Ankle ligament injuries

Injuries to the ligaments of the ankle are often called "low ankle sprains". If the tibiofibular ligament or the syndesmosis is injured it is called a "high ankle sprain". Inversion sprains with injury to the lateral ligaments of the ankle/foot complex are by far the most common. They occur with an estimated frequency of one injury per 10 000 people per day, amounting to about 27 000 injuries each day in the United States.' 2 Although many injuries are treated outside medical establishments, 7-10% of those who are visiting the emergency departments of the hospitals in Scandinavia have sprained ankles.3 Ankle injuries are the most common injuries in sports and recreational activity.i"12 For this reason, probably, these injuries tend to occur primarily to young people.'3 The sprained ankle also remains the most common injury regardless of whether the sport is primarily an upper or a lower extremity sport. Garrick7 noted that injuries to the ankle accounted for 53% of injuries occurring during basketball and for 31% of those occurring during soccer. Reviewing 41 soccer teams, Ekstrand and Tropp'4 found ankle sprains to account for 17 to 21% ofthe injuries. The "high ankle sprain" usually occurs as the result of an eversion injury in combination with fractures or lesions to the deltoid ligament. Isolated syndesmosis injuries occur in 3% of the cases. This article will deal mainly with the lateral ligament complex.


Ankle biomechanics
The passive stability of the ankle is the responsibility of the ligaments and the bony constraints of the ankle joint, while the active stability depends on muscular support. The talus has no muscular insertion. Active motion depends on the long foot muscles inserting into other tarsal or metatarsal bones. Dorsiflexion and inversion are effected by the extensor hallucis longus and the anterior tibial muscles. Dorsiflexion and eversion are guided by the peroneus tertius muscles and extensor digitorum longus and brevis muscles. Plantar flexion and eversion are effected by the peroneus longus and brevis muscles. Plantar flexion and inversion are regulated by the flexor hallucis longus, the flexor digitorum, and the posterior tibial muscles. '5 The ligaments of the ankle can be divided into the lateral group, the medial group, and the ligaments of the syndesmosis.
The lateral ankle ligament complex is traditionally considered to consist of the anterior talofibular (ATFL), the calcaneofibular (CFL), and the posterior talofibular (PTFL) ligaments. However, in inversion the subtalar ligaments, especially the cervical ligament, the interosseous ligament, and the ligaments spanning the calcaneocuboid and the talonavicular joints, also have to be considered.
Many studies have been done on talotibial ligaments to gain insight into how they function together to stabilise the joint. Of the talotibial ligaments, the ATFL is a thin 6-10 mm wide, 20 mm long, and 2 mm thick'6 weak ligament, being essentially a thickening of the anterior ankle joint capsule. It passes from the distal anterior origin of the lateral malleolus to the talus in front of the proximal part of the lateral articular surface. In neutral position its direction is parallel to a long axis of the foot and in full plantar flexion it is more parallel with the tibia (fig 1). The CFL is a 20-25 mm long rounded ligament with a diameter of 6-8 mm. '6 It is an extra-articular ligament closely associated with the peroneal tendon sheath. It runs obliquely downwards and backwards to be attached to the lateral surface of the calcaneus. There is a great variety in its direction and in its attachment sites."' A rupture to this ligament will also cause a rupture of the tendon sheath, and occasionally also damage the peroneal tendons.
The ATFL acts as a primary restraint against plantar flexion, as well as internal rotation of the foot. '7 In strain studies Renstrom et al" found that the strain of the ATFL significantly increases with increasing plantar flexion. In the neutral position the ligament is relaxed. '9 The CFL does not have an independent role in talotibial joint stability, but acts instead as a guide for the axis of subtalar motion. '7 In dorsiflexion the ligament has increased strain.'8 In a normal standing position the ligaments remain relaxed.
The lateral talocalcaneal (LTCL) extends from the talus to the calcaneus and blends its fibres with CFL and ATFL fibres. The exact incidence of injury to this ligament is not known. Transecting the subtalar ligaments results in very limited increase in motion when measured in degrees, but as they have very limited motion in the first place the increase after rupture is about 40%.2°The incidence of rupture here is also unknown.
The PTFL connects the posterolateral tubercle of the talus to the medial aspect of the lateral malleolus. The PTFL has an average diameter of 6 mm. In plantar flexion and in the neutral position the ligament is relaxed, whereas in dorsiflexion the ligament is tensed.'5 The clinical significance of PTFL injuries is somewhat unclear, but it is not commonly damaged. The ATFL, CFL, and PTFL ligaments function as a unit for the talotibial complex, though one may resist a specific Through the full range of motion the ATFL and CFL act in synergy22"2 ( fig 1). As the foot plantar flexes, the strain in the ATFL increases while the strain in the CFL decreases.'8 Shybut et a125 measured ankle ligament loads directly by using implanted buckle transducers. The results indicated that ligament loads remain low within the functional range of motion (10 degrees of dorsiflexion to 20 degrees of plantar flexion). This supports the concept that ankle ligaments act as kinematic guides rather than primary restraints during normal activity. Stormont et al26 studied the stabilising capacity of the ligaments and articular surface in the ankle with and without physiological loading. With loading, the results indicated that the articular surface becomes an important stabiliser, accounting for 30% of stability in rotation and 100% of stability in inversion. Without loading, the results indicated that the primary and secondary ligamentous constraints vary with testing modes and ankle position.

