Ottawa Ankle Rules for assessment of ankleinjuries

Kwok-biu Tsui 徐國標, Leo CW Kong 江志宏, Luke CY Tsang 曾昭義

HK Pract 2011;33:22-28

Summary

Foot and ankle x-ray examinations are commonly ordered by family physicians for ankle injuries. The 1993 Ottawa ankle rule project demonstrated that more than 95% of patients with ankle injuries had radiographic examinations but that 85% of the films showed no fractures. To address the problem of cost-effectiveness, a group of Ottawa emergency physicians developed two rules to identify clinically important fractures of the malleoli and the midfoot. Use of these rules reduced radiographic examinations by 28% for the ankle and 14% for the foot.

The Ottawa Ankle Rules are examples of clinical decision rules, which are multifactorial tools used to aid in clinical decision making. In radiology, clinical decision rules are an important method for determining who should undergo imaging. To be useful, clinical decision rules should be clinically important and relevant, have face validity, be reproducible and easy to use. They should also be validated in subjects distinct from those used to develop the rules.

This paper aims to introduce Ottawa Ankle Rules, and shows that evidence supports it as an accurate instrument for excluding fractures of the ankles and mid-foot. The applicability of the Ottawa ankle rules in Asian populations and in primary care setting is also discussed.

摘要

家庭醫生會經常為腳踝受傷進行足部和腳踝的X光檢驗。於1993年按渥太華腳踝規則進行之研究顯示， 在為超過9 5%的腳踝受傷患者進行X光檢查中，95%的X光片並沒有顯示骨折。為關注這項檢查的成本效益問題， 一組渥太華急診科醫生重新建議用兩項規則，來鑒別臨床上重要的足踝和足弓骨折。按這些規則進行檢查， 減少了28%的腳踝和14%的腳部X光檢查。

渥太華腳踝規則是臨床決策規則的例子。它是一種工具，利用多項因素的特性來輔助臨床上決定。在放射科， 臨床決策規則是一種重要方法以決定誰應該進行影像檢查。為能達到成效，臨床決策規則具有臨床上的重要性和相關性，有表面度效， 可重複性和易於使用。而在不同的病人組別仍然有效使用。

本文旨在介紹渥太華腳踝規則，並列出支持它作為一種準確排除足踝和足弓骨折方法的證據。 渥太華腳踝規則在亞洲人士和基層醫療環境的適用性亦加以討論。

Background

General practitioners frequently encounter ankle injuries, especially inversion injuries. It is one of the commonest musculoskeletal injuries in athletes as well as sedentary persons, accounting for an estimated 2 million injuries per year and 20% of all sports injuries in the United States.¹ Ankle injuries are usually due to simple fall or a twist. Majorities are just ligamentous soft tissue injuries, or small avulsion fracture of little clinical significance. A small number of them suffer from more serious fractures, which require plaster treatment or internal fixation. Though prevalence of fractures is low in a primary care setting, x-ray is often routinely ordered to exclude bony injuries. At emergency departments, the management of ankle injuries is daily routine. Though most patients undergo radiography, fracture occurs in less than 15%.^2-5

Studies in the past have tried to identify clinical indicators to determine the presence of an ankle fracture. These included colour change, mechanism of injury, age of patient and presence of swelling.^2,4,6-12 However, these studies were not properly validated. Some studies even had contradictory results. For instance, in the two studies by deLacey G et al8 and Cockshott WP et al ⁹, swelling was considered to be a sine qua non of fracture. But in the study by Dunlop MG et al, 11% of malleolar fractures manifested no swelling.⁴

In view of the inconsistent assessments and use of radiological investigations, and the small yield of positive results, Stiell and colleagues at the University of Ottawa developed a set of clinical parameters-/clinical decision rules to help guide the management of ankle injuries and to determine the indications for radiography of ankle and midfoot.

Methodology

A review of literature on Ottawa Ankle Rules and ankle injuries was undertaken. The Cochrane Library, Ovid and Medline databases were searched for the following keywords and MeSH terms: ankle, fracture, ankle injuries, ankle sprain, ankle fracture, midfoot fracture, radiography, radiology, x-ray*, physical examination, Ottawa Ankle Rules (OAR), clinical decision rule*, clinical prediction rule*, guideline*, Asian*, Hong Kong, primary care, family medicine, family practice, and general practice. In addition, the references in the articles obtained were scanned to identify published studies of potential interest, and other relevant and up-to-date articles.

