Monday, February 18, 2013

Kaiser Health News/Philadelphia Inquirer on InformaticsMD: "The flaws of electronic records"

At my Dec. 2012 post "How an interview for Kaiser Health News rekindled memories of health IT dysfunction in the 90's that persist in the 10's" I mentioned an interview by a reporter from the Kaiser Health Foundation regarding health IT flaws.

His article appeared in both the Philadelphia Inquirer and Kaiser Health News today under the title "The flaws of electronic records":

Philadelphia Inquirer / Kaiser Health News
Feb. 18, 2013
The flaws of electronic records

Drexel University's Scot Silverstein is a leading critic of the rapid switch to computerized medical charts, saying the notion that they prevent more mistakes than they cause is not proven.


Scot Silverstein of Lansdale, one of the most ardent critics of electronic medical records, works on an antique computer. A growing collection of evidence suggests poorly designed medical software can obscure clinical data, generate incorrect treatment orders, and cause other problems.

RON TARVER /Staff Photogapher

Posted: Monday, February 18, 2013, 3:01 AM

Computer mistakes like the one that produced incorrect prescriptions for thousands of Rhode Island patients are probably far more common and dangerous than proponents of electronic medical records believe, says Drexel University's Scot Silverstein.

Flawed software at Lifespan hospital group printed orders for low-dose, short-acting pills when patients should have been taking stronger, time-release ones, the Providence-based system disclosed in 2011. Lifespan says nobody was harmed.

But Silverstein, a physician and adjunct professor of health-care informatics who is making a name for himself as a strident critic of electronic health records, says the Lifespan breakdown is part of a much larger problem.

"We're in the midst of a mania right now" as traditional patient charts are switched to computers, he said in an interview in his Lansdale home. "We know it causes harm, and we don't even know the level of magnitude. That statement alone should be the basis for the greatest of caution and slowing down."

In an ethical world, it would be.  Medicine, though, is in the throes of a loss of ethics, as many stories at this site and many others attest to.

Use of electronic medical records is speeding up, thanks to $10 billion and counting in bounties the federal government is paying to caregivers who adopt them. The consensus among government officials and researchers is that computers will cut mistakes and promote efficiency. So 4,000 hospitals have installed or are installing digital records, the Department of Health and Human Services said last month. Seventy percent of doctors surveyed in September by research firm CapSite said they had switched to digital data.

But the notion that electronic charts prevent more mistakes than they cause just isn't proven, Silverstein says. Government doesn't require caregivers to report problems, he points out, so many computer-induced mistakes may never surface.

The recent ECRI "deep dive" study of "EHR events" (link) is just the latest to raise red flags on that point.

Even Dr. David Blumenthal, former chair of the Office of the National Health IT Coordinator (ONC) in HHS seems to have changed his message about reporting of medical problems.

From a Feb. 16, 2013 New York Times article "Doctors Who Don’t Speak Out":

....  TRADITIONALLY, doctors have brought problems to the attention of colleagues by conducting research and publishing their findings in a medical journal. The advantage of that system helps ensure the credibility of study data and protects a researcher from random attack, said Dr. David Blumenthal, the president of the Commonwealth Fund, a group that studies health policy issues.

But getting a study published can take a year or two; some Johnson & Johnson consultants did publish studies about the hip’s flaws, but they largely appeared after it had been recalled.

Dr. Blumenthal said there was probably a need for more immediate ways for doctors to share their concerns, like forums supported by professional medical organizations.

Back to the Inquirer article:

He doesn't discount the potential of digital records to eliminate duplicate scans and alert doctors to drug interactions and unsuspected dangers.

But the rush to implementation has produced badly designed products that may be more likely to confound doctors than enlighten them, he says. Electronic health records, Silverstein believes, should be rigorously tested under government supervision before being used in life-and-death situations, much like medical hardware or airplanes.

In fact, arguments otherwise are specious.  "Harm to innovation" is the one I've heard most often.  Yet, those proffering such claims cannot point out what aspect of regulation - adherence to GMP's (good manufacturing processes), pre-market safety and quality testing, post-market surveillance, etc. - will "harm innovation."  Indeed, they opine as if innovation in medicine without objective safety and quality checks is a virtue, rather than a potential vice.

Silverstein "is an essential critic of the field," said physician George Lundberg, editor at large for MedPage Today and former editor of the Journal of the American Medical Association. "It's too easy for those of us in medicine to get excessively enthusiastic about things that look like they're going to work out really well. Sometimes we go too far and don't see the downside of things."

