Monday, June 15, 2015

Challenging the meme that [yes, there are all these drastic flaws and problems - BUT] ... EHRs improve patient safety

One of the most persistent memes in healthcare IT is that, for all their deficits, bugs, flaws, interferences in care, and so forth, these systems "improve patient safety."

I find the meme remarkable.

37 medical societies can issue a complaint letter about how EHR systems interfere in care and pose patient risk (http://hcrenewal.blogspot.com/2015/01/meaningful-use-not-so-meaningul.html).  The Joint Commission can issue a detailed Sentinel Event Alert outlining the myriad ways that these systems "introduce new kinds of risks into an already complex health care environment where both technical and social factors must be considered" (http://www.jointcommission.org/assets/1/18/SEA_54.pdf).

ECRI Institute can, year-after-year, report health IT as among the top ten technology risks in healthcare (2015 list at https://www.ecri.org/press/Pages/ECRI-Institute-Announces-Top-10-Health-Technology-Hazards-for-2015.aspx).

This writer can casually aggregate quite a few examples of EHR flaws, risks and harms without really trying very hard (http://hcrenewal.blogspot.com/search/label/glitch).  Some of these include incidents where EHR flaws could have or did affect thousands, a feat nearly impossible with paper (http://hcrenewal.blogspot.com/2011/11/lifespan-rhode-island-yet-another.html).

Outages that make all records unavailable can occur with regularity (e.g., http://hcrenewal.blogspot.com/2015/05/another-day-another-ehr-outage-medstar.html).

The ECRI Institute in its "Deep Dive" analysis can gather voluntary reports of 171 IT mishaps in just 9 weeks from 36 hospitals capable of causing harm, with 8 injuries and 3 possible deaths resulting (http://hcrenewal.blogspot.com/2013/02/peering-underneath-icebergs-water-level.html).

Medical malpractice insurers can reveal an increasing number of medical malpractice cases (and injury) involve EHRs (e.g., http://hcrenewal.blogspot.com/2014/02/patient-safety-quality-healthcare.html, also http://cci.drexel.edu/faculty/ssilverstein/cases/?loc=cases&sloc=norcal, also http://www.msms.org/AboutMSMS/News/tabid/178/ID/2595/System-Dangers-How-EHRs-Can-Contribute-to-Medical-Malpractice-Claims.aspx).

Yet, the "BUT" phrase seems to reliably appear in articles about these flaws:

"BUT" EHRs improve safety.

Of course the comparator in such statements is the paper record.

For instance, in the June 11, 2015 Politico report "Why Health Care IT Is Still on Life Support" (http://www.politico.com/magazine/story/2015/06/electronic-medical-records-doctors-118881.html), Arthur Allen sums up the problems very well such as:

  • In surveys, doctors describe the EHR as the biggest cause of job burnout—worse than long hours, billing and other nuisances.  [Burnout is not exactly contributory to patient safety - ed.]
  • One frequent complaint is mental strain.
  • The doctors can’t tell one patient from another in the absence of idiosyncratic impressions. The memorable rash or symptom a patient reported is buried in screen after screen of seemingly trivial data [what I've called "legible gibberish" on this blog - ed.] In an ER or ICU, with time of the essence, this can become a critical safety problem.
  • EHRs are inevitably listed among the 10 top safety concerns for doctors because they introduce new kinds of errors.
  • “All the clicking saps intellectual power and concentration and blocks normal conversation."
  •  “The computerization of medicine will surely be that long-awaited ‘disruptive innovation,’” but “today it’s often just plain disruptive: of the doctor-patient relationship, of clinicians’ professional interactions and work flow, and of the way we measure and try to improve things.”

Yet with all of the above, the following familiar claim is made about these systems:

  • Overall, EHRs are probably improving patient safety—they have replaced illegible medical scrawl with typing, for instance.

At least the word "probably" was used.  Not to single out this article, as the refrain seems commonplace.

I opine in any case that the advantages of occasional handwriting illegibility problem resolved by EHRs are quite thoroughly nullified by critical data being "buried in screen after screen of seemingly trivial data" and other information-clouding issues related to EHR outputs.  See for instance "Two weeks, two reams" at http://hcrenewal.blogspot.com/2011/02/electronic-medical-records-two-weeks.html.

(Missing in this report, like most others on EHR problems such as the May 2015 American College of Physicians report "Frustrations with EHRs rampant as development slows" (http://www.acpinternist.org/archives/2015/05/EHRs.htm) are mentions of patient harm and deaths.  That topic seems verboten.)

In view of all the above, let me state this clearly:

With the increasing amount of knowledge about the flaws of these systems, coupled with the reports of harms in an environment where our top medical organizations and officials admit that the true rate of harms cannot be known due to inadequate reporting infrastructure, policies, and procedures (see http://hcrenewal.blogspot.com/2014/04/fda-on-health-it-risk-reckless-or.html), my belief is that these systems in their present form do not improve patient safety.

My belief is that these systems as they are today decrease patient safety, perhaps markedly, over a reasonably-staffed clinician paper records system. 

To take the enthusiast view is to ignore all of the above.  

For instance, extrapolating the ECRI Deep Dive figures alone is alarming, and to date I have not seen any arguments whatsoever as to why those figures should not be extrapolated.

The situation is only to become worse as more and more hospitals without strong internal expertise increase the complexity of the in-house clinical information systems.

The line that "EHRs increase patient safety" in view of all the problems that are now apparent even to the most hyper-enthusiastic EHR pundit is, I believe, wishful thinking run amok.

Such statements defy common sense.

The need for a very robust reporting mandate on EHR-related close calls and actual harms sorely needed.

It is the only way to know for sure whether we've moved from the occasional paper record-related mishap to a more pervasive EHR-confusion related medical misadventure circus.

Unfortunately, I don't see such mandatory reporting taking place any time soon.  A "health IT safety center" without regulatory authority and receiving HIT mishap reports on a 'voluntary' basis is favored by the industry and its government sponsors (see http://hcrenewal.blogspot.com/2014/07/new-onc-director-karen-de-salvo-no.html).  A safety center will quite likely be "safely" ignored by the sellers and users of the systems, when it suits their financial interests (which is nearly always).  It is a band-aid solution to a very serious problem.

It seems apparent to me, considering all these problems, that health IT incentives should stop.  Further, new EHR rollouts need to be put on hold until this technology is more thoroughly vetted.  Until then, harms and deaths of patients are in part the fault of those who knew, should have known, or should have made it their business to know of the risks of bad health IT.

-- SS

2 comments:

Anonymous said...

The zealots' convictions are undermined by evidence.

Anonymous said...

Maybe the evidence data in support of EHRs got hacked. Perhaps if we asked Eric Snowden to develop a EHR data integration program, China could run the numbers for us and then we could determine the true value of EHRs.