There are significant bodies of research dedicated to the conversations around radiology error types. These diagnostic errors are the most common cause of malpractice suits against radiologists (75%) and the average error rate is between 3-5% per year. That’s a global average of 40 million diagnostic errors annually. Multiple factors influence these statistics, ranging from the internal – cognitive – to the external – time and technology.
What’s needed is a pathway through the challenges that not only supports the radiologist but provides a deeper and richer framework within which they can operate more effectively. Ultimately, the human factor is critical in ensuring the diagnosis is made correctly but the inordinate pressure on people has to be mitigated or removed to ensure this diagnosis is accurate.
The cognitive causes of diagnostic errors in radiology
A paper released by Degnan et al in 2018 – Perceptual and Interpretive Error in Diagnostic Radiology – Causes and Potential Solutions – stated that: “…there has been increasing recognition of the limits of human image perception and other human factors and greater acknowledgement of the role of the radiologist’s environment in increasing the risk of error.”
There are numerous elements that make up the cognitive causes of error in radiology. These include bias, expertise, state of mind, and time. Cognitive bias is, of course, a complex web of brain decision-making patterns that can potentially influence how any given image is analyzed and interpreted.
It has been discussed and analyzed to the edge of understanding and yet there is no clear map that allows for any one person to identify their own cognitive bias or catch it when it affects performance. The brain has its limitations and cognitive bias unconsciously affects the way people think, react and behave. Anything from self-serving bias, to fundamental attribution error, to confirmation bias can influence the practitioner.
This is further influenced by factors such as mood, illness, tiredness. The sheer volume of images that the practitioner is expected to analyze on any given day has increased dramatically over the past few years. According to a study published in 2015, radiologists are expected to interpret one image every 3-4 seconds just to keep up with their workloads. If that pressure was driven by the rise in technology-driven scans and accessibility nearly five years ago, imagine what it is today? Image volumes have scaled to the point where it is unrealistic to expect the radiologist to keep up.
In addition to bias and personal factors, research has found that around 60-80% of interpretive errors in radiology are caused by perceptual errors. These errors influenced by attention and perception are complex to identify and control. Human interpretation of anything is prone to this type of diagnostic error and, as research has indicated, the layers of inattentional blindness that influence error and decision making cannot be entirely prevented.
Causes of diagnostic errors in radiology
“Identifying contributing factors is one of the keys to developing interventions that reduce or mitigate diagnostic errors.” – Fundamentals of Diagnostic Error in Imagine, Jason et al.
The issues that influence diagnostic errors in radiology are not confined to the practitioner alone. There are significant system errors that can further impact on diagnosis and patient care as well. These system-related factors can account for up to 65% of radiology errors types and include factors such as teamwork, communication, technical failures, equipment failures, processes, policies and procedures.
Many of these factors are interlinked and have a knock-on effect with regards to efficiency and accuracy of diagnosis, and the communication of that diagnosis to the relevant treating practitioner. They also influence the human factors outlined above. Poor internal process, policy and procedure can play a significant role in practitioner mood, capability and fatigue. The same can be said for the environment within which they operate and the technology with which they work. Poor lighting, levels of work, speed of reading times, length of shifts – each of these factors directly affect how well any radiologist operates.
Within all this, it is critical to always consider the impact of fatigue on practitioner and radiology mistakes. Medical practitioners are under constant pressure to perform to levels of accuracy that are often not anticipated in other roles. This pressure is not only driven by image backlog and system delays, but by the fact that life can potentially hang on an incorrect diagnosis. The above research pointed to several sub-categories that fall within the fatigue arena: visual fatigue, decision fatigue and, of course, physical fatigue. Each of these has to be considered when assessing the common errors radiologists make and mitigating them.
Looking to the future: addressing some of the symptoms
There are numerous strategies, policies and systems that have been put in play to support the radiologist and to minimize diagnostic errors. These span anything from structured reporting systems to peer review to refined processes and expectations. One of the most dynamic solutions lies within the realm of technology. While improvements in capability and quality have been one of the primary causes of radiologist fatigue and stress – the influx of images from MRI, CAT, X-Ray et al has become so prolific that these are the reason why radiologists are scanning more and more images per second than ever before – they are also the cure.
One area within which technology has played a significant role in reducing radiology mistakes is in the introduction of artificial intelligence (AI). While some have touted it as that which shall replace the radiologist, the reality is far more interesting. With AI, radiologists are handed a tool that allows them to fine-tune the process, reduce diagnostic errors, improve decision making and minimize admin. The always-on functionality in solutions such as those developed by Aidoc, ensure that all images are constantly assessed by an AI system, one that is customized to detect all anomalies and ensure they are seen by the relevant practitioner. AI has the potential to become the precise surgical instrument that, when wielded by the right practitioner, saves time, lives and diagnostic errors.