Twenty-four articles were identified. Eleven were about general discussion on ankle injury and x-ray examination of the injury.^1-11 Six were about development, validation and implementation of OAR.^12-17 The following questions were used in evaluation of these six papers: whether the test was performed in an appropriate spectrum of patients, whether there was an independent comparison with a reference standard of diagnosis, and whether the test was validated in a second, independent group of patients. All answers were positive, which showed that the results of the studies were valid. For discussion of clinical decision rules, two papers regarding trauma imaging were included.^18-19 For systematic review of the accuracy of OAR, one important article was identified.²⁰ Two international guidelines on imaging in ankle injury were found.^21-22 One is a publication and the other can be accessed through the internet. For discussion on applicability of OAR in Asian and in primary care, only two of each can be found.^23-26 Though the studies were of small scale, they are included for discussion since there are no other similar articles identified.

The Ottawa Ankle Rules (OAR)

To address the problem of frequent unnecessary radiographs for acute ankle injuries, Stiell and colleagues conducted a series of studies in phases.¹³

1. Phase zero: Efficiency of use of ankle x-ray and attitudes of clinicians ¹⁴

A two-stage study was designed. They included patients who presented with ankle pain or tenderness after they had sustained blunt ankle injuries. "Ankle" was subdivided into two zones representing the malleolar region and the midfoot, which included the following anatomic structures and their overlying soft tissues: (a) the malleolar region (distal 6 cm of tibia, distal 6 cm of fibula, and talus) and (b) the midfoot (navicular, cuboid and cuneiform bones, and the anterior process of the calcaneus and the base of the fifth metatarsal bone. In the first stage, the efficiency of the use of ankle radiography was examined by retrospective chart review. They concluded that emergency physicians relied heavily and rather inefficiently on radiography. In the second stage, the judgments and attitudes of experienced clinicians in their use of ankle radiography were studied through a prospective survey. They showed that experienced emergency physicians could accurately distinguish fracture from nonfracture cases but they were reluctant to use this clinical skill.

2. Phase one: Development of the OAR ¹²

The study objective was to develop decision rules that would predict fractures in patients with ankle injuries. A prospective survey of 750 emergency department patients was carried out in 1990. 32 clinical variables were assessed. A few variables were found to be highly associated with fractures. They were included in the first rule which stated that an ankle x-ray series was only necessary if there were pain near the malleoli and one or more of the following findings: 1) age of 55 years or more, 2) inability to bear weight both immediately and for four steps in the emergency room, 3) bone tenderness of either inferior tip or posterior edge (distal 6 cm) of lateral malleolus, and 4) bone tenderness of either inferior tip or posterior edge (distal 6 cm) of medial malleolus.

Lastly, 364 patients preferred HbA1C monitoring; most of them, 219 patients (60.2%), said they preferred it because it was accurate.

A similar process determined the best predictor variables for midfoot fracture on a foot radiographic series. The second rule stated that a foot x-ray series is only necessary if there is pain in the midfoot and there is bone tenderness at the navicular, cuboid or the base of the fifth metatarsal.

In this study, 689 ankle radiographic series were performed. All 70 significant malleolar fractures were identified when the first rule was applied. Similarly, all 32 significant midfoot fractures on the 230 foot radiographic series performed were found when the second rule was applied.

3. Phase two: Validation and refinement of the OAR ¹⁵

The study objective was to validate and refine previously derived clinical decision rules. 1,032 patients were studied. All 121 malleolar zone fractures were identified, thereby achieving the goal of sensitivity of 1.0. The foot series rule identified 48 out of 49 significant midfoot fractures. The refined Ottawa rules retained a sensitivity of 1.0 even after exclusion of the age and cuboid criteria. To validate the refined rules, 453 patients were studied. The refined rules correctly identified all 50 ankle series fractures and all 19 foot series fractures. No fracture was missed.

The refined rules were:

First rule:

An ankle x-ray series is only necessary if there is pain near the malleoli and any one of the following findings: 1) inability to bear weight both immediately and for four steps in the emergency department, or 2) bone tenderness of either inferior tip or posterior edge (distal 6 cm) of lateral malleolus, or 3) bone tenderness of either inferior tip or posterior edge (distal 6 cm) of medial malleolus.

Second rule:

A foot x-ray series is only necessary if there is pain in the midfoot and any one of these findings: 1) inability to bear weight both immediately and for four steps in the emergency department, or 2) bone tenderness at the navicular, or 3) bone tenderness at the base of the fifth metatarsal.

