Underappreciated Harms of the Electronic Medical Record
Part I: Electronic doctor/patient communication
There is a classic saying in Chicago, “There is no bad weather, just poorly chosen outfits.” I am sure this is a saying everywhere.1 Recently, “It’s not the EMR (electronic medical record) that is bad, it is how we use it” has been echoing in my mind. Why? Maybe it is because I recently listened to the Acquired podcast episode about Epic. (I finish that podcast in awe of every company they cover). Maybe it is because of my chronic contrarianism, my need to disagree with people who say things I reflexively agree with.
Colleague: God, Epic is ruining my life!
Me: No, I love Epic!2
I like to think that the real reason I defend the EMR is because with 30 years of experience, I know that EMR has made things better. I remember a time when:
All notes were handwritten.
I often had to practice without medical records.
When I did have a record, it had to come from the medical record department and usually arrived hours or days after the patient. (Sometimes it took even longer, when the chart had to be retrieved from “deep storage”.3)
Every prescription had to be handwritten or called into a pharmacy.
Viewing an x-ray meant waiting at the window of the “file room.” File rooms, at every hospital, were staffed by the grumpiest of humans – the result, not doubt, of terrible working conditions and having to deal with impatient (inpatient or outpatient) doctors all-the-damn-time.
Now, I am not an idiot, I know our EMR is not perfect. The fact that these systems were built as billing tools, as well as record repositories, has always made them less than ideal. Even here, though, I’d argue that their shortcomings have more to do with billing requirements than the EMR itself.
Much of what truly bothers us (doctors and nurses) about the EMR, is our fault. What irritates me most, is the poor-quality of notes. Notes overflow with copied radiology, lab, and pathology reports. Notes swell with incorrect, out of date, or just plain useless information pasted forwarded for generations. Notes often contain no evidence of a reasoned assessment and plan.
All that said, I do think that there are two areas in which the EMR’s intended and realized benefits are overwhelmed by its harms – electronic communication between doctor and patient and immediate access to results. This will be a two-parter; today, electronic communication.
Because so much our communication in the 21st century is electronic, it makes sense that this would be offered within the EMR. Certainly, if doctors and patients are going to communicate electronically, it is better that this happens within the EMR than on millions of mobile phones. And, if the question is simple, requiring neither an appointment nor a time-sensitive response, what could be the harm? The harm comes in two forms, excessive access and providing patients AND doctors the option to avoid conversations — actual, necessary human to human conversations that need to be had.
Excessive Access
It might seem impossible that a patient could have to much access to his or her doctor. Sit down with me when I open epic and you will see what excessive access looks like. Dozens of questions, sent at odd hours, sometimes, seemingly, to satisfy passing curiosities. Questions that might be better asked of parents or friends. (I am commonly alerted, one week after responding that my answer has not been read.) Buried among these questions are appropriate ones and emergencies that should have been attended to when the message was written, not when I open the chart.
We talk about physician burnout; I talk about primary care general internal medicine being a dying field. Here is a part of the problem. Yes, patients need be able to contact their doctors, I have had conversations with nurses telling them, “Don’t feel like you need to protect me from the patients.” The current situation, however, needs to change.
Avoidance of Necessary Conversations
SS is a 78-year-old man. He is in good health other than hypertension, controlled with an angiotensin receptor blocker/diuretic combination. On a Sunday afternoon, while working around the house, he develops lightheadedness and chest pain. Worried he is about to faint; he sits down on his stairs. He has never had similar symptoms. He sits for about 45 minutes until he is feeling better and then walks slowly upstairs and goes to bed.
The next morning, he sends his doctor a MyChart message describing his symptoms. He has a phone number for the doctor’s nurse, which is forwarded to the on-call doctor after hours. He also has his doctor’s cell phone number, to be used in case of emergency.
Patients sometimes use EMR based electronic communication to communicate things that deserve a real conversation. Sometimes they do this so as “not to bother the doctor.” Sometimes they do this because they fear what reporting a symptom will lead to. This was the case with SS. He knew his symptoms warranted immediate attention. He himself was worried about the symptoms. But instead of facing them urgently, he used the MyChart as an out, a way to avoid facing reality.
Doctors are not immune to this type of behavior. They sometimes use electronic communication to avoid a conversation that needs to be had. In my own practice, and while precepting trainees, I have watched electronic conversations continue when someone just needs to pick up the phone.
I know the flaws with this essay. (I am venting here, don’t be too critical). We could do a better job of educating our patients on the appropriate use of EMR communication. We could have better triage of electronic messages. Hell, we could use AI to answer the messages.4 Although it is the users’ who are really the problem here -- rather than the EMR -- I think we should not allow the EMR to lead us into temptation.
Next week, a related rant on patient access to ALL of their information.
Just like one of my most despised comments, “If you don’t like the weather, wait five minutes.”
I had nightmares about deep storage during internship.
I don’t actually think that…
My main concern with the EMR is that it has effectively killed the patient narrative. Patients have been reduced to an age, gender and long list of diagnoses, and consequently, we are technicians of their illnesses.
When I started in nursing 45 year ago, knowing and documenting the "backstory" was part of our responsibility. With consumer confidence in health care at a record low, restoring this information to the EMR would add humanity to the work we do. In the past week, I discovered the following among the inpatients I worked with:
- The 82 yo female being treated for a TIA was a 28 year veteran of the hospitals medical transcription department.
- The 21 yo male with a spontaneous pneumothorax had 9 visitors in his room because it was his graduation day. He had just earned a degree in aerospace engineering.
- The 71 year male who refused to go to a skilled nursing facility for rehab did so because he was the primary caregiver for his wife, who had undiagnosed dementia.
This information is not only interesting to know, but provides talking points for providers, and helps us to know our patients beyond their diagnosis, aiding our informed decision making. I believe it will also help reduce burnout, as we deepen the connections that led us into healthcare in the first place.
The VA system is wise to this issue and has initiated the VA My Life, My Story program, where a veterans life history is written up in a concise narrative and then integrated into their EMR. Their stated goal is to "humanize the patient experience". This history is the first thing visible to all providers as they enter the EMR. We can, and should, do this too.
I am certain this will be covered in a future post, but the EMR allows entry errors to be persistent. My mother experienced, (suffered), a periprosthetic fracture after hip replacement, requiring 12 weeks zero weight bearing. If I can get time detailing that scenario warrants its own post.
The morning after the surgery a perky therapist popped into her room, where I was camped out, announced, "OK, 50% weight toe touch today".
I corrected, "Nope, zero weight bearing for 12 weeks".
After arguing with her for some minutes, she confirmed I was accurate, and we determined that the weight bearing section of the record which had a weight bearing drop down selection button was "mis-clicked".
Had I not been there my mother could have destroyed the repair and potentially lost her leg.
That same error propagated through to the SNF, and 2 other facilities as she convalesced.
I have comment in my talks that my real job is keeping the health system from killing my mom, which gets a laugh, until they see I am not laughing and realize I am serious.
But I also agree that immediate access is also a huge improvement over shipping papers here and there.