Recently, a good friend of mine rolled her ankle playing recreational volleyball. I was on the adjacent court and could see someone crawling her way off to the sidelines. Over the next few minutes, a few tears were shed, the ankle was elevated on a bench, an ice pack was applied and then the question arose – “should I get an xray?”
The Scenario: Rolled ankle. Instant pain. Unable to weight-bear. It feels pretty bad. Within 15 minutes it’s looking pretty swollen. Should I get an xray?
How often does an xray performed on a sprained ankle reveal a fracture? The answer is – not very often. Several of the most referenced studies put the number of sprained ankles seen in an emergency room that result in a fracture as less than 15%. I'm sure it's probably lower than what most of us would have thought.
Let’s look at 3 of these studies a little closer.
#1: A physician named SC Brooks and his colleagues from Edinburgh, Scotland published an article in 1981 that is commonly cited as one of the earliest studies investigating the likelihood of experiencing a fracture post-ankle sprain.1 The idea was pretty straightforward: for a period of 10 weeks at the Regional Accident Unit, Royal Infirmary, in Edinburgh, everyone between the ages of 12 and 65 that came in with a sprained ankle would be xrayed, and then statistically the number of fractures could be expressed as a percentage. In those 10 weeks, 241 people with a rolled ankle came in, 30 of the ankles (12.5%) showed up with a fracture.
#2: Dr. T. Vargish and his team from the University of Iowa School of Medicine followed Brooks’ study with one of their own.2 In their study, they initially retrospectively examined 2 years of data to investigate the proportion of ankle sprains that produced a fracture – out of 600 ankle sprains, 73 ankles, or 12%, ended up being a fracture. They followed this analysis by performing a study similar to that of Brooks - xray-ing all traumatic ankle sprains over a 6 month period to investigate how many of these actually revealed a fracture. The findings: 150 ankle sprains, with only 19 ankles (12.7%) presenting with a fracture.
#3: Alan Montague and Robert McQuillan, 1985, from the Accident and Emergency Department, Derbyshire Royal Infirmary in Derby, England.3 During a four month period, a total of 311 people arrived to the emergency department secondary to a sprained ankle. Every single ankle xray was performed within at least 24 hours of the person spraining their ankle. Their findings? 55 people, or only 17.7%, revealed a fracture.
So it seems that the less than 15% estimate is fairly reasonable. Although in the grand scheme of things the number is fairly low, the fact is, fractures because of an ankle sprain can and do happen, and we definitely wouldn’t want to miss one. So, is there any way of knowing when an xray is appropriate?
Each of these previous authors tried to answer that question by coming up with certain clinical features that could correctly estimate the presence of a fracture. Moderate to severe pain? Brooks found that this was present in 70% of the fractures… but in 66% of sprains. Any bruising present? Brooks found that this was present in 80% of fractures but in only 41% of sprains. OK, maybe bruising vs pain would be better at identifying the presence of a fracture. Some tenderness on the outside ankle and some ability to weight-bear? Vargish found that this identified 97.5% of sprains – meaning that answering yes to these 2 questions meant less than 2.5% chance of an xray revealing a fracture. Even better.
But is there a way of 100% of the time not missing a fracture?
Enter Ian Stiell, an emergency medicine physician out of Ottawa, Ontario, Canada who asked this very question.4 In the early 90s Ian and his colleagues set out to establish some “rules” that could help predict a 100% chance of never missing an ankle fracture.
His study, performed in 1992, seemed simple enough: a team of emergency medicine experts identified a list of 32 variables that could make one suspicious of an ankle fracture. These included the presence of bruising, tenderness, the ability to walk, hearing a “cracking” sound, etc. These variables made up a standardized exam. This standardized exam was then applied to a whole lot of sprained ankles – 689 to be exact – and then an xray would be taken. Of these, 70 ankles (representing 10.2% of the study population) revealed a fracture.
The question Ian wanted to answer – were there any variables within the standardized exam that could predict with 100% certainty the absence of a fracture? The answer was – yes. There were 3 – which can all be applied in the form of a question:
Inability to weight-bear 4 steps in the emergency department?
Bone tenderness on the lower or back edge of the medial malleolus?
Bone tenderness on the lower or back edge of the lateral malleolus?
When all 3 of these questions were answered with a “no”, not a single one of the 70 fractures were missed. Seems pretty good.
Now, question #2: could these results be reproduced?
In a following paper published in 1993, Ian and his colleagues applied the same set of rules to a new group of 1485 people with ankle sprains before an xray was taken, and then the xray would be read and analyzed for the presence of a fracture.5 The results? Again, when all 3 questions were answered with a no, not a single fracture was missed.
2 studies. 2174 ankle sprains. 3 rules. Not 1 fracture missed.
Some pretty good rules. Now, there were some nuances to the data. For example, even when the rules were positive – the likelihood of identifying a fracture was only 49% (meaning answer a “yes” to any of the rules, there was approximately a 50% chance of a fracture being present). And no one under the age of 18 was included in the study meaning the same rules may not apply to that population (although data for children do exist). But with regards to adults over the age of 18, excellent data nonetheless.
So let’s get back to the scenario. Should she have gotten an xray? Well, unfortunately we would need more information in her case before we could really answer. But if the ankle bones aren’t tender to the touch, and she can take 4 steps, how confident can we be that a fracture is unlikely? 100%.
References:
Brooks SC, Potter BT, Rainey JB. “Inversion injuries of the ankle: clinical assessment and radiographic review.” British Medical Journal 1981;282:607-608.
Vargish T, Clarke WR, Young RA, Jensen A. “The ankle injury – indications for the selective use of x-rays.” Injury 1983;14:507-512.
Montague AP, McQuillan RF. “Clinical assessment of apparently sprained ankle and detection of fracture.” Injury 1985;16:545-546.
Stiell IG, Greenberg GH, McKnight RD, Nair RC, McDowell I, Worthington JR. “A study to develop clinical decision rules for the use of radiography in acute ankle injuries.” Annals of Emergency Medicine 1992;21(4):384-390.
Stiell IG, Greenberg GH, McKnight RD, Nair RC, McDowell I, Reardon M et al. “Decision rules for the use of radiography in acute ankle injuries. Refinement and prospective validation.” Journal of the American Medical Association 1993;269:1127-1132.