Mechanism of injury
The extent of tissue damage that will occur with the trauma depends not only on the mechanism and magnitude of the forces that act on the ankle but also on the position of the foot and ankle during the trauma. '5 The most common mechanism causing lateral ligament injuries is a situation where the ankle goes into a combination of plantar flexion and inversion. The ATFL tears first followed by rupture of the anterolateral capsule. With further inversion, the CFL will be ruptured followed by variable injury to the PTFL and the anterior part of the deltoid ligaments. 27 With weight bearing, the articular surface can provide 30% of stability in rotation, and 100% stability in inversion.26 This ability is a function not only of the axial load but also of the close packed position.28 Ankle destabilisation thus occurs during loading and unloading, but the joint is stable once it is fully loaded.

Frequency of lesions caused by inversion sprains
Owing to the ATFL's vulnerable position in plantar flexion, it is the most commonly ruptured ligament in a lateral ankle sprain..323 29-34 In 1964 Brostrdm'6 surgically explored 105 sprained ankles and found an isolated ATFL tear present in about two thirds of the cases. The second most common injury was a combined rupture of the ATFL and CFL, occurring in about 20 to 25% of the cases. ' of the joint occurs. In a grade III injury, there is complete rupture of the ligament with severe swelling, haemorrhage, and tenderness. There is loss of ability to bear weight on the foot, limited function, and considerable abnormal motion and instability of the joint. An improved and validated classification of ankle sprains is needed. The terms mentioned above were validated by Lindenfeld50 and found to be quite subjective. Others prefer to classify lateral ligament injuries as single or double ligament tears. A single injury would imply lesions to only the ATFL, a double injury would affect both the ATFL and the CFL. The distinction in clinical terms is difficult here also.

Diagnosis
Diagnosing the extent of an acute lateral ligament injury is not considered very accurate because of pain, swelling, and muscle tenderness.'5 In the first few days after injury local palpational pain is often diffuse, with no maximal point of tenderness.' 51 52 The extent of swelling does not depend only on the magnitude of the ligamental injury but also on the initial treatment. Extensive swelling is a predictor of ligament rupture, but the positive predictive value has been found to be only 60-70%. 5 The characteristic haematoma suggesting ligament rupture usually does not develop during the first few days, and joint range of motion is mainly determined by the severity of pain and does not differ between simple sprains and ligament rupture.5455 Furthermore, the interobserver variation of the acute examination concerning tenderness, swelling, discoloration, and the anterior drawer sign is considerable.56 The anterior drawer sign seems to be a good predictor of lateral ligament disruption. Brostrom30 found that the anterior drawer sign was positive in all of 239 patients with ligament rupture. His examination was, however, done with general or local anaesthesia. Without anaesthesia he could only elicit a positive drawer sign in two patients. Based on these results van Dijk57 suggests that the physical examination should be delayed four to five days after the initial injury. The specificity and sensitivity of delayed physical examination for the presence or absence of a rupture to the ATFL were found to be 84% and 96% respectively, and the delayed examination, done by observers of varying degrees of experience, gave information of ligament quality that equalled that of arthrography. After four to five days a combination of tenderness at the level of the anterior talofibular ligament, lateral haematoma discoloration, and a positive drawer sign indicated a ligament lesion in 95% of the cases. A negative drawer test and the absence of discoloration always indicated an intact ligament.
Radiographs to exclude fractures are suggested according to the Ottowa strategy for ankle injury.58 Plain radiographs are taken if there is bone tenderness at the tip or posterior aspect of the lateral malleolus, at the tip or posterior aspect of the medial malleolus, at the navicular tuberosity or at the base of the fifth metatarsal, or if the patient is unable to bear weight immediately after injury and at the initial examination. With these rules the authors could reduce the use of radiographs from 80% to 63% of the injuries.

