Update the next afternoon: be sure to read the comments, with important updates as the conversation continues.
Bulletin – I just learned about this tonight:
Last Friday the board of the American Medical Informatics Association (AMIA) published a position paper in its journal (abstract here) saying that the “hold harmless” clause is unethical. One of the paper’s authors is Dr. Danny Sands, currently President of our Society.
I hope to write more about it this week, after attending the AMIA conference in DC, but here’s the basic issue:
- For ages, makers of electronic medical record systems (EMR) have insisted on a “hold harmless” clause in the contracts a system buyer must sign. It says, in essence, that if any harm comes to anyone because of a system problem, the buyer (the hospital) will hold the manufacturer harmless.
- In other words, if anything goes wrong with the system and someone gets hurt, it’s not the manufacturer’s fault. The reasoning has been “Hey, you doctors are smart – if our system displays a wrong value, you’re supposed to notice it.”
I’m told this policy has been one big impediment to adoption of EMR systems, because it removes all motivation for vendors to fix things that make their product hard to use: if there’s a bug or the system slows someone down, and a patient gets hurt, the hospital gets sued, not the vendor.
If you were a hospital, wouldn’t that make you eager to buy? What would that do to your trust of the vendor?
Patients, how do you feel about that? Providers?
This came up repeatedly during the Meaningful Use meetings I attended last winter; I wrote about it in “What to do about the cream of the crap?” … those strong words were how an unhappy buyer described the system they’d just chosen.
That post links to a video of a lecture last December by Ross Koppel of the U of Pennsylvania, another author of the AMIA paper. Trust me, if you aren’t yet familiar with how ghastly some EMR systems are, this 30 minute lecture (and 30 Min Q&A) will take you to school quick.
The hold harmless clause has also been an immense impediment to improving the quality of software. Imagine if no matter how badly you did your job, you were legally not responsible. Better, think about that schlumpf down the hall – imagine if s/he was not responsible for errors in his/her work.
Last spring as the Meaningful Use rules were being written, there was talk of mandating that vendors have to make their systems usable for providers in a harried hospital environment. According to rumor, an industry executive said “Over my dead body.” Very funny, making my nurses and doctors use crappy, buggy software. So when I spoke in June at the Agency for Healthcare Research & Quality (AHRQ), I titled my talk “Over My Dead Body: Why Reliable Systems Matter to Patients.”
Thanks to the AMIA board for standing up for quality, and thanks to Dr. Sands, Ross Koppel, and the others who wrote the paper. It doesn’t have the force of law, but let’s hope it moves things in the right direction – for the benefit of the patients – and yes, I mean your mother, your child, your best friend. Our providers need tools that serve them well in serving us.
HOORAY!!!! I have seen enough crappy software to last a lifetime. I am no longer able to work, which puts me firmly in the patient category. It’s no better from the other side of the bed; it’s a nightmare both ways.
But worse still: Try being the EDUCATOR responsible for teaching the system. You won’t have a job if you don’t hype the software as if it’s God’s gift, even if it’s dangerous to the patient & the nurse. It is very difficult to keep quiet about something you know is garbage.
If you want support for your position, you might be able to get it from hospital educators & preceptors/mentors who have to deal with teaching new hires to use terrible systems – the worst of which are the ones with low IT maintenance costs relative to others.
Usability is inversely proportional to the cost of IT maintenance.
Can we get a full text copy of this position paper, about an issue that concerns all of us?
It is truly impossible to make sense of this complex issue by just reading the short abstract. Just as no one should ever make a medical decision based solely on an abstract.
Here is a link to the full document
The situation is indeed not as simple as it may sound. First of all, the “hold harmless” clause is not as widespread as people think. The limited liability clause is the major culprit, since the “limits” are very low. There are other contractual terms that I find more disturbing.
What people need to understand is that software does not come without bugs. Never. The constructive thing to do is to have commitment from the vendor regarding time frames for addressing bugs. That’s the part that, IMHO, is the most important, and often missing, piece.
It is easy and fashionable to demonize EHR vendors, but please remember that nobody wants to be involved in frivolous malpractice suits, particularly large corporations that make very easy targets.
As to Ethics education, that is, how should I say this, a bit naive.
For a more detailed analysis of this paper and the issue in general, see this post:
Hi Margalit – as always your counsel is most welcome. I just arrived in DC for the AMIA conference – will soak up what I can about people’s opinions on all this, and will read your post tonight. Thanks!
I should add that as someone who’s been involved with software all my life, I’m well aware of practical issues. Otoh, Ross Koppel’s lecture was astounding to me: he describes a case of a system that’s coded so archaically that warning messages can disappear into the background color. That kind of coding practice was supposed to have expired in the 1980s.
I’m no expert in this field but that was a giant red flag to me. Same for sloppy, silo’d UI issues like a datum that’s entered in one unit on one screen and displayed (unitless) on another. That’s just plain obsolete.
Looking forward to learning more – thanks for coming over here!
Thanks again to Margalit for the PDF of the full article. When I tried to get it last night it asked for a login.
I just skimmed the 5-page position paper. At first blush I’m pleased with the depth and the transparent presentation of rationale. For instance:
The paper lists 21 specific findings recommendations and several suggested next steps. Gut reaction: I’m pleased that it has a pragmatic tone, rooted in sensible patient-centered values.
I’m looking forward to reading the whole thing tonight …
” I’m told this policy has been one big impediment to adoption of EMR system”
The largest EHR vendor in the county will have 1 in 4 providers using its system by the end of their current roll-outs and all of the major hospitals in Seattle have already implemented an EHR as well as most of the major clinics. I assume that is true in Boston as well?
So I don’t actually see that clause has played any role in implementing and EHR. People will find an excuse to not do something.
Medical liability defense attorney Chad Brouillard (whose recent article I’ve cited on my blog) agreed with my take that HIT “hold harmless” clauses are unlikely to withstand the first serious court challenge.
It seems to me that the ‘hold harmless’ clause would be of limited legal use. Health organizations are unlikely to sue EMT providers because EMT software errors resulted in their making mistakes that harmed patients and harmed patient will sue the people/institutions that caused them harm.
The real utility of the ‘hold harmless’ clause is that it impedes any push from users to improve software once it is sold.