(Dr. Lundberg mentioned me in Nov. 2011 at MedPage today in a piece entitled "Health IT: Garbage In, Garbage Out" as here.)

The patients - including ourselves and our own family members, I might add - are the ones who pay the price of our hyper-enthusiasm.

... The FDA's Jeffrey Shuren, a neurologist, has said such cases "likely reflect a small percentage of the actual events that do occur."

"Tip of the iceberg" were his exact words (link).

... At conferences and working from home on the "Health Care Renewal" blog, Silverstein chronicles digital failures and criticizes hospitals in the same dogged way he applied himself to building the 1970s-era Heathkit computers [and amateur radio equipment - ed.] he still keeps in his home, say people who know him.

"His message has been consistent": Health IT "provides far less benefit than is claimed by its proponents and opens new, sometimes potent, routes to failure," said Richard I. Cook, a medical error expert at the University of Chicago who sat on a panel examining electronic record safety at the authoritative Institute of Medicine. "No one wants him to be visible. But his message and tone have not wavered."

Dr. Cook wrote the lone dissent to milquetoast IOM recommendations on health IT risk - the magnitude of which IOM itself admitted is unknown - in their 2012 report "Health IT and Patient Safety: Building Safer Systems for Better Care" available here.  See Appendix E.

The last scientific conference at which I chronicled these failures was at the Health Informatics Society of Australia's HIC2012, as a keynote speaker on health IT trust (link).  I would have been at HIC2011, to which I had been originally invited, but was helping care for a relative injured by bad health IT at that time.  Her death freed me to travel Down Under in 2012.

I have presented at a number of plaintiff attorney's meetings since then, however, such as the American Association for Justice Winter Convention (AAJ, formerly the Trial Lawyer's Association) just last week (link).  Trial lawyers don't seem to mind a very direct approach to the issues, unlike many so-called scientists who, as author Michael Crichton once warned (link), seem to believe in "consensus" rather than science.

The HIMSS Electronic Health Record Association, an industry group, declined to comment on Silverstein. A spokesman for the Office of the National Coordinator for Health Information Technology, the administration's proponent of digital records, said: "It's important to listen to all the voices" in the discussion of the subject. 

Some voices, unfortunately, are louder than others and backed by lobbyists and big money.  See, for instance, the May 2009 Washington Post article by Robert O'Harrow Jr. "The Machinery Behind Health-Care Reform: How an Industry Lobby Scored a Swift, Unexpected Victory by Channeling Billions to Electronic Records" (link to the article and my essay about it is here).

I do point out, however, for the benefit of those at HIMSS and ONC, that knowing of risks, while doing nothing substantive while "listening to all voices in the discussion" can be seen as gross negligence.

Trained as an internist and in medical information technology [Medical Informatics- ed.] as a Yale University postdoc, Silverstein, 55, served as Merck & Co.'s director of scientific information in the early 2000s and then as a full-time Drexel professor, shifting in recent years to part-time teaching and working on medical liability cases for plaintiff attorneys. His insistent warnings about digital health risks over more than a decade have effectively barred him from a lucrative career at a hospital or software vendor.

Perverse as that reality may be, it's also the reverse:  I would not want to work for a hospital or software vendor in 2013, where effecting change to protect patients from bad health IT is hard if not impossible, e.g., as at link, link, and link, and at the other case examples at that site.  I find it a far more effective use of my time to help enact change from the outside - and avoiding the pathological individuals who make such scenarios possible.

"I'm sure Scot would be better off by keeping his mouth shut and getting a job with a hospital that's just put in a big . . . system," said Matthew Holt, a Silverstein critic and cochairman of Health 2.0, which organizes health technology conferences.

I note that the raison d'être for this blog is the impact of many people doing just that sort of thing - "keeping their mouth shut" and making money, no matter what the ethical implications.  (The patients killed by such behaviors are, unfortunately, unavailable for comment.)

Many say he comes on too strong. Even admirers cringed when he began blogging about the 2011 death of his mother, which he blames in a lawsuit on a computer error that allegedly caused a hospital to overlook a key medication. Personalizing his campaign, some thought, made him seem less objective.

I'm at a loss here.  "Too strong" - on matters of life and death that affect everyone?  A proficient writer, who had been writing about health IT problems since 1998, should have kept silent about a first hand story of HIT harm in 2010 of potential great relevance towards public safety?  Writing about the incident 'lessens objectivity'?  That makes little sense, and is perverse.  Those who opined as such have their priorities in serious disarray.

