Friday, August 09, 2013

A War on Patients: Panel Says EHRs Should Not Be Vetted Before Marketing and Deployment

"First, do harm - it's a learning experience, and injured or dead patients are just a bump in the road, anyway" - the apparent creed of the healthcare computing hyperenthusiasts

Joe Conn and Modern Healthcare published the following article:

Work group says OK to some HIT safety regs (link), Joe Conn, Modern Healthcare, Aug. 7, 2013

What is important is what safety regs the Workgroup said "no" to.  It comes as no surprise:
A federally chartered special work group with representatives from three federal agencies has submitted its draft recommendations on establishing a regulatory framework for health information technology. Chief among those recommendations is that health IT should not be subjected to pre-market federal regulation, but there were a few exceptions.

The exceptions are narrow, and are likely already covered as Class III medical devices by FDA (see

The exceptions under which there should be FDA regulation, according to the work group, include medical device accessories to be defined as such by the FDA; certain forms of “high risk” clinical decision support systems, such as “computer aided diagnostics,” also to be defined by the FDA; and some “higher risk software” use cases to be defined by the committee's own safety work group.

They did acknowledge the need for postmarket surveillance:
... The group also recommended: developing a federally supported, post-market surveillance system for health IT products “to ensure safety-related decision support is in place,” creating a process for gathering information on safety issues, aggregated at the federal level and establishing a public process for “customer rating of HIT to enhance transparency.”

Dr. David Bates [a professor at Harvard Medical School], chairman of the Food and Drug Administration Safety Innovation Act work group, presented the preliminary findings Wednesday at a meeting of HHS' Health Information Technology Policy Committee.

Let me translate this to plain English:  the health IT systems that go in (and their upgrades and patches) are recommended to be free from pre-marketing regulation and regulatory vetting.  Patients are to be the guinea pigs for testing of the software.  

If patients are harmed or killed, they get the honor of being named as "postmarket surveillance learning cases" who gave their all for the betterment of healthcare information technology.  

(Without their consent, but who needs consent to test experimental and unvetted devices on guinea pigs?)

Bates did express some liability concerns:

Asked during a question and answer period following his presentation whether the committee had considered the liability implications of its recommendations, Bates said, “It's not something we discussed at length, but it's something we can discuss over the next month.”

I, on the other hand, as a legal consultant on health IT-related medical errors and evidence tampering, am considering liability issues.

Unfortunately, patients would rather be whole than in lawsuits (or dead).  Also, sadly, it's physicians and nurses who will bear the brunt, if not all, of the liability for bad outcomes due to defective IT such as at these two recent posts, with vendor alerts regarding serious flaws of medication and other orders not being retained:

A clarification for all those proletarians who lack Harvard educations, and for the Workgroup members as well. Allow me to point out that the above manufacturer safety alerts of life-threatening fundamental flaws (involving entered text that "disappears", apparently found in live-patient scenarios, and the other "glitches" that did cause life-threatening errors sometimes en masse involving thousands of patients such as another apparent Siemens debacle at would likely not have occurred if the systems had been vetted before being turned loose on patients.

Finally:  David and panel members, my mother and I thank you profusely. 

Oh mother can't thank you, she's dead from the toxic effects of un-premarket-vetted health IT on simple care processes at the very hospital where I performed my residency two decades ago.

She might have died a few times before she actually did thanks to other IT "glitches" that cropped up during her recovery from the first one, but I was able to (in one case, by sheer happenstance of showing up at  the right time) discover or provide staff with information to work around additional unvetted-health-IT flaws before those did her in.

It's taken more than a decade for critical-thinking, unconflicted writers and researchers ("iconoclasts") to force cybernetics-over-all hyperenthusasts (see here) like Bates and his panel members to own up the risks of health IT at all, e.g. via sites like this blog and this teaching site. These panel members IMO have their heads buried in sand.

Dr. Bates and his panel are, in my opinion, healthcare IT extremists, which is in part the apparent holding of the belief that computers have more rights than patients - and the other beliefs mentioned in this post:  "Another Health IT 'Glitch' - Can Digital Disappearing Ink Kill Patients?" at

-- SS


Anonymous said...

They did acknowledge the need for postmarket surveillance:

Bet that will work out well. The same requirement attached to rDNA insulin a couple of decades ago . . . we're still waiting for THAT to happen.


Anonymous said...

Vendor sponsored recommendations. The Siemens' Safety Advisory letters are a perfect example of why pre market assessment and validation is mandatory.

Anonymous said...

The fact that EHR software is treated this cavalierly is due to the sheer amount of cash they throw around. This needs to be treated like any other biomedical device.

Anonymous said...

As configured, the current iteration of EHR and CPOE devices are a disease, requiring more cognitive and intellectual power to solve and use from the user, than does the disease of the real patient.

The FDASIA Work Group's conclusion is flawed and as such, is detrimental to the entire health care system in the USA.