One of the more thought-provoking articles I’ve read this week was in the Journal of the American Medical Informatics Association. Of course, the title caught my eye: “Do electronic health record systems ‘dumb down’ clinicians?” The abstract was compelling as well, although I was able to read it only after entirely too many clicks were needed despite my AMIA membership.
The abstract discusses a panel that was held at the 2021 AMIA Symposium and sponsored by the American College of Medical Informatics. The panel sought to answer the title question, looking at how the incorporation of EHRs might be negatively impacting care delivery. Discussions centered on how less-than-optimal EHR workflows might impact clinician efficiency, thought processes, and knowledge both during system use and during the longer term.
In broader coverage of the topic, the journal goes beyond the panel discussion, starting with the evolution of EHRs including both homegrown and commercially developed products. It notes that although early EHRs improved safety and efficiency compared to paper-based systems, “several critical perspectives were lost.”
The authors note that data showing impacts on patient outcomes were lacking and that most studies have focused on the processes involved in delivering care. Since early systems were concentrated among a small number of academic medical centers, there wasn’t much portability across institutions. They go on to explain how the Meaningful Use program and the HITECH Act of 2009 incentivized provider organizations to not only expand EHR use in the marketplace. but also to focus on a core set of functions that would lay the groundwork for broader improvements in care.
Having been part of an organization that was already knee-deep in EHR implementation and adoption before Meaningful Use came along, I can attest that it actually slowed us down, because we had to focus on ensuring that prescribed workflows were followed versus being able to customize or configure workflows that worked best for our clinicians. It’s validating that the article notes some of the same negatives that were created in the name of progress.
A series of great quotes are included in the piece. They’re attributed to a New England Journal of Medicine article that was penned by John Halamka and Micky Tripathi in 2017. They are quoted as saying that the HITECH Act had some less-than-ideal consequences. “We lost the hearts and minds of clinicians … We tried to drive cultural change with legislation. In a sense, we gave clinicians suboptimal cars, didn’t build roads, and then blamed them for not driving.” One of my favorite family physicians and clinical informaticists, Jan Lee MD, used to refer to this as “paving the cow paths” when physicians actually needed high-speed roads with no obstacles.
The article goes on to discuss in detail how EHRs might impair clinicians in the short term. Although there are many beneficial features in modern systems, including allergy checking, order sets, and often a treasure trove of patient information, there are also interruptive alerts and distractions. Sometimes users are unwilling to question information supplied by the EHR, and alert fatigue can cause users to ignore warnings and alerts that might in fact be useful. The article gave specific examples, ranging from hundreds of thousands to millions of interruptive alerts where only a fraction (2-4%) were accepted. The way that EHR notes are organized can obscure the details of a patient’s situation. The use of copy-and-paste functionality in progress notes was specifically called out as potentially misleading and dangerous.
As far as long-term impacts of EHR use, the authors noted that standardization of EHR documentation has led to less-granular terms being used during the creation of History and Physical documents. At one institution, EHR templates reduced the possible descriptors from 1,800 to 360, meaning that some clinician documentation might be hidden from view by other clinicians. An unintended consequence of this might be the shrinking of clinical vocabularies used by medical trainees. As those trainees become faculty, and older faculty members retire, the broader vocabularies are ultimately lost, and notes become less optimal. Additionally, after using templates to order certain treatments, such as total parenteral nutrition, clinicians lost the ability to order these treatments manually. The authors note that this can be dangerous during system outages or when clinicians move to a less-automated environment.
The panel also discussed solutions, which fell into four general categories: institutional and end-user readiness and competency; EHR design and capabilities; regulatory policies and healthcare system-vendor partnerships; and decoupling clinical documentation from billing and regulatory requirements so that clinical notes contain only that information necessary to care for the patient.
I think you would be hard-pressed to find a clinician in the US that wasn’t in support of the latter suggestion. Technology could be a great booster of the latter as well. I’d much rather have a photo of a rash in a chart than a rambling description of someone’s idea of a “lace-like reticular rash with mild to moderate erythema and occasional popular features.” Unfortunately, in many situations, the words will get you paid, but a photo will not.
As far as the other categories, the panel called for improved EHR training, expanded downtime simulations, and greater incorporation of learning about EHRs during undergraduate and post-graduate medical education. Medical students and residents need to understand the “why” behind various parts of the EHR as much as they understand the data that they’re keying in. They call for greater clinician involvement in the design and validation of EHR systems and improvements to alert messages to ensure that such interruptions are clinically important and obtain enough clinically relevant information for clinicians to take action.
It will be interesting to see how EHRs evolve over the next five to seven years. More interesting will be assessing the approaches taken by health systems in how they implement and optimize EHRs. I still see far too many organizations that think that installing an EHR is some kind of “set it and forget it” process.
Physicians constantly complain about the impact of EHR upgrades, but in the last conversation I was in about that topic, not a single physician admitted to having seen any kind of upgrade documentation or educational materials that told them what to expect or what benefits might arise. It feels like those kinds of communications might be casualties of the rampant understaffing I see in many organizations. Gone are the days when we used to send a member of the EHR implementation to every office to make sure clinicians knew what to expect and that their questions were answered. I’m sure the materials are probably out there on some intranet site that physicians have long forgotten how to access.
On the whole, I don’t agree with the premise that EHRs have made us dumber, but I do think they have impacted our workflows tremendously, and not always for the better. There certainly is room for improvement and evolution of technology, but everything comes at a cost. When hospitals are trying to figure out how to keep beds staffed, they are less concerned about things like EHR adoption or end-user satisfaction.
Do you think EHRs have dumbed-down clinical practice? Leave a comment or email me.
Email Dr. Jayne.