Such personalized accounts are common and have caused great change.  Libby Zion's death due to hospital neglect, and her father's making the issues quite public, comes to mind, as do the many laws enacted that are named after people who've gone public after personal tragedy:

... Grieving the loss of their child, Zion's parents became convinced that their daughter's death was due to inadequate staffing at the teaching hospital.  Sidney Zion questioned the staff's competence for two reasons. The first was the administration of meperidine, known to cause fatal interactions with phenelzine, the antidepressant that Libby Zion was taking. The second issue was the use of restraints and emergency psychiatric medication. Sidney's aggrieved words were: "They gave her a drug that was destined to kill her, then ignored her except to tie her down like a dog." To the distress of the doctors, Zion began to refer to his daughter's death as a "murder." Sidney also questioned the long hours that residents worked at the time. In a New York Times op-ed piece he wrote: "You don't need kindergarten to know that a resident working a 36-hour shift is in no condition to make any kind of judgment call—forget about life-and-death." The case eventually became a protracted high-profile legal battle, with multiple abrupt reversals; case reports about it appeared in major medical journals.

An alternate explanation is that, lacking other credible means, this is an ad hominem reaction (of those "some") seeking an angle to attack the message ... and the messenger ... or is simply a reaction of, to put it bluntly, castrati who are more at home in a country club than in the world of ideas.

For a bona fide example of "coming on too strong", there's this:

In the NEJM article "The 'Meaningful Use' Regulation for Electronic Health Records", David Blumenthal, M.D., M.P.P. (ONC Chair) and Marilyn Tavenner, R.N., M.H.A. (10.1056/NEJMp1006114, July 13, 2010) available at this link, the opening statement is (emphases mine):

The widespread use of electronic health records (EHRs) in the United States is inevitable. EHRs will improve caregivers’ decisions and patients’ outcomes. Once patients experience the benefits of this technology, they will demand nothing less from their providers. Hundreds of thousands of physicians have already seen these benefits in their clinical practice.

I think it fair to say those are grandiose statements and predictions presented with a tone of utmost certainty in one of the world's most respected scientific medical journals. 

Even though it is a "perspectives" article, I once long ago learned that in writing in esteemed scientific journals of worldwide impact, statements of certainty were at best avoided, or if made should be exceptionally well referenced.

I note the lack of footnotes showing the source(s) of these statements.

Another bona fide example of "coming on too strong":

“We have the capacity to transform health with one thunderous click of a mouse after another,” said (former) HHS Secretary Michael Leavitt - 2005 HIMSS Summit 

We shall transform health (into what, exactly, is not specified) one thunderous mouse click after another!

It doesn't get any stronger than that, unless, perhaps, the thunderous wrath of God is invoked.

Back to the Inquirer once again:

"His refusal to temper his message makes it sometimes difficult to hear," said Ross Koppel, a University of Pennsylvania sociologist and digital health record skeptic.

As per a recent article by Joe Conn in Modern Healthcare entitled "Health IT Iconoclasts" (link), which wrote of Dr. Koppel, Deborah Peel, Lawrence Weed and me, Dr. Koppel has firsthand experience at his message being found "difficult to hear" by the hyper-enthusiasts:

... Researcher Ross Koppel started an uproar in 2005 when he and a colleague coauthored an article in the Journal of the American Medical Association that found a first-generation computerized physician order entry system (CPOE) at the Hospital of the University of Pennsylvania was simultaneously creating new errors even as it reduced others.

Koppel’s bombshell—he’s now an adjunct professor of sociology at the University of Pennsylvania— brought down the wrath of information technology boosters. The Healthcare Information and Management Systems Society, a health IT trade group, challenged the study’s “methodology and its subsequent outcomes,” and criticized its authors for their “limited view” and not “looking at the big picture.”  [Others wrote that his work was 'disingenuous', although it had similar findings to my own observation of the very same CPOE system at Yale-New Haven Hospital ... in 1992 - ed.]

... In 2009, he revealed in another JAMA article that health IT vendors’ contracts included “hold harmless” clauses that shielded software developers from legal liability for medical errors their systems caused, even if the developers had been warned about the defects. “That got me major upheaval,” the worst of his career, Koppel recalls.

Koppel, a sociologist, has probably done more for health IT transparency and safety than the physicians of the entire academic medical and medical informatics community combined.

... But Silverstein says his position today is the same as it has always been. He believes in the potential power of electronic records for good, he says. But any doctor who feels bound by the Hippocratic oath's injunction to "first, do no harm," he adds, should balk at what's going on.

