Another winters morning. You are freshly vaccinated, caffeinated and ready for another ED shift. Your first patient is a return visit. A 7 year-old who fell onto his shoulder at school a week ago. You read your colleague’s previous assessment. On examination there was no bony tenderness and the x-ray report of the right clavicle was normal. Yet today there’s a lump over the collar bone and he’s no longer using his arm normally. Has something been missed?

X-ray interpretation is a complex human enterprise vulnerable to a wide variety of errors. The extent of missed diagnoses has been estimated to be as high as 15-20% ^1,2.

There are two principle types of error:

• Perceptual errors – those where the abnormality is simply not seen
• Cognitive errors – where the abnormality is seen but its significance is not appreciated

You might think that such errors can simply be avoided through education, better imaging techniques and training. Yet since the 1960’s, despite doubtless advances in technology and improvements in medical practice, the rate of radiological errors has remained almost unchanged.

So what do we do? Admit defeat? Never!

Instead, let’s journey inwards and analyse these errors, why we make them and how we can improve ourselves and our approach to avoid missing fractures in children with injuries.

Causes of error

Perceptual errors are the most common and are due to many factors including:

• Clinician fatigue
• Distractions from colleagues and the working environment – the extrinsic cognitive load
• High workload
• Satisfaction of search (spotting one abnormality then failing to look for any more)

There is a reason your friendly radiologist is sat quietly in a dark room with a cup of coffee – a world away from a noisy, busy accident and emergency department. Consider yourself and your environment when reviewing an x-ray. Just as with prescribing, respect reviewing x-rays.

Even with the best conditions, what the eye sees the brain doesn’t always spot. Consider the now infamous Invisible Gorilla experiment that earned Christopher Chabris and Dan Simons an Ig-Nobel Prize in 2004. Participants were asked to watch a video and count the number of times the ball was passed between players. What they failed to notice was the large hairy simian playing the game. The brain failed to recognise what the eyes clearly saw.

Cognitive errors occur for a whole host of reasons. Some of these include:

• Lack of knowledge (e.g. how to interpret x-ray findings, ossification centres, etc.).
• Lack of clinical information (e.g. history or examination)
• Faulty reasoning (e.g. fracture identified but not cause of pain)
  • True positive, misclassified
• Complacency (e.g. fracture identified but from separate injury)
  • False positive finding
• Satisfaction of report (e.g. reliance on radiology report discourages further analysis).
• Satisfaction of search (e.g. finding one fracture discourages search for another).  

Then there are our own cognitive biases which may also influence our interpretation…

Availability bias– recent experience of a diagnosis/presentation makes you more likely to diagnose the same condition

“I saw a pulled elbow the other day, it looks the same”. May miss ulnar dislocation.

Confirmation bias– looking for evidence to support your hypothesis and ignoring evidence against

“It looks like a simple ankle sprain, I think that X-ray must be fine”. Can miss fractured fibula.

Outcome bias– opting for the diagnosis associated with the best patient outcome/prognosis

“If there is a vertebral fracture, we will have to immobilise this child. It probably isn’t that”.

Zebra retreat– history and findings are in keeping with a rare diagnosis but the diagnostician is afraid to confirm this

“As Dr Cox says, if you hear hoofbeats look for horses not zebras” …sometimes it’s a zebra!

Finally, no article on medical error would be complete without reference to the good old Swiss Cheese Model. We are but one step in a sequence of events that can either prevent or lead to error. For our example case, consider the following…

Can I have some examples please?

Most fractures in children are easy to spot however some may present with subtle findings, especially when they involve the epiphyseal growth plate.

Examples of where most missed fractures occur are shown below:

Many fracture patterns are unique to children. The paediatric skeleton is more elastic, more porous, and has a relatively stronger periosteum. That makes it uniquely vulnerable to torus fractures, buckle fractures, plastic bowing and greenstick fractures. Knowing to look for such subtleties sets paediatric fracture diagnosis apart. That coupled with odd growth plates and ossification centres explains, in part, why fractures are more easily missed in children⁵.

We have seen how even with the benefit of the patient in front of us and the luxury of radiology reports that we are vulnerable to making mistakes. Yes, we need to first know our ischial spine from our olecranon (our arse from our elbow), but we also need to train ourselves in techniques to avoid perceptual and cognitive traps.

So how do we prevent them?

Reducing missed fractures in children

Sadly the evidence is lacking and largely focuses on the performance of radiologists. Approaches centred solely on education and training are insufficient. Slowing down strategies, group decision-making and feedback systems are, as yet, an unproven step in the right direction. Checklists, however, have a growing evidence base in improving performance despite their poor popularity.

Whether or not you are a fan of the ‘Checklist Manifesto’, less controversial are principles around workplace culture and communication. Facing up to errors, avoiding blame and frequently just talking with colleagues (the clinician, the radiographer, the radiologist, the patient) remains incredibly important.

What else?

Systems-level thinking

A growing number of healthcare trusts now implement peer learning systems. Rather than being punitive, such groups create collective opportunities to teach using diagnostic catches as well as misses. At Leicester Royal Infirmary, Education Fellow Sarah Edwards set up one such weekly group teaching session for A&E staff. It gave them the opportunity to review images with the support of a Consultant Radiologist.

Evidence also supports “double-reading” to reduce the misses. At the Royal London Hospital, we are supported by our Radiology colleagues who review all images from our paediatric emergency department within 24 hours. Furthermore, within our ED we foster a culture of learning from each other through openly sharing learning points without risk of embarrassment and most (if not all) x-rays are reviewed by two or more clinicians to share knowledge and experience.

