Wednesday, January 09, 2013

Some Real-World Lessons for the Health IT Hyper-Enthusiasts

An article was published in Health Leaders Media yesterday by Scott Mace, senior technology editor entitled "Scot Silverstein's Good Health IT and Bad Health IT" at this link.

(Actually, the terms "good health IT" and "bad health IT" themselves came from Prof. Jon Patrick as a result of my discussions with him in Australia about my conviction, presented to the Health Informatics Society of Australia in my Aug. 2012 talk "Critical Thinking on Building Trusted, Transformative Medical Information:  Improving Health IT as the First Step", that to be trusted and do no harm, health IT must be “done well".)

Scott Mace observes:

Inevitably, when the subject turns to the pitfalls of bad health IT, you will find Scot Silverstein, MD, ready to comment. He has been writing about health IT difficulties since 1998.

Silverstein is an adjunct professor at Drexel University who I recently interviewed for an upcoming HealthLeaders magazine story on physician resistance to health IT.

A recent Silverstein blog post caught my eye for the following statement: "It is impossible for people, especially medical professionals, to be 'ready' for a system that 'is not ready for them.'"

I wanted to learn about the good doctor's thinking and so I gave him a call. We spoke for two hours and it felt like scratching the surface of issues that healthcare will be facing for a good while to come.

Indeed, the issues we discussed were just scratching the surface.  The real world is ever so complex.

Also noted was my observation that:

... Silverstein says it is wrong to think of the tension in healthcare as being IT modernists versus Luddites ... [he says] "I believe the proper framing of this tension between technologists and physicians is that of technology hyper-enthusiasts, who either are unaware of or deliberately ignore the downsides and ethical issues of healthcare information technology in its present state, versus pragmatist physicians who just want to get a job done."

The hyper-enthusiasts largely ignore the real world. 

Two recent "real world" posts on other blogs by practicing physicians caught my eye, that help illustrate the concepts of health IT's disruption of clinicians and of clinical care.  These disruptions increase risk of error (even under normal circumstances; in an emergency scenario, I fear the disruptions will become far more destructive).

These disruptions need to be thrust in the face of the hyper-enthusiasts as characteristic of a very flawed approach to healthcare improvement.

The real-world observations, courtesy KevinMD blog (who reposted them from the source bloggers), with my comments are in [red italics]:

Information overload for doctors increases malpractice risk
Wes Fisher, MD
January 1, 2013

I have used the electronic medical record (specifically EPIC) since 2004.  I have grown accustomed to its nuances, benefits and quirks.  There are parts about it I really like.  There are parts of it I’d like to do without but accept that they are necessary evils in our current health care climate.  I know that there will always be parts of any modified computer system that will suffer growing pains.  For any new and adapting technology this is understandable.

But there is a little-appreciated issue that I see brewing: doctors (and maybe even patients) are quietly being buried by electronic information overload.  As a result, I believe doctors are being placed at an increased liability risk. [Not just doctors, but all clinicians, and the acquirers and implementers of the technology, and those who force the use of it on the clinicians - ed.]

Let me explain.

In the past era of medicine, nothing happened without a doctor’s order.  Nothing.  If you wanted a medication, lab test, invasive procedure, opportunity to participate in rehab classes – anything – you needed a doctor’s order.   For the years of paper records and independent doctors offices, this work flow assured that doctors (1) knew what was happening with their patients, (2) saw their patients, (3) prescribed the proper therapy, and (4) assumed the risk for the intervention or treatment prescribed.  Information proceeded in a logical linear fashion and the doctor was always at the head of the information line.

But we are no longer in the old days in medicine.  We are in the era of near-instantaneous information flow, multi-directional electronic communication, and geographically disparate order entry by “caregivers,” (think nurses, nurse practitioners, advanced practice nurses, clinic operators, registrars, etc.) who help us take messages, continue care, and order things.  In this electronic process, messages are no longer passed from just one individual to another, but rather are passed to two, three, four, or more individuals simultaneously from any one of several different clinical locations – some of which might be many miles apart.  There is an incredible amplifying effect of all of these messages, orders, and notifications — so much so that even the most tech-savvy doctors are struggling to keep up. [This observation about an "amplifying effect" gives life to my own observation that the terms "EHR" and "EMR" are anachronistic and suggest to the layperson an innocuous file cabinet, when in reality today's "EHR" is an enterprise healthcare resource and clinician control system, with all that implies in terms of potential adverse unintended consequences - ed.]