4. Phase three: Implementation of the OAR ¹⁶

In phase 3, they put the OAR into practice in the Ottawa Civic Hospital Emergency Department. Compared with baseline levels and compared with a nearby "control" community hospital, they found that radiography for 593 patients was reduced by 28% for ankle radiographs and by 14% for foot radiographs. Patients discharged without radiography spent 36 minutes less in the emergency department, had much lower estimated costs for radiography and physician visits, were equally satisfied with physician care, and were no more likely to have subsequent x-ray examinations than were patients who did undergo radiography in the emergency department. The OAR accurately identified all 93 malleolar and midfoot fractures in the 593 patients.

The OAR as clinical prediction rules

Clinical prediction (or decisive) rules attempt to reduce the uncertainty of medical decision-making by standardizing the collection and interpretation of clinical data. A clinical rule is derived from original research and may be defined as a decision-making tool that incorporates several relevant variables from the history, physical examination or simple tests.¹⁷ The process of development and validation of clinical prediction rules are costly and time-consuming. For ankle injuries, we have the Ottawa Ankle Rules as clinical prediction rules to identify patients with ankle injuries who do not require imaging. Both the first and second rules are based on simple criteria. For the first rule (and second rule): inability to bear weight, tenderness at the posterior edge or inferior tip of the medial malleolus (navicular for second rule) and tenderness in a similar distribution at the lateral malleolus (base of fifth metatarsal for second rule).

To be useful, a clinical prediction rule should meet the following criteria:¹⁸

1. It should be clinically important and relevant. The underlying clinical issue must be of sufficient importance to justify the research needed to derive the rule. X-ray of ankle is a very common procedure performed in emergency department and is thus clinically important. Missing ankle fractures affects the patient's future ability for mobilization, thus it is relevant.
2. It should have face validity. Face validity is the extent to which the test actually measures what it is supposed to measure. For wide adoption, practitioners must accept the logic of the clinical prediction rule. For OAR, focal tenderness corresponding to common sites of fracture is one of the factors under study, thus here imparting face validity.
3. It should be reproducible. The performance of a clinical prediction rule in actual clinical practice is more important than its performance in the original research setting.
4. It should be easy to use by health care professional. The OAR consist of simple criteria that can be understood by persons with lesser medical expertise. Derksen RJ et al showed that specialized emergency nurses were able to assess ankle and foot injuries using OAR with high level of accuracy after attending a short and inexpensive course.¹⁹
5. It must have been validated in subjects distinct from those used to develop the rule so that it can demonstrate generalisability.

Validation of the OAR

L M Bachmann et al conducted a systematic review in 2003 on the accuracy of the OAR.²⁰ They electronically searched databases, checked the reference lists of included studies and contacted experts and authors in the specialty. Twenty-seven studies reporting on 15,581 patients were used for meta-analysis. Six of them were on assessment of the ankle or mid-foot in children. Sensitivities were consistently high, ranging from 99.6% in studies on application of the rules within 48 hours of injury to 96.4% in studies of combined assessment of ankle and mid-foot. The specificities ranged from 47.9% to 26.3%. Less than 2% of patients in most subgroups who were negative for fracture according to the Ottawa Ankle Rules actually had a fracture. They concluded that evidence supports the OAR as an accurate instrument for excluding fractures of the ankle and the mid-foot after sustaining injury', using x-ray as the gold standard. The instrument has a sensitivity of almost 100% and a modest specificity. They estimated that the OAR would decrease the number of unnecessary radiographs by 30-40%.

The Royal College of Radiologists also has similar recommendations for performing radiological investigations for acute ankle injuries. The guideline states that the features that justify x-ray include inability to weight bearing immediately and in the emergency room, point tenderness over the medial malleolus, and/or the posterior edge and distal tip of the lateral malleolus (Recommendation grade B).²¹ The American College of Radiologists adopts the OAR as guidance for ordering x-ray for suspected ankle fracture. The rule is ranked as most appropriate (=9) in the rating scale of 10.²²

The Ottawa Ankle Rules is an instrument that is calibrated towards high sensitivity. There is room for improvement of its low specificity. Further modification or refinement of the rules may increase the specificity, i.e. to decrease the number of false-positive results.

Are the OAR applicable in Asian populations?

