Boxer – is a good idea. As her office noted, this would help researchers identify where the problems are and how they can be fixed. I just don’t think we should act as if this is some kind of “20/20, dramatic music in to the interlude” problem. You know, the kind of problem that has your mother-in-law not eating tomatoes for a month because of an E. coli problem. I can hear the voice now:
They were designed to help patients but are electronic health records killing them instead? Evil computers, tonight at 10.