"Patients are being harmed and killed as a result of disruptions to care caused by bad health IT," he said. "I'm skeptical of the manner and pace" of implementation, "not of the technology itself. . . . My only bias is against bad medicine. And my bias is against people with complacent attitudes about bad medicine."

The issues are actually relatively simple, using terminology coined by Dr. Jon Patrick of U. Sydney at the aforementioned HIC2012 meeting in Australia during our discussions.  Bad health IT must be removed from the market, and good health IT must replace it.

Good Health IT ("GHIT") is defined as IT that provides a good user experience, enhances cognitive function, puts essential information as effortlessly as possible into the physician’s hands, keeps eHealth information secure, protects patient privacy and facilitates better practice of medicine and better outcomes.

Bad Health IT
("BHIT") is defined as IT that is ill-suited to purpose, hard to use, unreliable, loses data or provides incorrect data, causes cognitive overload, slows rather than facilitates users, lacks appropriate alerts, creates the need for hypervigilance (i.e., towards avoiding IT-related mishaps) that increases stress, is lacking in security, compromises patient privacy or otherwise demonstrates suboptimal design and/or implementation.  
It is this author's opinion that this change will not happen by "going along to get along" or "listening to all the voices in the discussion of the subject" (especially those with Big Money and Big Lobbyists behind them) while doing nothing.

I also note that "complacent attitudes about bad medicine" are not the sole province of IT personnel or healthcare management.  Physicians and nurses who acquiesce to bad health IT are part of the problem.

-- SS


Anonymous said...

Patients are being killed and injured daily from dysfunctional care processes being dictated by rigid devices programmed to tret every human as if they are all the same, like cars on an assembly line. Medicine is being stifled and shackled by these devices that are not fit for purpose

Keep working at this. The truth will rule.

Matthew Holt said...

Hey--I'm a critic! It says so officially, but actually what I was quoted as saying was praise, basically Scott chose to speak up at cost to himself. Of course I *am* a critic as I think there is a middle ground about the value or EMRs, and there was about a 30 minute "but..." I told the reporter which didn't get into the article. Scott, however, knows what that is.

Having said all that, Good Health IT is very closely associated with "Health 2.0" which Indu & I define as
--adaptable technology (that plays well with others)
--enhancing the user experience
--getting data to and from the point of decision to drive decisions.

I think Scott and I just disagree about how to get to the world of Good Health IT, and also about whether the world of "bad" health IT is a phase we sort of have to go through to get there, and whether it's better than paper and "No Health IT". I say yes to both, I think he says no.

Anonymous said...

Politicians owe big favors to their largest contributors, as revealed by the Menendez proceedings in New Jersey.

I can not imagine that the HIT industry, including HIMSS, did not give back to the political machine that gave it special accommodations by declaring their devices safe by whim.

Anonymous said...


I had a chance to read over Hancock's write-up about you and the EHR quagmire. I would characterize it as mostly fair, with a few jabs thrown in. The issues of which you are so passionate about were given their just measure. Having Lundberg and Cook on your side was weighty.

I had a good laugh over the "no comment" from HIMSS EHR . Whatever could they say? The comment that followed ("It's important to listen to all the voices") was lip service if ever I heard it. It makes clear who your friends and enemies are. Or perhaps those who are threatened by your outspokenness.

I was miffed by the negative connotation ascribed to your own personal experiences with your mother. Did anyone criticize John Walsh when he created America's Most Wanted after his son's murder? I think it's admirable. A powerful motivating force not mutually exclusive with objectivity.

Finally, who is this Matthew [D]olt? Is he even a physician? Does he have the slightest insight into the clinical thought process, or does he consider it a discipline of systems engineering?
For non-physician wonks in the health care arena, I remind them of the profound words of the late Earl Butz, "You no playa the game you no maka the rules".

InformaticsMD said...

Matthew Holt said...

It says so officially, but actually what I was quoted as saying was praise, basically Scott chose to speak up at cost to himself.

I respect that.

I think Scot and I just disagree about how to get to the world of Good Health IT, and also about whether the world of "bad" health IT is a phase we sort of have to go through to get there, and whether it's better than paper and "No Health IT". I say yes to both, I think he says no.

My views are clear. Health IT is a good thing...when it's good health IT.

Putting "bad health IT" in the clinic or in the hospital without rigorous and objective pre-market vetting, quality validation, post marketing surveillance (and mandated withdrawal if needed), etc. is not a "phase we need to go through" to get to good HIT, however.

For doing so is the very definition of unconsented human experimentation, and negligence.

-- SS

InformaticsMD said...