Treatment of lateral ligament lesions TREATMENT OF GRADE I AND II SPRAINS
In the presence of a grade I or grade II injury it is universally agreed that recovery is fast with non-operative management and the prognosis is good.455'5 Jackson et at5 found that early mobilisation resulted in a disability of eight days for a grade I and 15 days for a grade II injury. Functional treatment including early motion and use of ankle support and early weight bearing is today the accepted treatment for grade I and II ankle sprains.

TREATMENT OF GRADE III SPRAINS
In the case of grade III lateral ligament lesions the treatment of choice that is, whether to operate, to immobilise in a cast, or to allow early controlled mobilisation, is more controversial. The key consideration being the question of whether ligament healing with adequate tension can be achieved equally well with early controlled mobilisation as with direct visualisation and suturing.
If the ankle is kept from excess inversion or dorsal or plantar flexion the strain in the lateral ligaments remains low,'8 allowing for adequate healing conditions without elongation within this range of guided motion. With the development of lower extremity magnetic resonance imaging scanning, studies have visualised that total ligament (grade III) ruptures do heal with ligamental continuity if there is early mobilisation. There is thus experimental evidence supporting ankle ligament healing in the presence of early mobilising treatment.
Clinically many uncontrolled, nonrandomised studies have shown mechanical stability and satisfactory subjective results after both operative and conservative treatment. Of the many papers available, only 12 could be considered prospective and randomised30 55 60-69 when a comparison between operative and conservative treatments was performed by Kannus and Renstr6m.35 Six of these considered the three treatment groups in question: operation, immobilising and early mobilising modalities30 62 63 65 66 69 and the results and conclusions based on these reports will be mentioned in the following. Duration of follow up ranged from six months to 3.8 years, which is considered adequate time to identify persistent disability. Results were evaluated using selected outcome parameters. Return to work or physical activity was reported in four of the studies that included three treatment modalities.30 6166 69 They concluded that return to work was two to four times faster after functional treatment than after operation or immobilisation in a cast. Return to pre-injury level of activity was found to be faster after conservative treatment than after operative treatment in four cases.6 62 66 69 The opposite was found in three studies.0 5563 and no difference was found in the remaining five.
Pain, swelling, or stiffness with activity could be evaluated in four of the studies involving three treatment groups.62636669 The studies failed to show any differences. In the three studies that included three therapy groups and included mobility of the ankle,306669 mobility was found to be superior after functional treatment compared with the other two methods.
Better mechanical stability has been the primary argument for operation. Assessment of mechanical stability calls for an objective measurement with the evaluation of talar tilt and anterior drawer on stress radiographs. Based on five studies evaluating the three different treatments,62 65  Functional instability is not a well defined entity but designates repeated inversion injuries that are either unprovoked or the result of very little provocation. Functional instability seems to be a late disability in 15 to 60% of lateral ligament injuries depending on the definition applied. In the studies that reported on functional instability'062666971 opinions diverged, but no one treatment seemed superior to the others in minimising the chances of late instability. The scientific basis for treating chronic functional instability is discussed later in this review. concluded that in first time lateral ankle sprains, although