Validation of the Ottawa knee rule in adults: A single centre study.

INTRODUCTION: This clinical audit aimed to evaluate performance of the Ottawa Knee Rule (OKR) and degree of compliance by emergency referrers for acute knee injuries in adults. METHODS: Knee radiography requests were analysed retrospectively for eligibility. Data were extracted from eligible requests under headings describing the OKR criteria, patient history, diagnosis and referrer profession. Sensitivity, specificity, negative likelihood ratio and positive likelihood ratio were calculated with 95% CI for the entire sample and each profession (consultant doctors, resident medical officers [RMO], physiotherapists and triage nurses) individually. The frequency of each OKR criterion and correlation with fracture, referrer compliance to the rule and the relative reduction in radiography were also calculated. RESULTS: Of 713 patients identified, 149 were enrolled by the eligibility criteria. The overall sensitivity, specificity, negative likelihood ratio and positive likelihood ratio of the OKR for knee fracture were 71% (95%CI, 49-87%), 46% (95%CI, 37-55%), 0.64 (95%CI, 0.33-1.22) and 1.3 (95%CI, 0.96-1.76), respectively. Physiotherapists and triage nurses demonstrated better rule performance than consultant doctors and RMOs, with a sensitivity of 100% and negative likelihood ratio of 0.0. Physiotherapists were most compliant at 73% (19/26). Only 85 requests were OKR positive and, when abiding by the rule, this would have reduced radiography by 43% (64/149). CONCLUSIONS: In this first Australian study, moderate OKR performance and variable compliance by emergency referrers were observed. This led to unnecessary irradiation of patients without a fracture. The findings suggest emergency referrers could benefit from education on applying and documenting the OKR on radiography requests.