Radiography in acute ankle injuries: the Ottawa Ankle Rules versus local diagnostic decision rules

STUDY OBJECTIVE: We validate the Ottawa Ankle Rules and 2 Dutch ankle rules in distinguishing clinically significant fractures from insignificant fractures and other injuries in patients with a painful ankle presenting to the emergency department. METHODS: This prospective comparison of 3 ankle rules was conducted in the ED of a 580-bed community teaching hospital in Amsterdam from January 1998 to April 1999. Participants included 647 consecutive patients aged 18 years or older presenting with a painful ankle after trauma. All physicians received extensive and pictorial training on how to correctly score the respective items of the rules. The physician on call recorded these items derived from history and physical examination on a standardized data sheet. All patients subsequently underwent standard radiographic assessment. A radiologist and a trauma surgeon evaluated the radiographs blinded from the results of the data sheet form and the treatment given. The diagnostic performance of the 3 rules was measured in terms of sensitivity, specificity, and the reduction of radiographs. Receiver operating characteristic (ROC) curves were constructed, and the area under the ROC curves was calculated and compared. RESULTS: Seventy-four fractures were seen, of which 41 were clinically significant. The Ottawa Ankle Rules had a sensitivity of 98% for identifying clinically significant fractures; the local rules scored 88% and 59%, respectively. The potential savings in radiographs for the 3 decision rules were 24%, 54%, and 82%, respectively. The area under the ROC curve was better for both the local rules (0.84 and 0.83) compared with the Ottawa Ankle Rules (0.76). CONCLUSION: Because the identification of all relevant fractures is more important than a reduction in radiographs, the higher sensitivity of the Ottawa Ankle Rules makes these most suitable for implementation in The Netherlands.