I think you're referring to this: https://www.wired.com/2015/03/how-technology-led-a-hospital-to-give-a-patient-38-times-his-dosage/ It's a fascinating example of the swiss cheese effect, and should be required…
Curbside Consult with Dr. Jayne 4/25/22
This week’s Monday Morning Update discussed an EHR outage at two of Tenet Health’s hospitals in South Florida. Apparently Tenet didn’t like the media coverage from WPTV and suggested that a story about the downtime be removed. The story covered a patient’s concern about potential medical errors during the outage, with Tenet complaining that suggesting a downtime could result in medical errors is “preposterous.” As a physician who has been in the trenches for more than two decades and who has been through enough EHR downtimes that I couldn’t begin to count, I’m speechless at the thought that downtimes aren’t problematic.
I’ve been up close and personal with a downtime-related medical error in my career, and the situation certainly would have been different had the EHR been online. It was a bit of a perfect storm-type situation. First, I was a relatively new hire still getting used to the processes at a new urgent care employer. Second, due to someone calling out sick, I had been rescheduled from my usual location to a different site, which added a baseline level of stress to the day since I was working with an unfamiliar team. Third, due to a pre-existing diagnosis, the patient I was seeing for a fever was unable to contribute to the history of the presenting problem and was combative during exam, which is tremendously stressful.
After my initial attempts at history-taking with family members and a brief exam during which I detected no emergent problems, I ordered some laboratory studies and moved on to see other patients. When I left the patient, he was pacing around the room and showed no signs of being in distress or in pain.
At some point before the tests were resulted, the EHR went down. As a new employee, I was unfamiliar with the downtime process, but knew there should be one. I asked if there was a downtime binder or how we were supposed to handle it. The clinical team lead was resistant to instituting downtime procedures, giving excuses along the lines of “the EHR usually comes back in a few minutes” and “it really makes a lot of paperwork if we try to go to downtime procedures.” Knowing that creating paperwork is the point of a downtime procedure, I pulled some paper from the printer and began writing my own SOAP notes and documenting what I could.
I remember having probably half a dozen patients on the board that I was seeing. I tried to move them through the process while begging for a paper prescription pad so I could write discharge medications and keeping a clipboard with sticky notes on it as a tracking board to help me remember what patient was in what room. Lab results were being printed from the instruments on little slips of receipt paper rather than flowing through the interface to the EHR. The results were in an unfamiliar format, with the individual tests being out of order within a panel and the reference ranges being difficult to read. Despite the downtime, the staff continued to room new patients and expected us to move forward. I was surprised by that – none of the patients were emergent, and as a walk-in urgent care center it would have been within our rights according to state regulations to stop taking new patients.
I was managing patients the best I could and providing written discharge instructions that I was typing in Microsoft Word and printing two copies so we could scan them later. For my patient who had a fever, there wasn’t anything apparent on the exam or on my review of the labs that could have been causing it, so I recommended close follow-up at home and told them what to look for. This was during the usual season for viral illnesses, and in many patients, the illnesses begin with fever but don’t always declare themselves with other symptoms for a day or more. Since the patient couldn’t describe his symptoms and the exam was difficult, I didn’t suspect anything serious.
Every one of my hand-typed discharge instructions included my best recollection of the practice’s standard disclaimer, which would have been automatically applied by the EHR had it been online. It was something along the lines of “Your examination at XYZ Health today is limited by the capabilities of this urgent care facility, which does not include advanced imaging or moderate complexity laboratory testing. If at any time you feel your condition is worsening, we recommend that you be re-evaluated at the nearest hospital Emergency Department.” I reviewed this instruction with the patient (who could not verbalize understanding) and his adult caregiver, who said she understood.
Two days later, I was called before the practice’s owner and yelled at for “letting someone walk out of here with those abnormal labs,” because by that point, the patient ended up having a significant abdominal infection that required surgical drainage. I explained that at the time I saw the patient they had no features of a serious abdominal process and reviewed the examination that I had documented on my handwritten SOAP note. I was then asked to review the documentation that had been keyed into the EHR after the downtime ended. There it was, in bright red — an abnormal lab value. I had missed it when looking at the receipt-paper printout in an unfamiliar format and with confusing reference ranges. It wasn’t a critical value, but it was abnormal enough that it might have made me think about additional potential diagnoses, even if the physical exam didn’t point me towards an abdominal cause for the fever.
In reviewing the patient’s course, he hadn’t been taken to the emergency department for more than 12 hours after I had seen him, which wasn’t a guarantee that the process requiring surgery was yet present when I evaluated him. Usually if patients have a significant infection in their abdomen, they’re not likely to be pacing around the room – they are completely still on the exam table, and you can hardly touch them. Still, I couldn’t help but second guess the factors that went into my care of the patient – the unfamiliar staff, the new location, the downtime, and the patient’s individual characteristics and presentation.
I explained to the now shouting and red-faced CEO that this wasn’t a normal visit under normal circumstances and that I didn’t have the luxury of having the abnormal lab highlighted in red in the EHR during the visit because there wasn’t an EHR during the visit. He seemed surprised to hear that. Even after he admitted that the EHR downtime was an issue and there’s to way to know if my care contributed to the problem, I agonized over the situation. Several peers reviewed the chart and had no additional suggestions, but that certainly didn’t make me feel any better.
