Cognitive errors underlie most diagnostic errors made in the course of clinical decision-making in the emergency department, according to Pat Croskerry, MD, PhD. These range from anchoring and heuristics to one recently coined by Irish emergency physician Andy Neill, MD: location-based decision-making.
It's a new name, but realistically, it's a relatively familiar concept that can be generalized to other environments. In the emergency department, however, the idea is applicable and important and possibly life-changing.
“This is something that I've noticed happening to me for as long as I've been doing emergency medicine — that the physical part of the department that you see the patient in has a significant impact on my decision making process,” Dr. Neill wrote in his blog. (www.emergencymedicineireland.com.)
The emergency department has particular characteristics that predispose it to error, or more specifically, that predispose a physician to making a wrong diagnosis. Crowding in emergency departments is not a new issue, but it is one that largely affects location-based decision-making.
“Yes, our emergency departments can physically contain 80 patients in a department with nominated physical bed spaces for 25, but it is magical thinking to believe this has no impact on providing safe, timely, and effective emergency care,” Dr. Neill said in an email. “There's the most simple of things: It's really hard to examine a patient in a chair in a corridor. We often end up doing our best and examining patients in chairs and ambulance trolleys, as it's the only way to even make a start on the patient's workup and treatment. When things like examination are really hard to do, you can see why people take shortcuts.”
The high-stress environment of an emergency department where physicians may be repeatedly inundated with emails and texts and other alerts certainly affects the ability to make sound, appropriate decisions, said Benjamin Sun, MD, an associate professor at the Oregon Health & Science University in Portland.
“There's some indirect evidence of that demonstrating that crowded conditions are associated with worse outcomes, including mortality, increased time to pain medications, increased time to antibiotics,” he said. “And having worked under very stressful situations, I mean, if you're dealing with multiple folks who are critically ill and you're being interrupted constantly, that certainly is going to have an adverse effect on your ability to make the best decisions.”
Triage nurse-physician agreement is another aspect where cognitive biases affect patient care. “I think our mindset is different when you're walking into a room, say, in the fast track, where a patient was initially seen by the triage nurse and thought to be relatively low acuity, versus if you're walking into a resuscitation room where the initial triage evaluation was somebody who's critically ill,” Dr. Sun said. “As physicians, we have to be careful about these built-in biases because if there was an error made at the triage desk, you don't want that initial triage decision to overly influence your own decision-making.”
And the job of triage nurses is no walk in the park, either. “They're basically making a decision based on a very brief evaluation with very limited information, and they're processing many patients in a short period of time. The expectation is not that the triage nurse is always right. That would be impossible; no physician can do that,” Dr. Sun added.
Anand Swaminathan, MD, an assistant professor of emergency medicine at the Ronald O. Perelman Center for Emergency Services in New York City, said a lot of location-based decision-making rests on the triage nurse-physician agreement and anchoring biases. Just like location-based decision-making can stem from several factors, anchoring bias may come from an initial triage evaluation, a patient who's already been given a prior diagnosis, or a patient who's been in the emergency department several times for the same thing. How to control these biases, however, is based on the individual, he said.
“I think some people are very good at it, but I think it can be very difficult even for really seasoned physicians. When I think of this, I usually think of a fast-track area or an urgent-care area versus an emergent-care area just because that's what I deal with on a regular basis. When you're in a fast-track area, you're sort of expected to see a certain number of patients. Nobody makes a guideline, but there's the expectation that you should be moving through patients quicker because they're less complicated. So even if you go into it saying, ‘I'm going to not be biased by that,’ I think it's very difficult because you want to move through that patient to get to the next one,” Dr. Swaminathan said.
Dr. Sun said a physician always has to be on guard for anchoring bias when the same patient repeatedly winds up in the ED. “The next time you see the patient you may think, ‘Well, this is just the same thing as before.’ You're sort of anchoring a prior diagnosis, but it might be something different,” he said.
Cognitive biases aren't so simple, so very black and white, however, and X doesn't always happen because Y, said Robert Wears, MD, PhD. “They're not linearly causal. If they do anything, they set the stage for certain kinds of behavior to emerge. They don't cause them to emerge, but they set the grounds where they can emerge if other things point in that direction,” said Dr. Wears, a research professor at the University of Florida and a visiting professor in the clinical safety research unit at Imperial College in London.
Whatever name is trendy at any given time, there's truth to the idea of location-based decision-making, particularly that decisions are more often gray than not, and decisions must be made without taking shortcuts, no matter how tempting cognitive bias may be otherwise. Why? “When we don't undress people, for example, we mightn't see the rash,” Dr. Neill said.
Access the linksin EMN by reading this on our website or in our free iPad app, both available atwww.EM-News.com. Comments?Write to us firstname.lastname@example.org.