Such principles underpin the Irish National Radiology Quality Improvement (QI) programme. Through standard setting and measuring performance they pursue a cycle of continued quality improvement.

Individual level thinking

Michael Bruno, Vice Chair for Quality and Chief of Emergency Radiology at Penn State University says “there’s a very simple fix for errors of thinking- cognitive biases.… you must force yourself to ask really open-ended questions…. what else, how else, where else could a finding be… force your mind back open again.”

To be more technical, lets consider the “dual process theory of reasoning.” In radiology, automatic system 1 processes typically enable immediate pattern recognition. In contrast deliberate system 2 reasoning enables less obvious abnormalities to be detected. Normally there is a dynamic oscillation between these to forms of thinking. The lesson is not to eliminate type 1 processing, which is prone to mental shortcuts and mistakes, but instead to be aware of our own thinking with the ability to deliberately “turn on” our type 2 brain when needed.

This discipline is termed metacognition or meta-awareness. 

For those who find such talk nebulous, there a number of practical steps that come recommended from Andrew J. Degnan (Department of Radiology at Children’s Hospital of Philadelphia).

Maintain a healthy skepticism

Reflect on your diagnostic process, challenge your interpretation forensically and question yourself objectively.

Use a structure or checklist

Structured reports help radiologists. Find your own repeatable techniques and approach each x-ray systematically, including “review areas” that are often overlooked. 

Consider the clinical findings

What is your pre-test (pre-x-ray) probability? How confident were you in your clinical assessment? Is the x-ray a rule-in or rule-out? Marrying up a thorough history and examination with a careful focus on the relevant radiographic area often bears reward.

“Injuries that are missed because of failure to image are typically because the injury was poorly localized or because of the presence of other injuries distracted attention from the injured part.”

Mind your environment

Are you fatigued? Have you had a break? Clearing your mind for even a moment can actually improve overall efficiency. A quiet work space. A few minutes away from distraction. These will all empower your type 2 thinking.

Mitigate, mitigate, mitigate

Mistakes happen. Telling parents about uncertainty is critical to them re-presenting if their child’s soft tissue injury or sprain is not improving. Importantly, this is not the same as forgoing responsibility. Yet if your routine practice includes quality safety netting, discussing cases with your friendly radiologist and chasing up on cases you may not prevent mistakes but you might minimize the harm that comes from them.

What happened with our case?

A repeat x-ray was done but again no fracture was evident. Yet to examine there was an un-deniable lump mid-clavicle. In view of persistent pain and continued non-use of the limb (right arm) the child was discussed with the radiologist who agreed upon ultrasound. Ultrasound confirmed early callus formation and a break in the cortex that was not visible on X-ray. The child went home in a sling for outpatient follow up.

Take home messages  

• Missed fractures are more common in children and not necessarily subtle
• Know what to look for and how to look for it
• Process is important, don’t forget history and examination
• Communicate clearly, speak frequently with your radiographer and radiologist

Selected references

1. Berner ES, Graber ML. Overconfidence as a cause of diagnostic error in medicine. Am J Med 2008;121(5 suppl):S2–S23.

George MP, Bixby S. Frequently Missed Fractures in Pediatric Trauma A Pictorial Review of Plain Film Radiography Radiol Clin North Am 2019 Jul57(4)843-855. – Images 3,5 in carousel

Hernandez, J.A., Swischuk, L.E., Yngve, D.A. et al. The angled buckle fracture in pediatrics: a frequently missed fracture. Emergency Radiology 10, 71–75 (2003) – Images 1,2 in carousel

Jadhav, S.P., Swischuk, L.E. Commonly missed subtle skeletal injuries in children: a pictorial review. Emerg Radiol 15, 391–398 (2008). – Image 8 in carousel

2. Wachter RM. Why diagnostic errors don’t get any respect: and what can be done about them. Health Aff (Millwood) 2010;29(9):1605–1610.

5. Smith J, Tse S, Barrowman N, Bilbao A, (2016). P123: Missed fractures on radiographs in a paediatric emergency department, CJEM, 18 (S1), S119-S119

Swischuk, L.E., Hernandez, J.A. Frequently missed fractures in children (value of comparative views). Emerg Radiol 11, 22–28 (2004). Images 4,6,7 in carousel

Further reading

Brady AP. Error and discrepancy in radiology: inevitable or avoidable?. Insights Imaging. 2017;8(1):171-182. 

Kim YW, Mansfield LT. Fool me twice: delayed diagnoses in radiology with emphasis on perpetuated errors. AJR Am J Roentgenol 2014;202(3):465–470.

Michael A. Bruno, Eric A. Walker, and Hani H. Abujudeh, Understanding and Confronting Our Mistakes: The Epidemiology of Error in Radiology and Strategies for Error Reduction, RadioGraphics 2015 35:6, 1668-1676 

Martino F., Barbuti D., Martino G., Cirillo M. (2012) Missed Fractures in Children. In: Romano L., Pinto A. (eds) Errors in Radiology. Springer, Milano.

Miele V., Galluzzo M., Trinci M. (2012) Missed Fractures in the Emergency Department. In: Romano L., Pinto A. (eds) Errors in Radiology. Springer, Milano.

Wang CC, Linden KL, Otero HJ. Sonographic Evaluation of Fractures in Children. Journal of Diagnostic Medical Sonography. 2017;33(3):200-207.