both immobilisation and early mobilisation prevent late residual problems and ankle instability, early mobilisation allows early return to work and, possibly, is more comfortable for patients. In a recent prospective study comparing surgery with functional treatment in ankle ligament tears Kaikkonen and coworkers74 found that early mobilisation gave better results than surgery plus mobilisation in treatment of complete tears of lateral ligaments of the ankle. Nine months after injury excellent to good scores were achieved in 87% of the functionally treated patients and 60% of the surgically treated patients respectively. In a metaanalytical article Schrier75 found functional treatment was superior to immobilisation in casts, reducing both ankle instability and the cost of treatment.
Sommer and Schreiber76 considered the cost-benefit aspects by comparing immobilisation in a plaster cast followed by immobilisation with a brace, with early mobilisation using a brace only. Early functional therapy proved to be the treatment with the least direct costs.
Leanderson and Wredmark77 in a study of 73 patients with grade II and III sprains treated with either early mobilisation or immobilisation found that the socioeconomic savings were potentially significant with ankle brace treatment. Nationally, the potential yearly savings in Sweden using this treatment were estimated to be eight million American dollars. The treatment of choice thus remains functional treatment.
If acute surgery is considered necessary the indications could according to Leach and Schepsis78 be (a) a history of momentary talocrural dislocation with complete ligamentous disruption, (b) a major clinical anterior drawer sign, (c) 10 degrees more tilt on the affected side with stress inversion testing, (d) clinical or radiographic suspicion of tears in both the ATFL and CFL, and (e) osteochondral fracture. Most techniques described for repair of acute ligament injuries are similar to that of Brostr6m.'0 The results after acute surgery are in general very good with return to sport in 10 to 12 weeks. This must still be compared with the three to six weeks after functional treatment.
Acute treatment protocol BIOLOGICAL BACKGROUND In the post-injury phases of an acute severe ankle injury, an ideal treatment and rehabilitation programme should fulfil four requirements'5: 1 The RICE principle: rest, ice (cold), compression, and elevation aims to minimise haemorrhage, swelling, inflammation, cellular metabolism, and pain in order to offer the best possible conditions for the healing process.s780 2 Protection of the injured ligaments during the first one to three weeks is required. In this phase of healing (the proliferation phase), protective ankle support is followed by undisturbed fibroblast invasion of the injured area, which leads to undisturbed proliferation and production of collagen fibres. Mobilisation too early in inversion leads to more prolonged type III collagen formation with weaker healing tissue than during optimal immobilisation.8"8' Protection is also needed to prevent secondary injuries and early distension and lengthening of the injured ligaments. 3 About three weeks after the injury the maturation phase of the collagen and the formation of final scar tissue begins."8'-In this phase the injured ligaments need controlled mobilisation, and perhaps even more importantly, the ankle itself must avoid the deleterious effects of immobilisation on joint cartilage, bone, muscles, tendons, and ligaments.8 8390 Controlled stretching of muscles and movement of the joint enhance the orientation of collagen fibres parallel with the stress lines (that is, the normal collagen fibres of the ligaments), and they can prevent the atrophy caused by immobilisation.8°82 Repeated exercises will also increase the mechanical and structural properties of the ligaments.9' ready for a gradually increasing mobilisation and rehabilitation programme, keeping in mind that final maturation and remodelling of the injured ligaments takes a long time-from six to 12 months.

