Thursday, May 18, 2017

Death By A Thousand Clicks: Leading Boston Doctors Decry Electronic Medical Records

Channeling Lyndon Johnson on Walter Cronkite, in clinical medicine, when you've lost Boston (including MGH), you've probably lost the health IT war.

Death By A Thousand Clicks: Leading Boston Doctors Decry Electronic Medical Records
May 12, 2017
By Drs. John Levinson, Bruce H. Price and Vikas Saini

It happens every day, in exam rooms across the country, something that would have been unthinkable 20 years ago: Doctors and nurses turn away from their patients and focus their attention elsewhere — on their computer screens.

By the time the doctor can finally turn back to her patient, she will have spent close to half of the appointment serving not the needs of her patient, but of the electronic medical record.

I have observed this myself in my own personal experiences in recent years observing clinicians in situ.  I (unfortunately) have also been on the receiving end of the resultant distractions, via a dead parent, at the very hospital where I did my residency training in pre-HIT days.

Electronic medical records, or EMRs, were supposed to improve the quality, safety and efficiency of health care, and provide instant access to vital patient information.

Instead, EMRs have become the bane of doctors and nurses everywhere. They are the medical equivalent of texting while driving, sucking the soul out of the practice of medicine while failing to improve care.

"Texting while driving" is, in fact, an excellent metaphor.  The practice of medicine is often a highly-demanding activity, cognitively speaking, just as driving is.  Slip up even for a moment, and accidents and catastrophe can occur.
To fix them, hospital administrators and clinicians need to work together to demand better products from EMR manufacturers and to urge government to relax several provisions of the HITECH Act, the 2009 law that spawned many of the problems with EMRs.

I do not think this is feasible, nor do I think it would solve the problems.  More on that below.

How did technology that has increased efficiency in every other industry become such a drag on health care? For starters, people who take care of patients did not design or choose these systems. They were foisted upon us.

Doctors and nurses know that good diagnosis and treatment requires listening attentively to their patients. They spend years learning to parse the clues that patients offer, both the physiological and personal, in order to provide the right care.

The sound of medicine is not the click of a mouse. It is the human voice. Let’s bring it back.

To do this well takes time and undivided attention.

"Undivided attention to the patient" and "use of health IT" under the current paradigms are incompatible.  The terms do not belong in the same sentence.  Again, more on that below.

... Instead of making this easier, most EMRs create extra work. A lot of extra work, thanks to endless prompts with multiple choice answers that hardly ever fit the facts and that demand click after click to get anything done.

Want to order a simple test? That requires getting through multiple prompts. Need to write a prescription -- an exercise that used to take less than 15 seconds? Another set of clicks.

Template madness is the coin of the realm.

Four screens of hundreds, from just one EHR, of many EHRs that a clinician might need to use in practice of medicine.

Typing, filing, mailing results and placing referrals all used to be done by assistants. Now, EMRs put that burden on clinicians, and we must do it during office visits, or “encounters,” as EMRs call them. And when the wrong button is clicked, the wrong test or drug is ordered, or it does not go through at all, delaying medical care.

It’s death by a thousand clicks, and it happens every day.

Another excellent metaphor, a derivation of "death by a thousand cuts."  However, in this case, it is not just referring to death of the ability to maintain medical practice and retain sanity (and health of the practitioner), but also injury and death of patients.

We are frustrated by EMRs because they pull us away from our patients. We are driven mad by the fact that EMRs in different locations do not talk to each other. And we think it’s just wrong that much of the EMR’s busywork is about optimizing billing for the hospital.

The lack of interoperability is likely protective against, in effect, mass copy-and-paste on a grand scale through importation of a chart from one EHR to another, resulting in propagation of errors.  In any case, interoperability needs to take a back seat to operability (usability).

Who is to blame? Start with EMR manufacturers, who lobbied Congress to require every hospital and doctor’s office to install an EMR system; hospital administrators who bought technology that conveniently pushed billing duties onto doctors and nurses; and federal regulators, who imposed on EMRs numerous quality metric requirements that do nothing to improve care.

Everyone in healthcare is to blame, including clinicians who DID NOT SPEAK OUT until obvious issues crated the terrible state clinicians find themselves in now.

We do not want to go backward. We believe that computing is essential to the future of medicine. We simply want all EMRs to live up to their promise of improving care and making patient information readily available.

I've been calling the problems out for close to two decades now.

The issue becomes:  what to do about it?

My area of Medical Informatics research in the early 1990s was to move away from the traditional GUI paradigm of menus, widgets, control buttons, etc to explore novel (and menu-free) paradigms of EHR-based data content visualization, navigation and querying.  The commercial companies, when shown this, had no interest in even considering such ideas.

Other avenues to enhance the traditional GUI interface - e.g., AI-based "wizards" of one sort or another that anticipated likely down-the-line choices from currently entered data and presented the choices to the user, better-designed EHR roadmaps and/or AI based on data entry to-the-moment to  allow simpler and less time-consuming navigation, etc. - all proved not very helpful due to the complexity of the domain, not to mention its specialties and subspecialties.  This experience informed my decision down the line to move to specialized and malleable paper forms for the clinicians, and data entry clerical teams, for high risk and/or high volume procedural or critical care areas such as invasive cardiology, cardiac surgery and the ED.

See my Aug. 2016 post "More on uncoupling clinicians from EHR clerical oppression" at for a specific, highly successful example.

It is my belief that the "traditional" model - GUI-based EHRs with the typical paradigms, and clinicians doing the bulk of the data entry that itself is bloated relative to the actual clinical need, and tying all human-computer interaction to a computer screen roughly at the resolution of one or two legal-sized pieces of paper, can never be improved to the point of not impairing the clinical workday.

The entire health IT enterprise needs to be rebooted (rethought), especially regarding roles, workflow, and most especially the "data capture load per clinician".

When I was writing about EHR issues in the late 1990s and into the 2000s, I was called a Luddite and alarmist (or worse, e.g., see my 2010 posts "The Dangers of Critical Thinking in A Politicized, Irrational Culture" at and "More on Perversity in the Healthcare IT World: Is Meditech Employing Sockpuppets?" at

It appears I was correct.

My issuing the prediction that the current model of health IT is irreparably broken (or, said differently, that the health IT experiment - and it has been an experiment - under its current paradigms has failed), and needs a top-to-bottom rethinking, will probably be taken the same way.

Until it isn't.

-- SS


Anonymous said...


Live IT or live with IT said...

I'm sure someone would complain if you reduced their government funded gravy train. MORE MONEY IS NEEDED TO SOLVE A PROBLEM CREATED BY MORE MONEY!!!