This study was part of a larger prospective study of injuries of the syndesmosis of the ankle joint conducted at Oulu University Hospital between 2007 and 2009. All skeletally mature patients (≥16 years old) with an unstable unilateral ankle fracture that resulted from a supination-external rotation mechanism of injury (Lauge-Hansen type SE) and that was treated at our hospital within a week after injury during this period were considered eligible for the present study. Patients were excluded if they had bilateral ankle fractures, a pathologic fracture, a concomitant tibial shaft fracture, a previous fracture or other major injury of either ankle, major peripheral neuropathy, or soft-tissue infection in the region of the injured ankle. Eighteen of the 166 patients identified by the original screening met these exclusion criteria, and eight additional patients were excluded because the surgical procedure could not be carried out according to the study protocol because of operating room congestion. The final study cohort consisted of 140 patients (seventy-one women and sixty-nine men) with a Lauge-Hansen supination-external rotation type-4 ankle fracture resulting from a supination-external rotation mechanism. The mean age (and standard deviation) was 47 ± 16.0 years (range, sixteen to eighty-four years).
The fracture was classified with use of the Lauge-Hansen classification system by the senior orthopaedic trauma surgeon who was responsible for the patient (H.P., P.H., M.L., or J.R.). All surgeons were familiar with the Lauge-Hansen classification and had used it in their clinical practice for several years. Four types of ankle fractures resulting from a supination-external rotation mechanism of injury were considered unstable: (1) a bimalleolar fracture, (2) a trimalleolar fracture, (3) a fracture of the lateral malleolus with medial tenderness or hematoma and a positive stress test, and (4) a fracture of the lateral malleolus with any talar shift or talar tilt on mortise or lateral radiographs of the ankle. All surgical procedures were performed by either two senior orthopaedic trauma surgeons or a senior orthopaedic trauma surgeon and a senior orthopaedic trauma resident who had completed our trauma training. The principles of the hook test and the external rotation stress test were summarized and rehearsed before the surgery.
Internal fixation of the malleolar fracture was carried out according to standard AO (Arbeitsgemeinschaft für Osteosynthesefragen) principles: The lateral malleolus was reduced and rigidly fixed with two 3.5-mm lag screws, a one-third tubular plate, or a lag screw and a plate. A fracture of the medial malleolus was reduced and fixed with two 4.0-mm partially threaded cancellous screws. A fracture of the posterior malleolus was fixed with 4.0-mm partially threaded cancellous screws from anterior to posterior if the fragment involved >30% of the articular surface as estimated from the lateral radiograph.
After the malleolar fixation, one of the surgeons temporarily left the operating room while the other surgeon performed the hook test and the external rotation stress test under fluoroscopy. The surgeon then left while the second surgeon carried out the same tests.
In the hook test, the tibia was stabilized with one hand, the lateral malleolus was grabbed with a bone hook, and a lateral force was applied with use of the bone hook. The hook test was considered positive if >2 mm of lateral movement of the fibula was observed under direct vision.
In the external rotation stress test, the tibia was stabilized with one hand and an external rotation force was applied to the foot; the tibiotalar clear space was then assessed under fluoroscopy. The images were printed, and the surgeon measured the radiographic parameters to the nearest millimeter with a caliper. The external rotation stress test was considered positive if the medial tibiotalar clear space was ≥5 mm under stress.
After completion of these tests, the other surgeon returned to the operating room and the standardized 7.5-Nm external rotation stress test was carried out for each ankle under fluoroscopy with use of a fork-like F-tool as described by Jenkinson et al.8. The 7.5-Nm torque value was chosen on the basis of cadaver studies that used the application of external rotation stress to evaluate injuries of the syndesmosis12,19,20. The F-tool was applied to the medial aspect of the forefoot and the lateral aspect of the hindfoot, and the proximal aspect of the tibia was stabilized with one hand by the other physician. The angle between the tibial shaft and the foot was held at 90°. The F-tool was then used to apply the standardized 7.5-Nm external rotational torque at the level of the ankle mortise8. The operating surgeon intraoperatively evaluated the tibiotalar and the tibiofibular clear space in each ankle under fluoroscopy as described above. The tibiofibular clear space was measured at the level of the physeal scar approximately 1 cm proximal to the tibial plafond7,21,22, and the tibiotalar clear space was measured as the distance between the lateral border of the medial malleolus and the medial border of the talus at the level of the talar dome8,23. A positive stress test was defined as a side-to-side difference of >2 mm in the tibiotalar or the tibiofibular clear space on the mortise radiographs. The fluoroscopic measurements were normalized to account for magnification by using the dimensions of the head of a 3.5-mm small-fragment fixation screw and the arm of the F-tool as references.
The interobserver reliability of the hook test and the external rotation stress test was assessed, and the sensitivity and specificity were calculated with use of the standardized 7.5-Nm external rotation stress test as a reference. The sensitivity and specificity were calculated for three different scenarios: (1) at least one of the two surgeons reported a positive result on the clinical test being studied, (2) both surgeons reported a positive result on the clinical test being studied, or (3) at least one of the surgeons reported a positive result on one or both tests.
Local ethics review board approval was obtained before study recruitment began, and informed consent for study participation was obtained from each patient preoperatively.
Statistical Methods
Interobserver reliability was characterized with use of the percentage agreement between assessors and the kappa coefficient. The percentage agreement gives an overview of the agreement between two assessors, but does not take into account the agreement that can occur purely by chance. The kappa coefficient and associated 95% confidence interval (CI) were calculated with use of SPSS version 16.0 (SPSS, Chicago, Illinois). Interpretation of the kappa values was carried out according to the guidelines proposed by Fleiss24 and by Landis and Koch25. According to Fleiss24, a kappa value of <0.40 represents poor interobserver reliability, 0.40 to 0.75 represents fair to good reliability, and >0.75 to 1.0 represents excellent reliability. According to Landis and Koch25, a kappa value of 0.00 to 0.20 represents slight reliability, 0.21 to 0.40 represents fair reliability, 0.41 to 0.60 represents moderate reliability, 0.61 to 0.80 represents substantial reliability, and >0.80 represents almost perfect reliability.
