Last month, a resident physician at Harvard Medical School wrote an editorial in JAMA Internal Medicine about the day the electronic health record (EHR) went down. She reported that while most of the hospital staff knew how to deal with a several-hour planned downtime in the middle of the night, this was different because it was unexpected, occurred when the residents should be summarizing the overnight course of their patients, and was going to last an unknown period. Little did this doctor realize that the EHR problem would illustrate the importance of patient-centered care and, indeed, some of the essential aspects of the teamwork that is central to good medicine.
She writes, “How could we assess overnight progress without our [EHR] system? [The residents] eventually came to the same simple answer: Why not directly ask patients about their condition overnight? Or ask the night nurse who spent 12 hours checking their vital signs and documenting patient data, or speak with the day nurse who got a sign-out from the night nurse?” The young physician then concluded that perhaps the routine, daily lab results that she could not access were not as essential as she thought because the C-reactive protein value, for example, may not “have any meaning if a patient with pneumonia feels better and their oxygen requirement has improved.”
As the kids nowadays say, “Preach!” A million years ago when I was a resident and we didn’t have any of these fancy EHRs, we were taught the same lesson. On daily rounds, many an attending would look at us and repeat the same question: how will this proposed lab test or imaging procedure change your treatment of this patient? If you didn’t have a good answer, it was clear that you should not order the test. Even back in the 1990s, when we barely had figured out electricity and whatnot, we were trying to minimize daily labs and the omnipresent chest X-rays because, in general, they were not helpful and were even potentially harmful. (Also, see my blog “De-adopting low-value EHR practices” to emphasize the point that we have low-value routines in clinical informatics as well as in daily patient care.)
The author wrote that the unexpected downtime made her “realize that while the system seems to make our clinical routine convenient, it may not result in increased efficiency or better patient care.” Yes, yes, and yes! If we configure an EHR to simply make it easy and convenient to offer sub-optimal patient care, we have failed. In my book “Designing for Health: The Human-Centered Approach,” we have a chapter titled “Make It Easy to Do the Right Thing.” Allow me to let you, my esteemed reader, in on a little secret: it's not difficult to configure the EHR to make it easy to do the right thing. (Shhhh! Don’t tell your COO or CMO this!) Setting up the EHR this way is a no-brainer. The complicated part? Figuring out what the “right thing” is!
As this Harvard resident discovered, the “right thing” is probably to ensure that it’s not very easy to order daily labs. Hence, we should not add daily BMPs and CBCs to routine admission order sets (as seems to be the case in the EHR that this physician uses every day.) Does this mean that doctors cannot order those tests every day for those patients where it actually makes sense? Of course not! Physicians can always configure such an order, but there is a tiny bit of friction involved. And as Thaler and Sunstein have taught us in their seminal work “Nudge,” a little bit of friction goes a long way.
Finally, this unexpected EHR downtime led the physician-in-training to observe how the conveniences of the technology can detract from human interactions. She noticed that while it’s easy to enter orders on a computer, the need to talk with her hospital colleagues had been severely diminished. As she puts it, the EHR failure “forced [us] to have in-person communications between the members of the care team, which actually turned out to be much more satisfactory.” To this, I say, yay, EHR downtime!
Last year, I wrote about a study that posited that decreased communication among the hospital care team led to increased clinician burnout and decreased well-being. As it turns out, a lot is lost when we don’t communicate with actual humans, but instead put technology between us. To mitigate this problem, the study’s authors recommend establishing norms around EHR communication. In my world, we call this writing (and following) an EHR etiquette guide. What goes into such a document? One example might be the expectation that when seeking a consult from a colleague, the doctor requesting the assistance call the consultant and briefly discuss the case and the specific questions that are outstanding. To be sure, this was the expectation when I first started practicing.
Our resident finished the editorial by reassuring us that everything turned out alright. “We finished rounding earlier than usual, and our plan did not change when the system came back around noon, even with all the missing results now available. Our patient care on that day was the most patient-centered and most collaborative than ever in my 2½ years of residency. It was an epic day.” In my reading, the day wasn’t as epic as the important life lessons that many learned from the EHR downtime: technology can improve the way we work and live, but it can also automate and simplify bad practices that should be minimized, not enhanced.