The bottom line here is that EHR outages are difficult. They raise the potential for medical errors in a number of ways. They add stress to already overwhelmed staff. They remove safety checks that we’ve come to rely on. They increase cognitive load as clinicians look at data in unfamiliar formats. They reintroduce illegible handwriting to the environment. They also create time pressures when they end and staff is forced to key in data while they proceed forward with their usual assigned tasks.
I’m fortunate that the patient in this scenario had an uncomplicated hospital stay and there were no long-term consequences of the event, either for him or for those who cared for him. However, the long-term psychological impact on me as a physician makes me never want to encounter another EHR downtime again.
What do you think about Tenet’s comments regarding EHR downtimes? Leave a comment or email me.
Email Dr. Jayne.
As a patient, I was hesitant when the EHR made its way into my PCP’s office, because it seemed like the clinical staff spent more time looking at the laptop than they did looking at me. Now that technology has become a mainstay in most healthcare settings, I shudder to be a patient in a facility undergoing an outage. When the staff (especially the younger ones who have no “pre-EHR” experience on which to fall back) suddenly have to start doing things manually that technology normally does for them, it rachets the stress level up exponentially. The added stress of unfamiliar processes is not good for patient outcomes.
TL;DR – Tenet’s PR team needs to shut up and take their lumps.
Absolutely correct Doctor. A down system puts everyone at risk. It’s undeniable.
It’s scary to realize there is a CEO out there somewhere who doesn’t look at that situation and ask “What contributed to a physician missing an abnormal lab and how can we make sure it doesn’t happen again?” Yelling at someone rarely fixes a broken process.
Yelling at coworkers, subordinates, and customers does “fix” the “broken process” of “people coming to you for help with their problems”. People will learn to hide the truth from you if they’re scared of your reaction.
We have known the following for a long time:
-Standard national requirements are needed for when a downtime needs to be called, what needs to be available and what needs to be entered back into the EMR. Leaving it up to each hospital is lunacy.
– Public reporting of hospital downtimes, their cause And impact would create incentives to minimize the frequency and duration.
A downtime is like a plane crash – luckily few have died. Let’s call for standards, reporting and investigation for the sake of our patients and ourselves
Thanks for sharing the war story–a type that often is locked away in the Painful Memory closet when in fact should be hauled out often for discussion until the non-system wises up. In particular, to recognize that the increasing dysfunction in our industry is more and more focused on the CEO’s point of view rather than the clinician’s or –imagine!– the patient’s. It is all about the data and the dollars and not the quality as reflected in good outcomes.
Perhaps a few law suits might change Tenet’s attempt to manipulate the message. . .
Of course EHR downtime could cause medical errors. Less dramatically, but more commonly, it slows down delivery of care, stresses the clinicians, and overloads staff.
The simple fact is, an online system is nearly always better than any paper process that preceded it. Once we get the online system, therefore, we quickly become dependent on it. And downtime processes are a poor second choice (though better than nothing).
Tenet Health is almost certainly in full defensive mode when they make their “preposterous” claim. It’s a bureaucratic reflex reaction. If you actually admit that patient care could suffer… sacre bleu! The world might end, non?
Geez, even the word is “downtime”, which is the opposite of “uptime”. Which would you rather be, down or up?
The most important step for the IT team in the downtime declaration process, is having a process for when to declare that the EHR is down.
“give me another 5 more minutes boss” doesn’t make anything easier for anyone, especially when minutes turn into hours.
I agree with “Jimmy the Greek”: Tenet’s PR team needs to shut up and take their lumps.
I repeat this without remembering who I stole it from, but when people say a downtime introduces no possibility risk, what people hear is that the medical record software offers no clinical value. I doubt that’s the intended message.
Thanks for the example. It helps show the ups and downs of software giving a hand.
The 1990’s are coming back to me… now I remember: The first rule of downtime is that we don’t talk about downtime
What would blow your mind is the number of ambulatory clinics that don’t have Business Continuity Plans in place. there is no binder to consult. It should not be a burden to resort to a well documented BCP, to keep staff and patients safe. I’ve been through outages as an employee in a hospital, and as a vendor for ambulatory clinic EHR, without a good plan in place, chaos will ensue, mistakes will be made. BCPs are not the only answer, however; being prepared and testing your plans, is good preventative medicine.
Personally, I’m sorry that you or any doctor is put in that position. You should feel comfortable in your process for providing care. I’m glad the patient had a good outcome.
I don’t think that downtime itself adds absolute risk to patient care. Just as don’t think practicing a hand to hand martial art like kenpo adds significant risk. But, if I am trained in the use of the EHR, and suddenly I am expected to translate that training to paper charts — I would equate that to that hand to hand martial artist suddenly finding himself on the tournament floor with kendo (japanese sword art) and suddenly expecting to compete at the same level.
I do believe that paper charts are less safe, for a multitude of reasons, but expecting someone to use an EHR then suddenly go to paper without regular and consistent training is itself a patient safety event.
Also, an organization that is handling X volume using an EHR should not expect to be able to continue with that volume of traffic if they are under disaster recovery procedures. The fact that going to paper is a DR function, should also indicate that there is significant risk added.
Tenant’s claim that there is no impact is itself ludicrous.