Increase Your Awareness and Understanding of Radiologic Error
15 strategies to help minimize radiological errors in MRI, CT, and ultrasound imaging.
Editor’s Note: The authors of this article received a certificate of merit at ECR 2017 in Vienna for their research.
Around 250,000 people die each year in the US because of medical errors, according to a study by researchers at John Hopkins University School of Medicine. This number is bigger than those who die from car accidents, breast cancers, or work accidents. In fact, medical error constitutes the third leading cause of death in the United States.
Radiologist Leo Henry Garland (1903 – 1966) was the pioneer in the study of radiologic error. The prevalence of radiologists’ errors does not appear to have changed since it was first estimated in 1960. Today, it remains around 10 to 15 percent . Although some diagnoses are missed because of the limitations of the imaging modality, most of them are attributable to image interpretation.
What constitutes an error?
The definition of what constitutes an error in radiologic interpretation is subject to debate.
Some radiological investigations say that the use of the term “error” is often unsuitable. They state that it is more appropriate to use the word “discrepancies” between a report and a retrospective review of a study. Because of the subjectivity of image interpretation, the definition of error depends on an expert opinion. In that order, an observer makes an error if he or she fails to reach the same conclusion that would be reached by a group of experts; and errors can only arise in the cases where the correct interpretation is not in dispute.
Rad errors range from range of 2 to 20 percent
A review made in 2001 reported that the prevalence of clinically significant errors in radiology was in the range of 2 to 20 percent . Approximately 1 billion radiologic imaging examinations are performed worldwide annually and most of the resulting images are interpreted by radiologists. If these interpretations carried an average error rate of only 4% (the lowest estimate for the rate of radiologic error) this would be approximately 40 million radiologist errors per year .
In a recent study of second readings performed by experienced abdominal imaging radiologists, they disagreed with each other more than 30% of the time and disagreed with themselves more than 25% of the time . Some studies report that reducing the viewing time to less than 4 seconds increases significantly the miss rate .
Four categories of radiologic errors
MRI, CT and ultrasound radiologic errors can be classified in four broad categories, with perceptual errors being the more common of them with a 60 – 80% miss rate[1, 5, 6]:
- Technique or image acquisition errors: missed because of the technical or physical limitations of the imaging modality. Contributors to these types of errors include staff shortages, inexperience of the staff and inadequate equipment.
- Artifacts: imaging artifacts or technical factors might obscure the pathology due to image distortion (false negative) or might mimic pathology (false positive).
- Inappropriate study: when the indication of an image test is not correct, and the capability of a particular radiologic technique might not be suitable for the question that is being asked.
- Incomplete study: when the indication of an image test is correct, but does not include all the areas that must be studied.
- Perceptual errors: when an abnormality is not identified.
- Under-reading: when the finding is not detected. It is the most common type of error.
- History: when a finding is missed because of an inaccurate, incomplete, or misleading clinical history.
- Satisfaction of search: when a finding is missed because of failure to complete a systematic search after discovering an abnormality. It is the second most common type of error.
- Satisfaction of report: when a finding is missed because of over reliance on the radiology report from a previous examination.
- Location: when a finding is missed because it is outside of the area of interest.
- Image manipulation: when a radiologist fails to perceive an abnormality because of an inappropriate window (CT) or pulse sequence (MR).
- Cognitive / interpretative errors: when an abnormality or a normal variant is seen but its meaning or importance is not correctly understood, resulting in an incorrect diagnosis.
- Complacency / Normal variant: when a clinically unimportant finding was appreciated but attributed to the wrong cause (false positive).
- Faulty reasoning: when the finding was appreciated and interpreted as abnormal, but attributed to the wrong cause (a true positive misclassified).
- Lack of knowledge: when the finding is correctly identified on the image, but its diagnostic importance is missed because of the reader’s lack of knowledge.
- Complication: when the finding missed is a complication from a procedure.
- Prior examination: when a finding is missed because of failure to consult prior studies or reports.
- Communication errors: when the radiologist fails to effectively communicate the results including errors in the report, in making recommendations and in communicating important findings.
Most common reasons for diagnostic errors
A review of 182 cases made by Blake A. Johnson in 2016 revealed that the most common reasons for diagnostic errors were: failure to consult prior studies or reports; limitations in imaging technique (inappropriate or incomplete protocols); inaccurate or incomplete history; location of the lesion outside of the region of interest; failure to search systematically beyond the first abnormality discovered; and failure to recognize a normal variant .
