How to Manage Mistakes in Radiology

August 8, 2019

Humans are error-prone, but in medicine that can spell disaster. How do you recover?

Twenty years ago, the Institute of Medicine published its seminal report, To Err Is Human, examining the number of medical errors that occur annually in the United States and the root causes behind them.

Two decades later, all of healthcare-including radiology-is still wrestling with this problem and how to minimize it.

“Even with the tremendous advancements in radiology science, an increased understanding of  what disease looks like, and greater education of radiology knowledge, we still have a persistent error rate that’s at least 3%,” says Michael Bruno, MD, vice chair for quality and chief of emergency radiology at Penn State University. “It’s very stubborn and won’t go away.”

With mistakes still occurring in the industry, providers need to know how to address them and, potentially, minimize the risk that they’ll happen again in the future.

Why mistakes happen

Part of the difficulty in combatting errors, Bruno says, is that it’s impossible to stop them from occurring overall. Radiologists are human, and each one will eventually make a mistake. Reducing that 3% error rate is also hard because mistakes happen for a wide variety of reasons.

According to Richard Gunderman, MD, PhD, Chancellor’s Professor at Indiana University School of Medicine, fatigue from working too long or too intensely can cause radiologists to make mistakes, or they could be distracted by phone calls or frequent interruptions. In some cases, a radiologist could make errors because they’re asked to interpret and render a diagnosis on an abnormality that lies just outside their zone of expertise.

Related article: What to Do When Facing a Medical Malpractice Lawsuit

Under these circumstances, radiologists can make several types of errors. While rare, Bruno says, errors of omission-a missed diagnosis-happen when providers simply don’t see an abnormality because it’s located in an unusual spot or it appears in only one image in the series.

Other mistakes, called “satisfaction of search” errors, occur when a provider overlooks one of two findings present in an image. Even though providers continue to analyze an image after identifying one finding he says, a second, whether obvious or inconspicuous, can be missed.

And, sometimes the fault lies with the organization’s systems and processes rather than with the individual radiologist. Faulty protocols can set a provider up for failure even under the best of circumstances.

What to do when errors occur

After identifying a mistake, whether it’s a diagnostic or reporting error, a provider’s first responsibility is ensuring the patient receives the appropriate care, says Stephen Brown, MD, associate professor of radiology at Boston Children’s Hospital and Harvard Medical School.

After that, he says, start communicating.

“Each institution should have laid out an articulation process within the department or the organization of how front-line providers can communicate when mistakes happen,” he says. “Maybe it’s going to the department leader, risk manager, quality and safety, or legal folks. Every institution must figure this out.”

Having this process in place can help identify the cause of the error, inform a root cause analysis, and help leaders determine whether informing the patient and family is necessary.

But, addressing mistakes means more than reporting an error and addressing it with the family and patient, says Alexander Norbash, MD, chair and professor of radiology at the University of California San Diego Health. It also presents a learning opportunity. In order for learning to occur, though, institutions must resist the initial impulse to assign blame and levy punishment.

“We have to start by creating a culture that values our being forthcoming and values our desire to learn from mistakes,” he says. “Mistakes are a part of life, but we have to learn so we don’t repeat them.”

In this vein, there’s a growing number of healthcare organizations that are implementing peer learning systems. Rather than being punitive for overlooked findings, these groups create collective opportunities to educate providers using examples of diagnostic catches, as well as misses. By reviewing these cases together, providers can create a corrective action plan that addresses problems when they occur, he adds.

It’s then incumbent upon the individual provider to remember the lessons learned and apply them in practice, Bruno explains. Doing so can help radiologists remember locations they’re likely to bypass or keep them from making too many assumptions about an image.

“There’s a very simple fix for these errors of thinking-these cognitive biases,” he says. “You force yourself to ask really open-ended questions, including ‘what else, how else, where else’ could a finding be. You force your mind back open again because closing it to diagnostic possibilities sometimes gets us in trouble.”

What to avoid

Although it can be tempting, Gunderman cautions against denying a mistake or sweeping it under the rug. Never falsify a medical record, he says.

“It might be tempting to take steps to ensure no one finds out about your mistake, but if anything comes to light later, the consequences could be dire, especially if it appears you hid something or changed something to avoid detection,” he says. “No one can learn if we don’t own up to our mistakes.”

It’s also important, Norbash says, not to succumb to a hierarchical structure that could inhibit providers’ abilities to speak up when they see potential mistakes. Organizations that rely heavily of hierarchy or a structure of silos can experience more errors because no one feels free to draw attention to brewing problems.

“It’s a challenge in healthcare because we are largely hierarchical, especially in academic environments,” he says. “But, if I’m making a mistake, I hope that someone would point it out to me before it accelerates into consequences that are negative.”

But, if errors do occur, and they lead to a conversation with a patient or the family, there are also several things a radiologist should avoid, Brown said. In these instances, he recommends:

  • Not talking to a patient or the family alone

  • Not talking to a patient or the family before talking with institutional leadership and receiving coaching or guidance

  • Not broaching a conversation without planning what to say

  • Not speculating or lying, but providing as much detail as possible

Potential assistance

With many errors revolving around a missed finding, artificial intelligence could be effective in providing a second set of eyes, Bruno says. Existing research shows double-reading reduces the number of missed findings significantly because it’s unlikely two readers will overlook the same things. In the past, radiologists functioned as both readers, but doing so is more difficult now due to increased patient volume and more stringent time constraints.

“In this healthcare environment, it’s possible that artificial intelligence could be the double-reader of the future,” he says. “A lot of algorithms have been developed to pick up on images and have 70% accuracy. That may be good enough for the double read.”

Ultimately, radiologists shouldn’t view mistakes as a torpedo to their careers. When errors happen, he says, they shouldn’t condemn themselves. Instead, they should practice self-compassion and understand their error wasn’t intentional.

“We are all subject to our biology and the human condition,” Bruno says. “Our brains evolved for something different than radiology, but we’re using our eyes remarkably well for diagnosis. But, they’re not perfect for that application. We have to understand that and have compassion for ourselves.”