The population chosen for development of OAR was originally Caucasian. To test whether the rules can be applied to Asians, two studies have been done: one in Singapore and one in Hong Kong. Both these studies were included in the meta-analysis by L M Bachmann et al. The study done by Tay et al in Singapore was completed in 1999.²³ There were 494 patients (age 12 years and above) enrolled in the study, predominantly Chinese, Malay and Indian. These 494 consecutive eligible patients presenting to the emergency department in a large teaching hospital with twisting injuries about the ankle were examined by emergency physicians for clinical criteria requiring ankle and foot x-ray studies according to the OAR. 488 of these patients underwent x-ray studies that were interpreted by a radiologist. The limitation of the study was that the radiologists were able to study the datasheet, and thus were not blinded to the clinical findings. The sensitivity and specificity of the OAR for predicting the presence of fracture were calculated to be 0.9 and 0.34, respectively. They also calculated sensitivity and specificity for a modified version of the OAR that included patients who were unable to bear weight either immediately after injury or in the emergency department (as opposed to both/and as described in the original rules). The sensitivity improved to 0.99. They concluded that the OAR cannot be used to screen for the need for x-ray studies in their Asian patients who had sustained twisting ankle injuries because of inadequate sensitivity, but when modified become acceptable, and can reduce the number of x-ray studies requested by 28%.

A prospective study in Hong Kong was conducted in 2000 in the Kwong Wah Hospital.²⁴ 773 eligible patients were enrolled. Consecutive Chinese patients age 16 and older with ankle and midfoot pain after twisting injury were examined by emergency physicians trained in the use of the OAR. All patients were sent for radiographs (the gold standard). The radiologists were blinded to the results of the OAR and the emergency physician's interpretation. 131 clinically significant fractures (53 ankle fractures and 78 mid-foot fractures) were identified. One significant distal fibula fracture was missed. The sensitivity and specificity of the OAR for ankle injuries was 98% and 40.8%. For midfoot, the respective results were 100% and 43.8%. Their conclusion was that the OAR were highly sensitive and were accurate clinical decision rules for guiding decisions for ordering radiographs for Chinese patients with foot and ankle injuries.

Since there is a paucity of studies for validation of the OAR for Asians, further trials of larger scale are needed to ascertain whether these rules are applicable to the Asian population.

Are the OAR applicable in primary care setting?

The original setting for development of the OAR was in accident and emergency departments in large hospitals. The rules have been validated in such setting in many studies. However, little research in family medicine discusses office use of the OAR. GF Smith et al conducted a study in 1993 on clinical evaluation of ankle inversion injuries in family practice offices. They did not develop any decision rule, but found out that the following clinical factors were significantly associated with the presence of a fracture: tenderness on the dorsum of the foot, impaired weight bearing ability, recentness of injury (less than 48 hours), and presence of additional injuries. Each of these findings had a negative predictive value of 94% or more. They also estimated that there would be a 30% reduction in x-ray utilization if these criteria were used, which was similar to the findings of OAR.²⁵

Wynn-Thomas et al attempted to validate the OAR in a New Zealand primary health care setting in 2001.²⁶ The research was conducted in two "afterhours" medical centres. 200 patients were recruited. Prospective data were collected on eligible patients over a six-month period. The general practitioner managed patients in their usual manner and recorded their findings on a data collection sheet that included (but did not emphasize) the observations necessary to apply the OAR. X-ray results were later recorded by the research team. Ottawa positive and negative status was determined retrospectively from the data for each patient by two GPs working independently and blinded to the x-ray results. In total, 113 patients were assessed as Ottawa positive, and 13 of them had significant fractures. A total of 133 patients received x-rays. 13 significant fractures were diagnosed. The overall sensitivity of the OAR was found to be 100%. The authors concluded that this study validated the OAR.

In the primary care setting, the OAR can be used as a reference when dealing with patients with ankle injuries. The rules are not meant to override clinical judgment. In conditions such as gross deformity, we clearly need an x-ray without using a clinical decision rule. We should also be cautious in patients with multiple painful injuries, diminished peripheral sensation, or altered mental status due to head trauma or drug intoxication, or if gross swelling makes palpation of bone tenderness very difficult. On the other hand, some non-clinical factors, such as demand from patients and the worry of medico-legal consequences, may affect a doctor's decision to order investigations.

Conclusion

The Ottawa Ankle Rules offer the attending doctor an opportunity to use clinical judgement to assess patients with acute ankle injuries for the need for radiography. The application of simple clinical observations would help to differentiate which patients are at negligible risk for a fracture and thus need not undergo radiography. The rules have been validated in the hospital setting. So far, however, the usefulness of the rules as a decision tool in primary care has been assessed only in one study. More discussion, refinement and dissemination of this approach among general practitioners may therefore be pertinent.²⁰