Anonymous February 18, 2013 at 11:53:00 PM EST wrote:

I was miffed by the negative connotation ascribed to your own personal experiences with your mother. Did anyone criticize John Walsh when he created America's Most Wanted after his son's murder? I think it's admirable. A powerful motivating force not mutually exclusive with objectivity.

The 'negative connotation' is a form of ad hominem. It's the last angle of attack ... of scoundrels.

Some such as Bill Hersh of OHSU have gone further, referring to my mother as an "anecdote."

He wrote "We make inferences and broad pronouncements about [anecdotes] at our own peril, though you seem to do this years after your one anecdote [my mother's death]."

Not only is the statement without merit, but it's brutally insensitive and obnoxious.

It is a sign, I believe, of contempt of my views, taken to an extreme.

Not even militant (and perhaps mob-connected) transit union leaders I once worked with were so callous.

-- SS

Anonymous said...


If you want to get there, let's hold the vendors acountable, have experiments on protocol approved by IRBs, and present the findings to the FDA. Additionally, after market surveillance of these defacto illegal devices should be run by the FDA, and, doctors and other users should be so informed and given protection from the teams of hospital-vendor who have been known to retaliate against those who complain when patients die from the toxic affects of the (bad)hit.

Matthew Holt said...

Fair enough, anonymous, IF you're prepared to do the same thing to the "placebo" which in this case is paper-based health care processes. The answer is you can't.

Simple fact is ALL health care delivery has been poor and dangerous in this country and others for years. You only have to look at multiple IOM reports to see that. So to say that using even "Bad Health IT" is worse than that is a) unprovable and b) almost certainly wrong, given the generally good results of the 2 biggest systems using "Bad Health IT" -- Kaiser and the VA

Oh, and as you and Scot wont like me pointing out but I will anyway, FDA approval does not equal safety. (I won;t upset Scott by saying the word, but you can buy a V....)

InformaticsMD said...


The issues are actually relatively simple.

1. The IOM and FDA have stated the magnitude of health IT harms are unknown (due to systematic impediments to information collection, diffusion, etc., but the reasons are an aside here).

2. In medicine, we ethically are compelled not to force technology or treatments of unknown risk on patients, unless it's voluntary and we have their explicit informed consent.

3. The issue of paper's risks is thus irrelevant. In medicine we do not replace one intervention with another because the latter might be better - particularly under the conditions of not knowing the risk of the latter.

There's really nothing to argue or debate here from an ethical perspective.

More generally, a major problem seems to be a subset of health IT enthusiasts who believe their right to experiment on patients with computing exceeds the rights of patients to a safe medical environment, and clinicians not harassed by bad health IT.

Finally, I can say with a reasonable degree of confidence, after presenting on HIT risks and attending sessions on trial examination and cross-examination techniques by some of this country's top trial attorneys at their recent Winter Convention (unfortunately Mr. Goldfinger and 007 were not at the hotel that day):

People expressing anything like that view - or views that HIT is of less risk that paper thus it should be rolled out without rigorous safety, reliability, fitness for purpose etc. testing - will be ripped to shreds in court.

And then made to pay for their ideas if they were in a position of authority.

-- SS

InformaticsMD said...


Two more items.

1. Review the Nuremberg Code of Ethics at this link. I would like to know if, ot why, a person would believe it should be ignored in the domain of health IT.

2. I would like your opinion on Jon Patrick's work on Cerner Firstnet. That work is located at this link (item 3).

Is it legit for the government of NSW/AU, in your opinion (or of any other country) to ignore this work, as has been done, and continue with rollout?

-- SS

Anonymous said...

I can think of more questions for Matthew.

He writes: So to say that using even "Bad Health IT" is worse than that is a) unprovable and b) almost certainly wrong, given the generally good results of the 2 biggest systems using "Bad Health IT" -- Kaiser and the VA


1. Maybe you could provide the papers that show the "generally good results" (that seem to be attributed solely by you to computer), and show that the methodologies of these papers are sound and not P.R.?

2. Considering as covered at this blog the fiasco of thousands of wrong prescriptions in Rhode Island, the orders going into the wrong charts at Trinity health system, the cases of injury and death, system outages putting entire hospital's worth of patients without records, the new ECRI study showing 800 EHR 'events' in 9 weeks, the disruptions of care for entire cities like San Francisco, and others - a careful person might surmise that "Bad Health IT" might be worse than paper.

3. What is the exact rate of errors due to paper records, anyway, especially that could not be fixed by simpler means than spending billions on computers?