Reflecting on the Electronic Medical Record

Happy MayDay, btw.

Ten years ago this month I took my retirement from HealthInsight, the Nevada-Utah Medicare QIO. It'd been my 3rd tenure with them, and the focus of this work was on the federal "Meaningful Use" initiative, via which to accelerate the conversion of outpatient primary care clinics from paper-based medical records to HHS "Certified EMRs" (Electronic Medical Records, also more broadly dubbed "EHRs"—Electronic Health Records).

 

The goal—and assumption—was quantifiably improved health care outcomes resulting from having digitized current and longitudinal patient data always close at "hand" (literally, via keystrokes or mouse clicks).

 

I was certainly a Kool-aid drinking True Believer at the outset, particularly in light of my lengthy former work as a laboratory programmer and QC statistician (pdf) in Oak Ridge, and my abiding interest in all things QI as an ASQ Certified Quality Engineer. Things would turn out to be considerably more complex, however.

 

This morning I opened my New Yorker to find this:

The Curious Side Effects of Medical Transparency
When we peer into our patient portals, we don’t always see ourselves more clearly.
By Danielle Ofri

One afternoon not long ago, I sat entering notes into a patient’s medical record. She was in her forties, and her labs showed anemia. The causes of anemia range from menstruation to cancer, and so pinpointing the correct underlying diagnosis is critical. Physicians are trained to formulate a full roster of possibilities, known as the differential diagnosis, and then to work down the list systematically. We’re taught to cast a wide net—celiac disease, parasitic infections, thalassemia, lead poisoning, liver disease, B12 deficiency, myeloma, sickle-cell disease, G6PD deficiency—because you’ll never make a diagnosis if you haven’t included it in your differential.

But I hesitated before entering my differential into the computer system. Should I include the more serious possibilities, even though they were much less likely? In the past, I wouldn’t have thought twice about it, as the chart served primarily as a tool for the medical team to communicate among ourselves. But a new law, the 21st Century Cures Act, had recently been fully implemented, making medical records open to patients by default, in real time, including doctors’ notes. My in-box was already jammed with panicked messages from people convinced that they had catastrophic illnesses, based on minuscule lab discrepancies and panic-inducing Google searches. How would my patient react to seeing my ruminations about possible colon cancer or duodenal ulcer in the note?…

From my end of the stethoscope, it’s always seemed obvious that patients should own their medical records, and be able to see them. But openness can be challenging in practice. When our hospital initially rolled out its patient portal, a few of my older patients asked that I remove references to erectile dysfunction from their medical records. Their adult children handled the household tech, they explained, and my patients preferred to keep their Viagra prescriptions private. In other cases, multiple family members can access a patient’s chart, messaging me about test results and treatment plans, and it can be complicated to deduce the hierarchy of responsibility. Advocates have raised increasing concerns about the ease with which abusers can gain access to victims’ medical records; health-care settings have traditionally been secure places for people experiencing domestic violence, elder abuse, and human trafficking, but, with medical records becoming more accessible, patients may feel less certain that their words are safe.

At the same time, there are remarkable upsides to the transparent medical record. Patients are able to review their diagnoses, medications, and treatments at their own pace. They are able to share information with family members if they choose. They can spot errors—medications they are no longer taking, medical history that’s missing, allergies that aren’t noted. They can prepare for their next visit, and perhaps feel on more equal footing with their doctor. And they can obtain test results without the interminable wait for the doctor’s call.

Medicine is a highly specialized field, and like philosophy it involves the challenge of bridging the gap between experts and nonexperts. Regulations coming out of the Cures Act prohibit “information blocking,” and in effect require that test results and doctors’ notes be released immediately to patient portals. In practice, this means that patients often see results before their doctors do and are presented with a fire hose of raw data, shorn of context. And yet on many patient portals, because of algorithms that offer only a binary distinction between normal and abnormal, lab values that stray a meaningless half percentage point out of range are labelled as “abnormal” along with results that have grave portent. Doctors’ ubiquitous use of multicomponent lab panels, which bundle together many different tests, virtually guarantees that every patient receives at least one “abnormal,” which, in our hospital’s system, is written in blood-red letters with garish yellow highlighting, plus an exclamation point, just in case the takeaway hasn’t been fully conveyed. Even seemingly straightforward yes-or-no results, such as those from H.I.V. or covid tests, can be easily misinterpreted, since false positives and false negatives occur depending on the course of the disease and the operating characteristics of the test. Meanwhile, results such as CT scans and MRIs often contain paragraphs’ worth of information that looks alarming but isn’t. Try reading the CT report of your lumbar spine and not coming away convinced that you’ll be paralyzed for life.