Residual disability after inversion injury
Inadequate rehabilitation is the primary cause of residual disability after ankle sprains. Many athletes return to sports before they are fully rehabilitated and therefore they are often subjected to reinjury or additional injury. Examination may show loss of range of motion, especially limited dorsiflexion, and atrophy of lower leg muscles. Residual disability after ankle inversion injury can be divided into primarily instability problems and pain-giving entities.

CHRONIC ANKLE INSTABILITY
Chronic ankle instability can be subdivided into mechanical, functional, and subtalar instability and the sinus tarsi syndrome.
Mechanical instability Mechanical instability is characterised by ankle mobility beyond the physiological range of motion, which is identified on the basis of a positive anterior drawer or talar tilt test, or both.92 The radiographic criteria for mechanical instability vary. Most authors agree, however, that mechanical instability is present when there is more than 10 mm of anterior translation on one side or the side-to-side difference is over 3 mm or the talar tilt is more than 9 degrees on one side, or the side-to-side difference is more than 3 degrees.93 Pure mechanical instability of the ankle is seldom the sole independent reason for the development of late symptoms. Mann et al, showed that 81% of patients with radiographically documented instability experienced recurrent sprains.94 Functional instability Functional instability is the chronic disability after an ankle inversion injury that can be attributed to lateral ankle ligament deficiency and is characterised by "giving way" problems. Forty eight per cent of patients with an acute first time sprain have recurrent sprains and 26% report frequent sprains.94 The definition of the symptom is ambiguous and the pathogenesis is unclear. It does not seem to be dependent on the grade of the initial injury69 nor does it correlate with the degree of initial mechanical instability.6' 64 69-95 No histological ligamentous changes except scarring have been found in chronic ankle disability.30 As anatomical studies showed the presence of mechanoreceptors in joint ligaments and capsule, and clinical studies showed decreased postural control in patients with functional instability, Freeman et al suggested that functional instability was due to motor incoordination secondary to a proprioceptive disorder.96 Postural control was later measured objectively using stabilometry by Tropp95 who, like Freeman, found increased postural sway in subjects with functionally unstable ankles. It is, how-ever, clear that functional instability is a complex syndrome where mechanical, proprioceptive, and muscular disabilities either alone or in combination are at fault. Known factors include mechanically insufficient ligaments, peroneal muscle weakness,95 subtalar instability,97 and proprioceptive deficit.996 Although increased postural swaying has been used as a measure for a proprioceptive deficit, there is no direct causal connection between swaying with an increased amplitude when balancing on one foot and sustaining recurrent inversion injuries. Rather it seems reasonable to assume that a common and superior deficit is responsible for both disabilities. The nature of such a disability remains unclear. The latency of the peroneal muscles has been found to be significantly slower in functionally unstable ankles by some98.'00 but not by others.10'-'03 An increased error in detecting ankle inversion movement or in matching ankle inversion angles has again been found by some'04-'06 but not by others.'07 There is, however, evidence that (a) a proprioceptive deficit may result after an acute ankle inversion injury,46 70 108 (b) proprioceptive defects measured in different ways are seen in subjects with functional instability, and (c) models exist that may explain how measurable kinaesthetic deficits may cause an increased frequency of ankle inversion injuries (Konradsen et al, unpublished data).

Subtalar instability
The incidence of subtalar sprains is unknown, but it is widely accepted that most subtalar ligamentous injuries occur in combination with injuries of the lateral ligaments of the ankle ( fig  2). Subtalar instability is estimated to be present in about 10% of patients with lateral ankle instability. The symptoms of chronic subtalar instability include giving way episodes during activity and a feeling of instability when walking on uneven ground. These symptoms may coincide with chronic talocrural instability and therefore careful clinical examination is necessary. Localised tenderness on palpation over the subtalar joint is suggestive of a subtalar sprain. The diagnosis can be verified with subtalar arthrography38 or a subtalar stress view or stress tomography.
Functional treatment similar to the regimen used for lateral sprains is the treatment of choice. Surgery with non-anatomical procedures has been described.'09110 Anatomical repairs, however, show equally good results.97 "'I Sinus tarsi syndrome The sinus tarsi syndrome is often accompanied by a feeling of instability and giving way of the ankle. Seventy per cent of the patients will have experienced a severe inversion sprain.92 If the calcaneofibular ligament is torn the interosseous talocalcaneal ligament occupying the sinus will often also be sprained. This sprain will cause synovitis in the sinus tarsi with rather localised pain. The diagnosis can be made based on a complaint of pain and tenderness of the sinus tarsi combined with the giving way feeling. The pain and the giving way feeling can usually be relieved by local injections of anaesthetic and corticosteroids. Excision of tissue filling the sinus tarsi can give good results if conservative treatment fails.

