Only 25% of ankle fractures are displaced, although they occupy most of the literature and textbooks.
Residual mortise incongruity increases the risk of late osteoarthritis (Berkes 2013), although many patients with some incongruity or even OA have few symptoms. Therefore, the goal of treatment is to restore tibiotalar congruity, maintain this until fracture union and provide opportunities for rehabilitation. Even with optimum treatment, most patients will have some residual symptoms at long-term follow-up and they need to be advised of this before treatment begins.
Initial management includes:
- management of other injuries
- management of open wounds with cleansing, dressing, documentation and prophylaxis against pyogenic infection and tetanus
- reduction of serious incongruity, especially if the medial skin is at risk
- splintage, usually in a plaster backslab
- pain relief
Traditionally, most displaced ankle fractures were managed by closed reduction and casting, with prolonged nonweightbearing and follow-up radiography. Early internal fixation techniques emphasised stabilisation of the medial malleolus with screws, believing that the lateral malleolus contributed little to ankle stability. In 1977, Yablon published a group of biomechanical and clinical observations which led him to the conclusion that “the talus follows the lateral malleolus” and that the key manoeuvre in internal fixation of the ankle was accurate reduction and stable fixation of the lateral malleolus. Yablon’s work, along with stable fracture fixation techniques through the AO school, has led to open reduction and internal fixation of most ankle fractures as a standard technique. What is the evidence to support this change?
MUA or ORIF - the evidence
Five RCTs have compared surgical fixation with closed reduction and casting (Bauer 1985, Phillips 1985, Rowley 1986, Makwana 2001, Sanders 2012). Makwana’s trial included only patients over the age of 55, and was the only trial to show any functional advantage at long-term follow-up for surgically treated patients. Bauer’s trial showed no difference at 6-8 years between patients treated surgically or by closed reduction and casting, but the surgical group recovered quicker. Rowley found that surgically treated patients took longer to recover normal movement and gait. Phillips’ paper is often quoted to show better outcomes in surgically treated patients, but in fact the clinical outcomes were the same – only the radiological outcomes were better after surgery. Sanders studied patients with undisplaced fractures who had increased medial clear space on stress radiography and were then randomised to ORIF or closed reduction and casting. There were no differences in Olerud or SF36 scores, but 20% of conservatively treated fractures had MCS 5mm or more, two of whom were operated on. Ten patients who had ORIF required secondary surgery.
These studies should not be taken to show that ORIF is unnecessary. For one thing, the post-operative management was restrictive (only Rowley et al allowed early weightbearing and none allowed early movement to surgically treated patients). Outcome measures were non-standardised and there was significant loss to follow-up in Phillips’ and Makwana’s series. In addition, there were patients in each series (10-30%) who could not be managed closed and required ORIF. Michelson (2007)’s meta-analysis showed that the risk of adverse events was 15% in the non-surgically treated groups and 10% in the surgical groups. On the other hand, redisplacement can be treated with surgery and conservatively treated fractures which remain undisplaced do as well as surgically treated fractures. It is reasonable for patients to choose after evidence-based counselling (Tornetta 2012). Further trials, using modern methods of post-operative care, may show additional advantages for surgery. Surgery would, of course, be mandated in open fractures and highly-unstable fractures, and strongly indicated in polytrauma and patients who will particularly gain from early mobilisation.