Patients turn from the portal to Google, then flood doctors’ in-boxes with messages, panicked by the possibility that they have cancer or multiple sclerosis or any of an assortment of inglorious diseases that appear on the Internet grossly out of proportion to their actual prevalence. But some abnormal test results will indeed be real, and some will be seriously concerning. In our current transparent system, tests showing brain atrophy or liver masses are released to the portal in the same manner as cholesterol levels and blood-pressure readings. For some patients, discovering these results on their own can be empowering; for others, it can be catastrophic. In one devastating stretch of twenty-four hours, two of my patients learned of their cancers’ metastatic reappearance by way of the portal. Their in-boxes pinged with new test results; they read them before either their oncologist or I had even seen the scans, let alone called.

Each morning, I log into our medical-record system with a certain amount of dread. Which will I tackle first—the torrent of test results or the deluge of responses it’s inspired? Some patients will have to be talked down from a ledge over a minor lab abnormality. Far worse is the serious result to which I haven’t yet had time to formulate a response: a biopsy result for which I’ve not yet tracked down the appropriate specialist to sketch a road map, or an adrenal mass on a CT scan for which I’ve not yet orchestrated the intricate endocrine evaluation. In the past, I’d do the legwork before I called the patient. Now that buffer is gone, and I am pressured to act immediately: the patient has seen the result, and further delay would be unconscionable. This timbre of rush imperils thoughtful analysis, and I worry incessantly about missteps…

It's a paywalled long-read, and well worth your time.

Long-time Danielle Ofri FanBoy here.

Her most recent book.

CHAPTER ONE
JUMBO JETS CRASHING

“Is this really true?” my editor at Beacon Press emailed me skeptically. It was a spring afternoon in 2016 and she had tacked on an article from the British Medical Journal (BMJ) that caused headlines in the mainstream media (though it stirred up a healthy dose of criticism in academic circles).1 Medical error, the article concluded, was the third-leading cause of death in the United States.

I floundered for an answer to her question, and not just because I hadn’t been keeping up with the medical journals that pile up relentlessly in my clinic, in my mailbox, in my inbox, and, okay, even in my bathroom. I floundered because I genuinely could not answer her question. Third-leading cause of death? Really? Did medical error really beat out breast cancer, stroke, Alzheimer’s disease, accidents, diabetes, and pneumonia?

As a practicing internist for the past twenty-five years at Bellevue Hospital, one of the largest and busiest hospitals in the United States, I feel as though I see a reasonable cross-section of medicine today. My patients suffer overwhelmingly from the ailments of a twenty-first-century “developed” society—obesity, diabetes, heart disease, hypertension, cancer.

If medical error is the third-leading cause of death, then I should be seeing it all the time, right? I should be hearing about it from friends and family. If medical error clocks in just after heart disease and cancer as a killer, it should be part of my everyday medical experience.

But it isn’t.

Or at least it doesn’t feel that way.

I have witnessed medical errors, of course, and I’ve certainly made my share of them. I’ve heard bone-chilling tales in hospital corridors and read shocking, heart-wrenching stories in the media. Yet these all feel like exceptions—rare and horrible. These deaths don’t pop up in my clinical practice with a frequency remotely near that of congestive heart failure, lung cancer, or emphysema.

Yet the data keep coming. From the first Institute of Medicine report in 19992 that estimated 44,000 to 98,000 deaths per year from medical error to this BMJ analysis suggesting upward of 250,000 deaths per year—medical error seems on the verge of a public health emergency. Even if the numbers aren’t completely precise—the methodologies of these papers have been challenged—researchers are in agreement that the number of errors is not small at all.

Are the data wrong? Or am I wrong?

Am I—and most medical staff—simply not seeing this epidemic? Are we biased? In denial? Are we clinicians killing our patients at an unprecedented rate and somehow remaining blithely unaware? If that’s really the case, perhaps we should take down our collective shingles and spare our patients the damage. We could just tack a note onto the door: “Eat quinoa and beans. Take the stairs. Stay away from the healthcare system.”

While the “third leading cause of death” claim is likely an overstatement, there is definitely a yawning gap between the published statistics of medical error and the experience of the everyday clinician. And then there are the experiences of the everyday patient, which also diverge from the data but in different ways.

As a practicing physician—and an occasional patient myself—I feel I have to get to the bottom of this. What I experience and what the published data conjecture seem at complete odds. One of us is calling it wrong, and my goal is to find out who…

Ofri, Danielle (2020-03-31T23:58:59.000). When We Do Harm. Beacon Press. Kindle Edition.

FanBoy.
 

After QIO retirement, I continued on for several years covering the (mostly Silicon Valley) Health IT startup space. But, now I'm just a cranky retired 77 yr-old patient with Parkinson's, no longer up-to-date competent to opine on the status of EHRs.