Rehabilitation
With the uncertain pathogenesis of functional ankle instability the elements of treatment protocols will rest primarily on personal experiences and not so much on scientific data. Before deciding upon surgical treatment, a well planned rehabilitation programme based on peroneal muscle strengthening and coordination training should be carried out. Regaining muscle strength is primarily achieved by increasing neural activation through functionally oriented training and dynamometer training. '12 In functional training the patient is required to activate his/her musculature in normally occurring activities. Dynamic stability, including proprioceptive information, motor control, and appropriate muscle force development, will be enforced by the training.
Treatment of the proprioceptive ankle deficit consists of a progressive series of coordination exercises to re-educate the ankle. Tropp95 found a significant improvement in balance during single limb stance in subjects with a functionally unstable ankle after six weeks of balance board exercises (fig 3). The same group reported a decrease in the frequency of inversion injuries and ankle giving way feelings when compared with an untrained control group with unstable ankles. Tropp9" also found that the subjects in question showed a change in their strategy of single limb stance, going from a broken chain strategy before training to the normal inverted pendulum strategy after completed training. So it may well be, as suggested by Glencross and Thornton,'04 that rehabilitation was as much a relearning programme as it was a physical recovery after injury. Often the injured subjects themselves have a clear feeling of a perception/ proprioception deficit in the ankle joint areathey cannot "feel" the ankle as they used to. When this feeling returns it seems to coincide with complete rehabilitation. Leanderson et al'08 found that return of single limb balance to normal values coincided with a subjective feeling of full rehabilitation in dancers. It can, however, be argued that in this group balance during single limb stance was as much a specific functional test as an unspecific test for ankle proprioception. In summary, training proprioceptive function through coordination exercises using unilateral balance boards, uniaxial and multiaxial teeter boards, and jumping ropes relies primarily on empirical data but produces very favourable results. Restitution of normal function relies on the individual's subjective feeling of return to normal sensation. Full effect of training is achieved within 10-12 weeks as evaluated by postural sway. 95 About 50% of patients with functional instability will regain satisfactory functional stability after such a programme."' EXTERNAL SUPPORT When considering the mechanical effect of the different external support modalities it is important to remember that even though they do increase the resistance to inversion, the effect cannot be compared with that of active evertor muscles. Ashton-Miller et al'.' found that active evertor muscles, acting isometrically on a 15 degrees inverted ankle, can provide more than three times the protection against further ankle inversion than tape or an orthosis worn inside a three-quarter-top shoe.

Tape
The primary purpose of taping is to provide increased stiffness to the ankle and a semirigid and sometimes rigid splinting around it. Studies have shown that taping the ankle prevents ankle injuries. Garrick et al found that the combination of taping prophylactically and the use of high top shoes in basketball reduced the frequency of ankle sprains.' 4a Lindenberger et all5 found that tape had a significant prophylactic effect in handball team players who were studied prospectively for two full seasons. Six top level teams participated in the first year. In three teams the players wore tape and in the other three teams the players did not. During the first season 13 ankle sprains occurred all in the non-taped group. In the second season nine teams participated. Twenty one ankle sprains occurred and 20 of these were in the non-taped group.
Contrary to this, several studies have shown that tape loses as much as 50% of its supportive effect after only 10 minutes of active exercise"'-"8 and offers virtually no support after one hour."9 The mechanism behind the effect of ankle taping thus remains unclear. However, tape does not only restrict extensive joint motion but also enhances proprioceptive feedback mechanism and shortens the recruitment time of the dynamic ankle stabilisers. '20 Many athletes may have a skin reaction to tape and therefore skin protection may be used. There seems to be no difference in support between tape directly on the skin and tape with skin protections. Because of these skin problems, tape is used mostly by top athletes and not by recreational athletes.

