Hard agree, and not just because I'm a spreadsheet nerd. Why are we all here? Isn't it in the service…
Readers Write 2/1/10
Submit your article of up to 500 words in length, subject to editing for clarity and brevity (please note: I run only original articles that have not appeared on any Web site or in any publication and I can’t use anything that looks like a commercial pitch). I’ll use a phony name for you unless you tell me otherwise. Thanks for sharing!
Virtual Medical Devices and Vendor Liabilities
By Scot M. Silverstein, MD
I found the Jan. 20, 2010 news release on Sen. Grassley’s latest health IT industry inquiry, Sen. Grassley asks hospitals about experiences with federal health information technology program, quite interesting. Based upon direct experience in hospital IT and in Big Pharma, I believe the inquiry justified and potentially beneficial to ensure proper accountability for taxpayer money and for patient safety.
One statement in particular caught my attention. In question #9: the Senator’s letter of inquiry states “… for example, one vendor stated that it is accountable for the performance of its [clinical information technology] product as long as the client uses the product appropriately. Another vendor stated that it is not liable when harm or loss results from the client’s use of the product in diagnosing and/or treating patients.”
Denial of inclusion of explicit "hold harmless" provisions on the one hand, and statements about being "accountable for the performance of its HIT product as long as the client uses the product appropriately" and "not [being] liable when harm or loss results from the client’s use of the product in diagnosing and/or treating patients" on the other hand, are at odds.
An EMR and other clinical IT systems are virtual medical devices, that is, medical devices that happen to reside on a computer.
The first part of the quoted statement above dismisses virtual medical devices that don’t “use the client” properly, for example, through presenting a mission-hostile user experience. Human computer interaction is a science, and its quality has major effects on results. For example, the Air Force — for obvious reasons — has been concerned with HCI and long ago, even before the GUI, wrote a treatise on its importance in mission critical settings (“GUIDELINES FOR DESIGNING USER INTERFACE SOFTWARE”, August 1986, http://hcibib.org/sam/).
Some vendors have opined that they are not liable when their products are used outside their intended purpose. I am not sure how an EMR can be used for anything else other than the "intended purpose" — that is, providing information and providing some type of actionable advice or recommendation based on the information (for example, decision support). On the one hand, if non-validated changes are made by the customer that cause malfunction, a vendor is clearly not liable. On the other hand, if inherent design problems (such as in the display of information) confuse or misinform the user, these vendors seem to be excusing themselves from liability on the grounds that the user was “using the product inappropriately.”
This position will not sit well with clinicians.
On the second issue about use in diagnosing and/or treating patients, vendors may feel they are “not in the business” of diagnosing and treating patients, therefore they are not liable when their products are used for such purposes.
They may have been correct two decades ago, but are in error today. Yes, HIT vendors are in the business of diagnosing and treating patients; in fact they are in the business of practicing medicine — via machine proxy.
HIT vendors are in the business of practicing medicine in the same way my medical supervisors (when I was a trainee) or medical consultants (when I practiced) were in that business when I presented a case to them and they made an assessment and treatment recommendations on what I should do.
When errors in presentation of information or errors in advice-giving occur, a "learned intermediary" defense does not absolve the consultant unless the error is so gross as to be implausible (e.g., a male with toxemia of pregnancy, or a serum creatinine rising from 1.0 to 10.0 in 24 hours). A physician should not be expected, on the other hand, to be able to with 100% reliability detect when a computer — or consultant — is lying when they provide a falsely low INR blood-thinning value or false diagnoses of another patient.
To emphasize these points further, a small thought experiment is in order:
1. If I set up a shop where I was allowing patients to bring me records and I simply looked at the records and dispensed advice on what evaluations or treatments I thought they needed, or what was wrong with their current regimen, but did not go further than that (just being a consultant), would I be practicing medicine? (The answer is yes; I actually held this role in the performance of independent medical evaluations and required both licensure and malpractice insurance in my state to perform this function.)
2. If I were to stand in the corner of the clinic where the computer terminal is, and take its place, i.e., offering data and advice on evaluation and treatment (perhaps Maelzel’s Chess Player-style) in place of the machine, would I be practicing medicine? (The answer is yes.)
3. If the advice I offer is based simply on the advice of another, or on following to the letter the rules and algorithms provided by an HIT vendor’s code, becoming a human computer, so to speak — am I not an intermediary to the other person’s, or the HIT company’s collective practice of medicine? (The answer is yes.)
I believe today’s HIT products involve the practice of medicine by machine proxy.
