Time on Task

There's a research paper making the rounds in the press and on social media, about Facebook usage in an emergency department.

It's called, "Online Social Network Use by Health Care Providers in a High Traffic Patient Care Environment" by Erik Black et al, and it's in JMIR, a popular informatics journal that I thought had a reputation for quality. I say this, because the methods were poorly described and the conclusions are grossly misleading.

Health policy wonks like Sarah Kliff made the following conclusion:
For every hour emergency department workers use a computer, they spend an average of 12 minutes on Facebook — and that time on the site actually goes up as the department becomes busier.
Researchers at the University of Florida monitored the workstations of 68 emergency department workstations for just over two weeks. They couldn’t see who was using each work station — it could have been a doctor, a nurse or another health-care worker — but they could see what they were looking at on the computer.
Over the 15-day period, the staff cumulatively visited 9,369 Facebook pages, spending an average of one of every five minutes of computer use on the site. Usage was much higher overnight. During the 7 p.m. to 7 a.m. shift, emergency department employees spent an average of 19.8 minutes per hour on Facebook. During the day, it came down to a much smaller 4.3 minutes.
It's just one academic ED, in Florida, over 15 days, more than three years ago - so, hardly representative of the specialty . And there's no way to tell if the usage is from doctors or nurses or techs or consultants from upstairs. And didn't we all look at FB more, back in 2009-2010, when this study was conducted?

But take a look at what the study results actually say:
In a 15-day period, health care workers spent an accumulated 4349 minutes (72.5 hours) browsing Facebook on workstations in one ED. ED staff cumulatively visited 9369 Facebook pages and spent, on average, 12.0 minutes per hour browsing Facebook.
This is not "using Facebook for 12 minutes per computer per hour" as Klift and others concluded. This is "using Facebook for 12 minutes per hour" total, across all of this large ED's staff and sixty-eight workstations.

If you broke usage down by computer, it's around 11 seconds per hour - which makes the author's opinion that FB usage is "unacceptably high" well, unacceptable.

Thanks to Jonathan Handler and the ACEP Informatics Section listserv for pointing out this article, and their analysis. UPDATE: Dr. Handler has a blog post on this topic up now, as well.

The politics of EHR implementation

I've pretty much kept my distance from the right-wing noise machine. They don't often talk about my professional interests, and when they do, it's not usually cited by a colleague or fellow academic.

But this morning, someone I respect shared this screed from Michelle Malkin on the great EDBA listserv - which I'd always equated with intelligent discussion of applied emergency medicine informatics.

So, let's dive into Malkin's piece on "Obama's crony," the CEO of Epic Systems:
The stimulus law provided a whopping $19 billion in “incentives” (read: subsidies) to force hospitals and medical professionals into converting from paper to electronic record-keeping systems.
I take issue with the past tense "provided" because these $19 billion will be allocated over many years, and only a small fraction has been given out already. And while $19 billion seems "whopping", healthcare spending was $2.7 trillion in 2011, and Medicare spending alone was $557 billion that year.
Obamacare bureaucrats claimed the government’s EMR mandate would save money and modernize health care.
This had nothing to do with "Obamacare" and in fact I don't think that term had been coined when the stimulus bill passed in February 2009. Lots of people thought, and still think, that EMR will save money in the long run (and moving from paper to electronic pretty much modernizes care, by definition). And of course, there's the reasonable expectation that patient care will be improved, too.
After hyping the alleged benefits for nearly a decade, the RAND Corporation finally admitted in January that its cost-savings predictions of $81 billion a year — used repeatedly to support the Obama EMR mandate — were, um, grossly overstated.
Among many factors, the researchers blamed “lack of interoperability” of records systems for the failure to bring down costs. And that is a funny thing, because it brings us right back to Faulkner and her well-connected company. You see, Epic Systems — the dominant EMR giant in America — is notorious for its lack of interoperability.
OK! Malkin has made a point that can't be dismissed out of hand: Epic systems do tend to be closed. There's lots to criticism about that system and the state of EHR in general. And yes, the 2005 Rand report (which, by the way, was funded by EHR vendors) estimated big savings and the "Obamacare bureaucrats" paid attention to it (though one has to wonder what Malkin would write if they had ignored the report.) Since that time, experts agree interoperability has limited the expected savings - but those many-billions a year are still anticipated soon .

And hey, allocating $19 billion over many years to generate annual savings greater than that - on a $550 billion dollar program - just doesn't seem so crazy to me. If additional savings are delayed a few years, well, it's still a reasonable investment, to say nothing of the other benefits from adopting EHR. Who knows? Medicare spending is already slowing, maybe EHR is involved? At least you can't argue: this part of the stimulus accomplished the goal of, well, stimulating economic activity (I've seen the construction activity at Epic HQ first hand).

But here's something to think about: If Epic runs the table and becomes a monopoly, as Malkin (and others) allege will happen, doesn't that render the issue of interoperability moot? Wouldn't that accelerate the cost savings? Seen in this light, Epic political influence doesn't just benefit their company, but the taxpayer as well. It seems like this is something Malkin would be rooting for, instead of decrying.

Of course, I don't think Malkin has thought through her argument - she's just stringing together half-truths to score points with her audience. Because on the topic of political influence, she notes:
Epic employees donated nearly $1 million to political parties and candidates between 1995 and 2012 — 82 percent of it to Democrats
Again, I think some perspective may help - averaged over 17 years, Epic employees gave less than $50,000 a year - in total - to the Democratic party. I wonder if this is the reason Republicans candidates lost so many presidential elections over this period. If $50k per year is all it takes to be an "Obama crony" then what does the $18 million Google spent in 2012 alone mean? What do you call the Koch brothers?
The shadow of tyranny and the stench of corruption are unmistakable.
Goodness. Well, we can agree something stinks. EHRs, and Epic in particular, are a subject worthy of debate, but Malkin's piece does nothing to advance understanding of policy or this industry.

(The last time I waded into a right-wing leaning discussion of electronic health records, over at the WSJ blog, commenters compared folks like me, who help implement and study EHR, to the Tuskegee researchers. Let's hope things have improved since then).