In fact, it is not uncommon for a doctor these days to work for two hours on a procedure and return to the computer to find twenty or thirty new notifications, e-mails, or orders have been deposited there.  Head back in for the next case and then another thirty items appear.  Pretty soon, it’s an avalanche of items.  Worse: doctors must click on each one of these notifications individually to “verify” he or she has looked at each and every single one.  [Looking at the computer has likely become a source of dread to many clinicians; I used to get the same feeling when facing up to a day's emails in Pharma, sometimes more than a hundred - ed.]

Doctors understand that the reason we have to click on all these orders is because (a) no one gets paid in our system unless a doctor orders whatever-it-is [not the best motivation - ed.] and (b) someone has to be the fall guy if there’s a problem with a nurse, medical assistant, or lab technician that “orders” something on behalf of the physician.  [Ditto; the "social issues" of health IT include factors like these  - ed.]  There is even a trend to auto-order things (like a pneumovax vaccine, for instance) that assure the hospital maintains excellent public reporting metrics whether the doctor ordered them or not with the order later appearing in our inbox to be clicked.  [This observation gives life to my own that the computer is increasingly becoming the intermediary between doctor and patient - ed.]

But worst of all are the silent notifications sent from fellow physician colleagues buried amongst the other notifications. They tell of an important story, one that needs fairly urgent attention, but because people no longer pick up the phone, are not immediately noticed or highlighted. It’s like a landmine sitting in a doctor’s inbox waiting to be stumbled upon.

* Click* *Click* *Click* * Click* *Click* *Boom*  [The "silent silo" syndrome, as I called it, also affects lab results reporting.  It should be clear that health IT does not "automagically" improve communications over Alexander Graham Bell's invention - ed.]

With all these people and devices ordering and sending, the limited number of doctors out there are being bombarded from multiple directions.   It is getting harder to keep up these days.  Orders and notices come to us on names we don’t recognize or have been long forgotten.  (Computers don’t forget that you saw the patient eight years ago).  [These observations should put an end, once and for all, to the oversimplifications of comparing health IT to, say, mercantile or banking IT - ed.] And once an order is placed and acted upon without our knowledge these days, we click on the order to clear our notices and thereby assume all the legal risk for the care. The legal buck still ultimately stops with us.

Doctors need to speak up about this problem.  [I could not agree more - ed.] We are not in the old days any longer.  We are in the fast-paced, electronic medical record era where things happen (literally) at the speed of light.  We need the electronic medical record companies, payors, hospitals and legal community to come together to help us find a solution to this current information overload crisis that maintains patient safety and improves efficiencies while limiting legal risks to the doctors who are doing their very best just to keep up.

Wes Fisher is a cardiologist who blogs at Dr. Wes.

The above "anecdote" (I use that term somewhat satirically, see here) is likely characteristic of the lives now lived by most clinicians using today's health IT.  Hyperenthusiasts, take note.

The second real-world illustration of the naivete of the hyper-enthusiasts is as below.  I'd observed most of the points made in my own writings after my CMIO period in the late 1990's, which I highlight.   It is quite interesting to see these same points come from others without Informatics expertise, directly from the clinic:

Explaining the epic failure of EMRs
Kiran Raj Pandey, MD
December 19, 2012

It is no news a lot of doctors like to stick up a rather snotty nose to EMR. The defenders of the EMR tend to label such doctors as archetypal Luddites, sticking to their archaic ways and unbecoming of change and the new times. [In reality, the tension is between the hyper-enthusiasts or 'Ddulites' vs. pragmatist clinicians with real-world patient care responsibilities and obligations - ed.] But as is usually the case with any two heated but opposite arguments, the truth likely lies somewhere between the two extremes.

On an objective basis, there is no denying that automatisation of medical record keeping is the new way forward. In theory, if the machine could keep records for you and give it back to you when and where you want it, thus freeing up valuable time for the patient encounter, that should be winsome for everyone. That alas, is a vision of the EMR utopia [the path to Utopia usually has very bad unintended consequences, and Utopia never reached - ed.], and let alone being anywhere close to such utopia, it is difficult to ascertain if we are even set in the road leading us there. [As I've opined, we're on a speeding bullet train on a quarter-mile track - ed.]

Sometime ago, exasperated at the sheer waste of time that the clunky new discharge module was causing because it would not work the way it is supposed to (my hospital is means challenged, so they are building a patchwork of cheapskate EMR suite on top of their legacy system from the 90s, just to placate the gods of CMS [and the Lords of Kobol - ed.]), I complained to the IT guy that the thing barely works! The guy was sympathetic and said, “look I know the discharge module sucks, just bear with it until the end of the year when we should be able to weed out the bugs.” [Hospitals and clinics, as I've written, are NOT the proper place for software alpha and beta-testing - ed.]