Two additional points:
I’m reasonably certain the vendor(s) who claim "not being liable [for the performance of their product] when harm or loss results from the client’s use of the product in diagnosing and/or treating patients" also make claims in their P.R. or in executive speeches about how their products will transform/revolutionize medicine.
Claims of transforming or revolutionizing medicine via their products are irreconcilable to any reasonable person with their stated claim of non-liability for use of the products in diagnosing or treating patients. Unless, that is, these systems are intended only for playing games:
Furthermore — and this is perhaps the more important point — hospital executives have both a fiduciary responsibility to patients and practitioners, as well as Joint Commission safety standard obligations.
Would the executives be performing these duties in good faith if, for example, they sign a contract for a CT scanner or surgical equipment where the vendor disclaims liability when these tools are used in diagnosing and/or treating patients? I believe the very act of signing such contracts is a breach of healthcare executive responsibilities and obligations, at the very least by shifting an undue legal burden onto their own clinical staff and contractors, as well as potentially putting patients at risk with tools that vendors are not as highly motivated as they should be to make robust.
Imagine a pharmaceutical company saying "we are not liable for the performance of our products when used in the treatment of patients." Would you use their products?
Health IT should be no different, as I pointed out in my JAMA letter to the editor of July 22 2009 at http://jama.ama-assn.org/cgi/content/extract/302/4/382.
Finally, it seems a vendor claiming they are not liable when harm or loss results from the client’s use of the product in diagnosing and/or treating patients should put that disclaimer on every screen of their products. Do I hear volunteers?
Ultimately, I believe vendors would best serve themselves, their customers, their shareholders, and patients via considering and understanding these points. If a HIT vendor claims to be a partner to clinicians and clinical medicine, they should be willing to accept the responsibilities that accompany such a position.
Scot M. Silverstein, MD is with Drexel University, College of Information Science and Technology.
TPD’s Review of the Samsung N310 Netbook
By The PACS Designer
The Windows 7 Starter Edition software is the key operating system being employed for netbooks. Windows 7 Starter has all of the features of the Windows 7 system except the Aero user interface. The Aero feature primarily offers window transparency so that you can see below your current window similar to looking through a glass pane. Windows 7
Originally I was going to evaluate a Dell Windows 7 Wi-Fi Netbook, but due to some problems with their online ordering system, I decided to obtain a Samsung N310 Netbook from service provider AT&T for $199 after a mail-in rebate.
The Samsung N310 has a 10.1" WSVGA screen, 1GB of memory, and a 160GB hard drive. In its small form factor, it only weighs slightly over three pounds.
Other features of this netbook are an integrated webcam, a Samsung Recovery Solution that restores your system without a CD or DVD, and a method to download applications using Intel’s AppUp Center online service.
One other feature of the Samsung N310 is it offers a 60-day trial version of Microsoft Office Online.
I tested the download speed for Firefox, Chrome, and Internet Explorer. The following response times when downloading HIStalk were: Firefox – 12 seconds, Chrome – 18 seconds, Internet Explorer – 28 seconds. So using Firefox for your internet browser is the best choice for netbooks.
Using the N310 as a highly portable netbook makes it ideal for anyone who travels frequently and needs virtually continuous access to a network connection. However, the smaller screen form factor of 10.1" makes screen navigation more difficult thus you’ll find you are using Control + or Control – keys more often to see content better.
The AT&T network is the standard 3G cell phone system, and access is available anywhere there is cell phone connection towers. The AT&T 3G Network is a bit slower than a DSL solution, but not slow enough to deter its use by everyone.
If there is no cell phone service available, you can still access a Wi-Fi hot spot where your system indicates the possible connection opportunity.
The Windows 7 user interface is simple and easy to use as Microsoft has worked on this new software extensively to improve it over the previous Vista system.
As far as using this netbook daily, it still emulates the larger laptops because of its sizeable keyboard which can accommodate just about any size hand that would use the netbook.
Also, even with its reduced size touchpad, it still provides adequate space for all hand movements.
When it comes to healthcare, it appears to be a useable solution provided that network connectivity inside buildings is not sporadic or restricted.
Overall, the experience to date has been satisfactory, and I would highly recommend a netbook with Windows 7 Starter to anyone looking for a smaller system that is easily portable and uses a cell phone network provider for Internet access. I’m predicting that the Windows 7 netbook with its Intel Atom processor will be the hot product in the 2010 PC space! Windows 7 Popularity
Physician Reluctance to Share Data
By Joe A.