 

I sure had my fun. Here as well. “There’s a whole lot of things that I never done, but I ain’t never had too much fun.” For example, from a prior post in June 2019:

 

CLINIC MONKEY ERRATUM

Back when I was working for the HealthInsight REC ("Meaningful Use" program) I routinely bit the hand that fed me when I felt it was warranted. Among other irreverent things, I posted a spoof "Certified EHR" site I called "Clinic Monkey" (tangentially riffing on Survey Monkey, which we used all the time).

When I put it up I embedded "under the hood" an mp3 autoplay endless loop file of ambulance sirens and jungle critters screeching and yacking, for comic effect. No longer works in Safari. I think it's an html thing, no longer supporting the old legacy "embed code." Whatever. It was funny.

One morning I got to the office and found a toy set of simian "office workers" on my chair. One of my colleagues had bought it for me. A Clinic Monkey admirer, no doubt.

Off to the garage I go forthwith after work.

BobbyG's on-the-fly Dollar Store photoshoot cyc.

Yeah, I have an Attitude. One frowned upon by the Really Serious (and snark-challenged) People. It's OK to lighten up.

 

ALL JOKING ASIDE: DR. OFRI ON THE EMR

…One morning in my clinic, as I was seeing my first patient of the day, I noticed that the electronic medical record (EMR) looked a little different. Apparently there had been some sort of “rollout” of various updates in the wee hours of the night, and now a bunch of minor things were out of order and tripping me up.

My fingers automatically knew, for example, that Spanish was #41 when it came to the language spoken by the patient, since Spanish is our most common second language. Knowing the number by heart saved me the aggravation of scrolling through the whole list. But somehow in this rollout, another language had been added, bumping Spanish to #42. My fingers still went to #41, though, so every patient that day came out speaking Serbian.

And then all of sudden, three brand-new fields popped up that I’d never seen before: latex allergies, food allergies, and environmental allergies. We’d always had a required allergies field, and you could enter any kind of allergy—medication allergies, food allergies—or even free-text other kinds of allergies (a few times I’d been tempted to write in “EMR allergy,” but I restrained myself). But now these three new required fields popped up, and they each demanded my attention.

Now, it’s not that I think latex, food, and environmental allergies are unimportant, but their inopportune debut made a hard day even harder. I typically type most of my notes after the patient has left the room, to avoid having the computer be the focus of a visit. So now I would encounter those new fields and face the prospect of sprinting out to flag down patients before they entered the elevator, hollering incoherently about latex gloves, kiwi fruits, and cat dander, smack in the middle of a day when everyone was suddenly speaking Serbian.

I recognized that these additional fields in the EMR were interventions to prevent medical error. They were placed there as part of a well-meaning effort on behalf of our patients to avoid using latex gloves if they were allergic or to make sure the staff didn’t inadvertently deck the halls with boughs of ragweed during holiday season. But I found myself incensed by the whole process. The hospital was already using latex-free gloves, so the potential yield on that effort was distinctly low. But filling in these boxes took away time from focusing on things such as diabetes and heart disease that were really posing threats to my patients’ safety.

You can be sure there was a 100% compliance rate on this effort—it was a required field, so no doctor could close out a note until something had been entered. Somewhere in some office there was a midlevel manager proudly reporting to his supervisor that the medical staff was “100% compliant.” Did this effort actually advance the cause of patient safety in our hospital? I highly doubt it, since nearly everyone gave up in frustration and just clicked “no” to the latex allergy question (and faked the food and environmental allergy questions as well).

At the time I viewed this whole episode as just one more EMR annoyance and one more example of bumbling administration. But after reviewing the experience with checklists, I can see it as an implementation disaster. It was a laudable patient-safety intervention that lacked even the slightest attention to implementation. We had not been given any heads-up that we’d have to start asking patients about latex allergies or that our Hispanic patients would start yammering away in Serbian. No one thought through the unintended consequences of rejiggering the workflow that hundreds of doctors were doing automatically. No one thought about the time it would take. No one seemed to weigh the potential value of this intervention in a hospital that does not use latex gloves. (Yes, there are some latex catheters, but gloves are the overwhelming source of latex.) No one asked whether the juice was worth the squeeze.

This approach to allergies illustrates how checklists can become victims of their own success. Once you start checklisting everything, it devolves into checklist overload. One of the reasons the central-line and surgical checklists worked well was that they were the only checklists on the block. Once you have dozens in play, doctors and nurses can’t cope. There are so many things to check off that you can hardly take care of your patients. Everyone just checks off everything to make it all go away. It’s not necessarily a deliberate gaming of the system; it’s a survival mechanism. A number of lessons can be gleaned from these initial forays into decreasing medical error and improving patient safety. One is that you have to address the system as a whole; piecemeal efforts get you only so far. Another is that if you overemphasize “100% compliance,” you will most certainly end up with people gaming the system, even if they aren’t doing it with malicious intent… [Ofri, pp. 13-15]

Ouch. Yeah. I know this kind of stuff all too well.

__________