But that’s not all, I said, even if it were working just the way it is supposed to, the discharge still takes me longer than what it used to with paper. “That’s something you will have to learn to live with,” he retorted. “Computer records do take a longer time than paper, and there is nothing I can do to change that.” [This reflects healthcare IT culture's of arrogant acceptance of bad health IT, largely ignoring ways to ease human-computer interaction - ed.]

Right there, I think is where EMR loses a lot of ground against paper records. At any practice, time is the most valuable resource, and anything that doesn’t offer a straight off benefit to save time will have a hard time being adapted. [The reverse is also true - ed.] Add to that the inertia people have about their old ways and you have a deal breaker right there.

That’s not all. Driven by the constant government whip to adopt EMR, and an EMR industry that is hell bent upon imposing itself on healthcare [long ago I began writing of a territorial invasion of healthcare by the IT industry - ed.], a lot of makeshift EMR adoption has taken place. So you have hospitals where one part is using one system while the other is using a completely different one. At one clinic I recently worked at, we had to switch between 3 different EMR systems, just to get the patients records. And there still was the paper records not to mention the dictation.The constant juggling not only made the patient encounters time consuming and cumbersome [and surely tiring - ed.], it literally made us curse at the computers and ruin an otherwise perfectly normal day at work. [And increase risk of cognitive overload and error substantially - ed.] Patient volumes have gone down from 15-16 patients per day to a half of that after EMR adoption.

What’s wrong with the current adoption of EMR? Why are even the converted like me questioning EMR? [Converted to what? - ed.]

I think there are two reasons for such seemingly epic failure. First, how we interface with an EMR. Second how the EMR tries to impose its will on to us instead of the other way around. [I've written that HIT should work like the clinicians work, not the other way around.  Again, the IT has become the cybernetic 'governor' or 'regulator' of care, and is not just an innocuous records system - ed.]  A keyboard and a point and click device may well have worked for many other interactions with the computer, but with an EMR it doesn’t always appear to be nifty.

... On the same note, no EMR is going to be see a faster adoption if something like writing a prescription takes a minute when in paper it barely takes 10 seconds. Right now doing something as simple as writing a prescription feels like running through a bunch of fire breathing hoops. Someone may argue, you can at least read it better [when bugs and 'glitches' due to sloppy industry practices don't cause faulty output such as occurred here - ed.] , but don’t get me started on how the EMR can come up with its own ludicrous set of errors, something that would never be possible with paper. ["not possible with paper" is a theme I've written about as well - ed.]

Trying to impose a ready made architecture on to health care will not work. “It works for retail and banking,” some people seem to offer cluelessly [Business/mercantile computing and clinical computing are two different subspecialties of computing, I've long observed - ed.]. But a patient encounter is no visit to your bank cashier. And human body is not your bank account, it is way more complicated and it is bound to generate way more complex information that is difficult to straight jacket into the rigid and rudimentary pipeline of set information pathways. An ideal EMR is supposed to be a seamless body-glove; today they feel like the hangman’s cloak, not only are they cumbersome, dark and dreary and suffocating, under their apparition, they force things you to do things you wouldn’t otherwise do. [Hyper-enthusiasts don't really seem to care; if it's a computer, it must be better - ed.]

Such forced behavior modification may make the administrator, the insurance company, and the government happy [it does - ed.] but I can’t understand how selecting a dozen pesky radio buttons while doing the discharge makes the patient lead a healthy life or make his doctor particularly enamored with the EMR, just because the government said so, or that it made the IT companies a few million dollars richer. [Doctors are just supposed to obediently accept this technology by the hyper-enthusiasts and profiteers - ed.]

Kiran Raj Pandey is an internal medicine resident who blogs at page59.

I feel "anecdote" #2 is also quite common, and the sentiments shared by a large number of clinicians forced into using this technology in its present state.

Hyper-enthusiasts and other health IT promoters and grandstanders need to read the above accounts well.  They need to understand that the real-world effects of the technology, recklessly pushed, can be toxic, and not result in the utopia of better care and cost-savings they naively believe will deterministically occur. 

-- SS


Anonymous said...

The failure of EMRs to improve outcomes or reduce costs is simple: they just do not cut the mustard.

In other words, they are inferior and sufficient for the tasks of patient care.

Said another way, they are not fit for purpose.

Does anyone not get the picture?

Anonymous said...

Clarification by Anonymous: above should be "inferior and insufficient"...

The device I used to construct the comment thought I did not know how to spell and spelled the wrong word for me. At least there is not a dead patient because of this device error, but the user will always be blamed by the vendor as the learned intermediary.