One issue that I do not believe has had enough exposure is what may be underlying the reticence of MDs sharing data. It is the 800-pound gorilla in the room that seems to be hiding behind the sheets – specifically,
malpractice reform.
There appear to be no protections for physicians that "share" data from those fishing for torts — open season for lawyers hunting for malpractice once the data is shared and available.
We are all too familiar with punitive lawsuits, from Mickey D’s coffee to asbestos to tobacco to plastic toys. Putting your clinical data in view seems to me will invite second guessing and lead inevitably to legal actions. If Obama wants HIEs to flourish, he must first work for malpractice reform — something that this current Democrat Congress has zero interest in pursuing, but which is a necessary predecessor to HIEs.
Dr. Silverstein makes a good case for deinstalling HIT when it is used for treating patients since the HIT vendor(s) deny accountbility. Call me, the deinstaller.
Dr. Silverstein also states “If a HIT vendor claims to be a partner to clinicians and clinical medicine, they should be willing to accept the responsibilities that accompany such a position.” HIT vendors have abicated their responsibility to safe patient care. If they were interested in assuring that their gear is safe, they would not have ducked the radar of those entrusted with assuring said safety. A look inside a vendor by a complaint of a dismissed HIT worker shows abject depreciation of safety in the interest of profits. I will send to HIStalk when the time is right.
Dr. Silverstein, please keep shining the light on the deception of vapid vendor language, “partnerships”. Hogwash.
Silverstein is the real deal, a courageous protector of patient safety.
With everyone hopping on the gravy train to feed at the ARRA trough, he has the judgment and presence of mind to realize the serious problems of HIT and the adverse impact it has on patient care.
Silverstein points out that a toxic ingredient has been introduced to hospitals affecting care, and those doing the introducing have no accountability. Kudos and more kudos to Silverstein.
From my experience, if a vendor has some software that makes a calculation based on flowsheet values, that must be tested by the hospital until they are happy with it. Similarly with decision support. Now of course if a update comes down the line that breaks that functionality then some liability should be on the vendor. If people die as a direct result of this, no court in the land would protect the vendor based on these contracts. They are just like none-competes. Everybody has them, but they would not stand up in court.
If the hospital customize the formulas or best practice systems, then why should the vendor be liable? That seems to be what you are saying. “Use” does not just mean just medical use, therefore use the “product appropriately”.
As far as “are not liable for the performance of our products when used in the treatment of patients”, this would be outrageous is that’s in the correct context. I would hope those comments would be released to the general public so they can be informed.
If data is coming from an external system those systems, and the interfaces, must be FDA approved. The source systems are generally approved. What about nurse entered data, who is to blame if data is enter incorrectly? Who would be to blame if decision were made based of faulty transcriptions or speech recognition. Obviously you raise some interesting questions, but it’s not as cut and dry as making EMR’s medical devices in my opinion.
“I believe today’s HIT products involve the practice of medicine by machine proxy” that proxy is usually hospital policy. It’s design and implemented by the individual hospitals using tools built in to the EMR. I know of no hospitals taking large scales best practice advise from vendors. Maybe you do, but I have never seen it.
“Imagine a pharmaceutical company saying “we are not liable for the performance of our products when used in the treatment of patients.” Would you use their products?”
Yeah they would never do that! And if they did, nobody would use it! http://www.npdb-hipdb.hrsa.gov/legislation/title4.html
“If people die as a direct result of this, no court in the land would protect the vendor based on these contracts. They are just like none-competes. Everybody has them, but they would not stand up in court.”
Can you give some examples, or is this just speculation?
It’s speculation based on precedent and common sense. You cannot protect yourself from liability, in any legal system, just by saying “If something goes wrong I am not liable”. Come on. It’s a stupid notion. Maybe you guys should have apply for a job at Toyota, using your logic they could have easily protected themselves from the recall and any associated legal cases by doing something similar.
Even if you go and do some extreme sport, let’s say bungee jumping, you sign a waiver. But if the bungee jumping company were negligent that waiver means nothing.
“Liability for negligence
A business cannot exclude or limit liability for death or personal injury caused by negligence. For other types of loss caused by negligence, liability can only be restricted if it is reasonable for the business to do so.”
http://www.out-law.com/page-419
Some basic research on the exemption clause would show this argument that vedors can protect themselves in cases of injury or death to be false.
If you examine , say, Microsoft’s EULA for Windows and case law relevant to suits against MS for malfunctions, you will find that such agreements are binding. More so when they place liability for mishaps explicitly and squarely on the shoulders of endusers.