Source of Funding
There was no outside source of funding for this study.
In this study, we found that the hook test and the clinical stress test both had excellent interobserver agreement when used to diagnose instability of the distal tibiofibular joint following repair of an ankle fracture resulting from a supination-external rotation mechanism. The specificity of both clinical tests was also excellent. However, the sensitivity was poor for the hook test and fair for the clinical stress test when a standardized 7.5-Nm external rotation stress test was used as the reference. The sensitivity was even worse when both surgeons were required to agree on a positive finding of an unstable distal tibiofibular joint. When a combination of the hook test and the clinical stress test was used, the sensitivity was comparable to that of the clinical stress test alone.
The poor sensitivity values indicated that neither test had an adequate ability to identify instability of the distal tibiofibular joint after internal fixation of the malleolar fracture. An undetected injury of the ligaments of the syndesmosis may lead to late instability, pain, and arthrosis5,7,11. The excellent specificity of these clinical tests, however, would be able to prevent unnecessary syndesmotic transfixation, which often leads to malreduction26. Clinical studies have indicated that the accuracy of reduction of the syndesmosis is a critical factor in achieving good clinical results5,7,10.
These tests may have poor sensitivity for several reasons. During application of the hook test, it is difficult to determine precisely how much force should be applied to the fibula, in which direction the force should be applied, and how much displacement should be required to indicate instability17. A similar difficulty probably applies to the clinical stress test, since the operating surgeon may not stress the ankle with sufficient force in order to protect the osseous fixation. It is our experience that a 7.5-Nm external rotation force at the level of the ankle mortise is considerably greater than the force that is exerted during a typical clinical stress test, and this may explain the poor to fair sensitivity of the clinical test when the standardized test is used as the reference.
To our knowledge, this study is the first to report the interobserver reliability of intraoperative clinical tests of the stability of the syndesmosis. The excellent reliability that we calculated for each test was primarily the result of the high proportion of stable distal tibiofibular joints. Thus, our results must be interpreted very cautiously. We examined only ankles with fractures resulting from a supination-external rotation mechanism, for which a concomitant injury of the ligaments of the syndesmosis is believed to be infrequent. Since the prevalence and severity of such concomitant ligament injuries may be totally different for ankle fractures resulting from a pronation-external rotation mechanism (equivalent to AO/OTA or Weber type C), conclusions regarding the reliability of the two tests for detecting ligament injuries that do not result from a supination-external rotation mechanism cannot be made.
The intraoperative 7.5-Nm external rotation stress test performed under fluoroscopy is the only standardized method for examining the stability of the distal tibiofibular joint. In our study, this standardized test was able to detect ankle instability resulting from injury to the ligaments of the syndesmosis more frequently than the nonstandardized test could after internal fixation of the malleolar fracture, in accordance with the findings of Jenkinson et al.8. To our knowledge, no previous studies have compared intraoperative clinical tests of the stability of the syndesmosis. However, in a cadaver study, the hook test was found to be better able to detect instability of the distal tibiofibular joint than the external rotation stress test was20, which contrasts with the results of our study.
Previous studies involving the stress test have indicated that the maximum increase in the width of the syndesmosis that is compatible with satisfactory function is 1 to 2 mm5,27,28. In the present study, a positive result on the stress test was defined as a >2-mm side-to-side difference in the tibiotalar or the tibiofibular clear space compared with that in the contralateral, uninjured limb. Jenkinson et al.8 used a >1-mm side-to-side difference, which may be more susceptible to measurement errors than our >2-mm side-to-side difference was. The 7.5-Nm external rotation stress test was positive in 17% of the patients in our study after malleolar fixation, which is approximately one-half of the percentage reported by Jenkinson et al8. This difference is probably attributable to our use of a >2-mm side-to-side difference rather than a >1-mm difference as a positive finding.
The strengths of our study are its prospective design, large number of patients, and use of a reproducible, previously published method. The clinical tests that we used are familiar to orthopaedic surgeons and are the most commonly used intraoperative tests for detecting instability of the distal tibiofibular joint2,29.
Our study also has some limitations, including the relatively small number of patients with instability of the syndesmosis due to a ligament injury. As a result, the confidence intervals were quite wide, especially that for the sensitivity. Although we are aware of only one previously published clinical study of the use of the 7.5-Nm external rotation stress test to detect instability of the distal tibiofibular joint8, we believe that this test is less prone to error than other tests are. However, more study is needed to confirm its validity and reliability.
The clinical relevance of injury (and repair) of the syndesmosis associated with ankle fractures resulting from a supination-external rotation mechanism is poorly understood. It is unclear whether transfixation of the distal tibiofibular joint is necessary to achieve a good result following an ankle fracture resulting from such a mechanism. If an osseous injury heals anatomically after malleolar reduction, the injured ligament may also heal at the proper length. However, the effect of possible widening of an unfixed, unstable distal tibiofibular joint on the functional results is unknown. As we are not aware of any published clinical studies reporting the medium or long-term radiographic or functional results of instability of the syndesmosis following treatment of a Lauge-Hansen supination-external rotation type-4 ankle fracture resulting from a supination-external rotation mechanism, we have begun a prospective randomized trial to obtain information regarding the medium-term results of such injuries. We also recommend reexamining the validity of the two clinical tests for ankle fractures resulting from a pronation-external rotation mechanism.