Veterinary radiology: are you making diagnostic errors?

As a vet, it’s likely you take radiographs most days. Veterinary radiology is a great tool in your diagnostic arsenal – but how accurate is it? Could you be making diagnostic errors? In this article, we look at the results of a recent study "Veterinary radiologic error rate as determined by necropsy" (Cohen, J. et al, 2023) and explain key takeaways from it to help vets in practice improve their radiology. 

The importance of accurate diagnosis 

Cohen et al’s 2023 study was the first of its kind in veterinary medicine. Previous studies in human medicine have investigated discrepancies between autopsy and imaging findings. However, similar research in veterinary medicine is limited. This is an important study, because understanding the limitations and potential pitfalls of diagnostic imaging can help vets in practice minimise errors and provide the best possible care for their patients. 

Cohen et al’s study aimed to evaluate the accuracy of diagnostic imaging in veterinary radiology. They compared postmortem diagnoses in dogs that had corresponding radiographs, finding a significant discrepancy rate of 40%. This means that nearly half of all clinically significant abnormalities found at postmortem were not diagnosed with the radiographs taken at the time. The study highlights common errors in radiologic interpretation and offers insights into improving diagnostic accuracy. 

Methodology 

This study was conducted on a year's worth of patient records at The Schwarzman Animal Medical Center. Cases with both necropsy and antemortem imaging were included. Over 1000 patients had both necropsy and antemortem x-ray. Diagnoses from the two methods were compared and categorised as either agreeing or disagreeing with each other. The authors then determined whether the postmortem diagnosis could have been visible on a radiograph (taking into account how things change over time and the sensitivity of x-rays) and determined an error rate. 

Key results 

40% of the postmortem diagnoses disagreed with the antemortem diagnoses – a 40% discrepancy rate. However, once the authors excluded conditions that could have changed over time or wouldn’t be expected to be visible on a radiograph, the error rate was 4.6%, which is comparable to the 3-5% that similar studies have discovered in people. In other words, 1 in 20 cases have a clinically significant missed diagnosis or misdiagnosis which was visible in the antemortem radiographs.  

Factors that contribute to radiologic error 

The high discrepancy rate highlights the importance of careful interpretation of imaging studies and the need for continuous improvement in diagnostic accuracy. Several factors contribute to discrepancies noted in this study, including: 

• Limitations of imaging modalities: Certain conditions may be difficult to detect with radiography 
• Study limitations: The radiographs taken did not include the correct areas 
• Temporal indeterminacy: Changes in disease progression between imaging and necropsy (a huge 15% of the 1099 cases) 

Genuine errors (where the diagnosis was visible on the radiograph but not correctly recorded) fell into the following two categories: 

• Perceptual errors: failure to identify lesions despite them being present in the image 
• Cognitive errors: the lesion is noted, but is misinterpreted 

6 ways to reduce error rate and prevent mistakes in veterinary radiology 

So what can we learn from this study? Namely, that mistakes in radiology are common, and that many of them are down to external factors that clinicians can't always control.

To minimise radiological error, vets can: 

1. Utilise systematic approaches: Employ a consistent approach to image interpretation, including checklists and algorithms. 
2. Consider alternative diagnoses: Consider a broad differential diagnosis and continue checking the film even after an anomy has been found - there could be more than one!
3. Stay updated with the latest advancements:  Reviewing the relevant literature, attending conferences, and upskilling with diagnostic imaging CPD can help vets use technology to its full potential.
4. Collaborate with pathologists: Regular discussions with pathologists can help identify patterns of error and improve diagnostic accuracy. AI film reading can also help check for missed diagnoses.
5. Learn from mistakes: A culture of no-blame mistakes can help people own up to and learn from their mistakes. Check out our article on dealing with a veterinary mistake for more help with this.
6. Reduce systematic causes of errors: Many errors are not down to the clinician but to their environment. High workload, time pressure, staffing problems, and noisy or light interpretation rooms are all examples of how workplaces can increase error rate. Reducing these is essential.  

Conclusion 

This new study reveals radiologic diagnoses are missed/misinterpreted nearly 5% of the time, and incorrect 40% of the time. By understanding the common sources of error, vets can improve their diagnostic accuracy and ultimately enhance patient care. Taking regular veterinary radiology CPD, such as our Diagnostic Imaging Certificate Programme can help reduce error rate and improve confidence in using other imaging modalities to reduce discrepancies. Or, if you feel like a short practical course is more what you need, why not look at our veterinary imaging courses?  

Source: https://onlinelibrary.wiley.com/doi/epdf/10.1111/vru.13259