Braces
The use of ankle orthoses has increased over the past decade. In contrast with tape the mechanical effect of a brace does not wear off with activity as shown by Shapiro et al. '2' Not many clinical trials have been performed judging the effect of ankle bracing on the frequency of ankle inversion injuries. In a retrospective trial Rovere et al'22 found a reduction of sprains in football using a lace-on cloth brace. Tropp et al5 in a prospective study showed the effect of a semirigid ankle brace in reducing the frequency of injuries in soccer players, and in a prospective randomised study Sitler et al'23 found that ankle stabilisers significantly reduced the frequency of ankle injuries but not their severity in a group of basketball players at West Point, New York. The resistance to an eversion moment seems to be the same for ankle braces as that for newly applied tape."4 Like tape, braces have been shown to have a proprioceptive enhancing effect. '24 There have been discussions about the effect on performance of tape and braces. In sprinting a reduction in maximal speed has been found by most,'25 but not by all. '26 For vertical jumping most studies also show that a brace decreases ability...2.'2 Opinions about bracing compared with taping vary, with some studies showing a smaller effect and others a larger detrimental effect of tape compared with braces. It must be noted that most of the studies were done on subjects with stable ankles. The use of an ankle brace or taping, or both, is recommended as they both give support against ankle inversion when the evertor muscles are relaxed and increase the maximal resistance to inversion with evertor muscles activated. They do seem to be able to enhance proprioceptive awareness, and to decrease the frequency of inversion injuries and decrease the severity of the injuries in groups with functionally unstable ankles.
The role of shoes Biomechanical studies support the use of high top shoes for ankle sprain prevention because they limit extreme range of motion, provide additional proprioceptive input, and decrease external stress on joints.'29 Clinical trials, however, have not shown that the effect of this biomechanically improved stability translates into a lower rate of ankle sprains in high top shoes. Further studies are needed.
Supportive treatment modalities Passive physical therapy modalities are often recommended to promote healing in the early rehabilitation phase. The most commonly used are ultrasound, temperature contrast baths, short waves, and various current treatments including dynamic or interference current therapy, or electrogalvanic stimulations. Of these different types of passive physical therapy, only cryotherapy has been proved to be effective. 130 Non-steroidal anti-inflammatory drugs have been studied in prospective randomised double blind trials, and found to be more effective than placebo for ankle tenderness, swelling, and pain, though the difference was not striking and seemed to disappear during an extended follow up. 3 130 In subjects with chronic post-injury swelling, moist heat packs, warm whirlpool baths, contrast baths, electrogalvanic stimulation, and intermittent pneumatic compression have been tried.4 Prospective randomised studies have shown that only the last alternative has an independent ability to reduce the amount of swelling. 131 132   Surgery for chronic instability Chronic ankle instability, as demonstrated by pain, recurrent giving way, and positive stress tests, is often treated surgically. A combination of mechanical and functional instability is the most commonly reported indication for surgery. More than 50 procedures or modifications have been described for treating chronic ankle instability, and these can be loosely grouped as non-anatomical reconstructions or anatomical repairs (table 1).
Non-anatomical reconstructions use other structures or materials to substitute for the injured ligaments. Structures commonly used for grafting are fascia late and the peroneus brevis tendons in procedures such as Watson-Jones, Evans, and Chrisman-Snook. Numerous modifications of these classic procedures have been described. The Chrisman-Snook'09 modification of the Elmslie procedure is the most widely used non-anatomical reconstruction. In four clinical series with a total of 100 ankles, more than 90% of patients had good or excellent results, and stability was more than 95% of the normal ankle. Reported complications included neuromas in 0 to 16% of patients, restricted dorsiflexion in about 20%, and restricted inversion to some degree in all patients. An advantage of the Chrisman-Snook procedure is that less lateral weakness is produced because only half of the peroneus brevis tendon is used for the graft. A criticism of the procedure has been that it results in restricted subtalar motion. Furthermore, these tenodesis procedures are nonanatomical and do not restore normal biomechanics."33 Although short term results are excellent, long term results may not be as good. Thus Karlsson et a193 found that the Evans static tenodesis was satisfactory in fewer than 50% of cases after a mean follow up period of 14 years.
Anatomical reconstructions, in which the tissues of the damaged ligaments are used, have become increasingly popular as they permit reconstruction without sacrificing any normal tissue.3093 14 Brostrdm in 196630 reported that direct suture repair of chronic ankle ligament injuries was possible even many years after the initial injuries and that the ends of the ligament could be found. Others reported that the elongated ligaments had healed encased in a fibrous scar tissue. Several authors have reported successful imbrication or shortening and replantation of the ligaments to obtain a more anatomical reconstruction. We have obtained good results with a modified Brostrom technique (Peterson procedure) that includes shortening of the ligament, repair through bony tunnels, and imbrication with local tissue93 14 (fig 4). This anatomical technique repairs both the ATFL and CFL. The non-anatomical reconstructions, except for the Chrisman-Snook modification of the Elmslie procedure, repair only the ATFL. Repair of the CFL is important, because insufficiency of this ligament may be a factor in the development of subtalar instability. The postoperative treatment after an anatomical reconstruction is a splint for eight days to secure wound healing and, thereafter, a removable walking boot for five weeks. The patient starts with dorsiflexion and plantar flexion exercises after eight days or as early as possible out of the boot two to three times a day. Return to sport is usually possible in three months.
Anatomical repair ofboth the ATFL and the CFL through bony tunnels produces good long term results and is recommended as the initial procedure in most cases. If anatomical repair fails, a tenodesis procedure, such as the Chrisman-Snook reconstruction, is a good alternative. A nonanatomical reconstruction is also indicated in patients with moderate arthritis or lax joints.