As we all know, radiologic interpretation is not a binary process. It is a very complex process that takes far more time than the process of acquiring the images. It is also subject to a wide variety of factors such as work environment, volume of imaging studies, pressure for fast results, limited patient information, limited access to clinicians, and distractions inherent to our workplace. However, as complex as it may be, radiologic interpretation constitutes an important component of the information available to clinicians to formulate the diagnosis, staging, treatment, and follow up of their patients.
15 strategies to help minimize radiological errors
Therefore, to try to minimize the rate of radiological error, we propose the following general strategies:
- Adopt standardized and optimized imaging protocols.
- Recognize and correct imaging artifacts by having a basic knowledge of CT and MR physics.
- Use appropriate and variable grayscale settings when evaluating CT studies so images have an appropriate contrast resolution to optimize the visualization of pathologic findings.
- Optimize working environment with respect to ergonomics. Avoid interruptions, distractions, and glare affecting the monitor. Human factors such as excessive workload and fatigue should be avoided to maintain constant vigilance in interpretation of studies.
- Make a systematic search or a structured image evaluation. A well-designed checklist might remind radiologists to take a second look and to pay special attention to the first and last images in viewing series in cine mode to avoid “scrolling error”.
- Review prior reports or imaging studies after the initial interpretation. Information from previous studies might improve diagnostic accuracy and might help to avoid significant errors. “Alliterative” or “satisfaction of report” errors occur more often when radiologists read the reports of previous examinations before looking at the new exam, so radiologists must perform their own interpretation before reading previous reports.
- Consult the literature when formulating conclusions about an unknown case to formulate a sufficiently broad range of differential diagnoses and create a program of continuous learning to prevent knowledge gaps.
- Attempt to overcome cognitive biases. Be familiar with causes of diagnostic pitfalls such as anatomic blind spots or normal anatomic variants to reduce a common cause of misinterpretation.
- Consult liberally with colleagues and/or referring physicians when a challenging case is encountered. Information obtained from these conversations often influences the final report and helps to avoid interpretative errors.
- Create multidisciplinary committees, particularly for cancer care. An important element here is the double reading of images that is made within the appropriate clinical scenario.
- Adopt communication protocols. Many errors are due to a poor communication at some stage in the imaging / reporting process. The adoption of a protocol for communication of urgent or unexpected radiological findings may reduce this type of error.
- Train radiologists to improve the clarity and effectiveness of their written reports so their message is not lost or misunderstood.
- Try to make double readings to safeguard the quality of the service. However, this strategy implies a significant manpower issue for its routine adoption. Another option is to establish a peer-review program in which a percentage of studies are evaluated by colleagues to provide feedback about any diagnostic errors.
- Pay attention to individual physician factors such as illness or advancing age to ensure they do not significantly affect diagnostic performance.
- Use computer-assisted detection if it is possible. Several such systems are available, and although their clinical effectiveness remains controversial, this area holds much promise.
In summary, we believe that increased awareness of the facts commented above will direct the actions to help detect and reduce the rate of error. #radiology
Read the authors’ blog on “Strategies to Avoid Diagnostic Pitfalls in Head and Neck Imaging.“
Paola A. Barón Ródiz, M.D. is a Radiology Resident at Hospital Universitario Severo Ochoa in Leganés, Madrid – Spain
Concepción Ferreiro Argüelles, M.D. is a Neuroradiologist at Hospital Universitario Severo Ochoa in Leganés, Madrid – Spain.
- Michael A. Bruno, Eric A. Walker, Hani H. Abujudeh (2015) Understanding and confronting our mistakes: the epidemiology of error in radiology and strategies for error reduction. RadioGraphhics 35: 1668 – 1676.
- Goddard P, Leslie A, Jones A, Wakeley C, Kabala J (2001) Error in radiology. Br J Radiol 74(886): 949 – 951.
- Abujudeh HH, Boland GW, Kaewlai R et al. (2010) Abdominal and pelvic computed tomography (CT) interpretation: discrepancy rates among experienced radiologists. Eur Radiol 20(8): 1952 – 1957.
- Robinson PJ (1997) Radiology’s Achilles’ heel: error and variation in the interpretation of the Röntgen image. Br J Radiol 70(839): 1085 – 1098.
- Blake A. Johnson (2016) Avoiding diagnostic pitfalls in neuroimaging. Appl Radiol 45(3): 24 – 29.
- Adrian Brady, Risteárd Ó Laoide, Peter McCarthy, Ronan McDermott (2012) Discrepancy and error in radiology: concepts, causes and consequences. Ulster Med J 81(1): 3 – 9.