Ajay Malviya1, Ali Raza2, Johan D Witt3. 1. Northumbria Healthcare NHS Foundation Trust, Wansbeck General Hospital, Ashington - UK. 2. Northern Deanery Orthopaedic Rotation, Newcastle upon Tyne - UK. 3. University College of London Hospital NHS Trust, London - UK.
Abstract
PURPOSE: To evaluate the degree of inter-rater agreement in identifying important radiographic features of femoroacetabular impingement (FAI) and hip dysplasia. In addition, the effect of supplementary investigations (CT and/or MRI) and clinical experience of the observer was determined. METHODS: During the British Hip Society meeting 2013 participants were asked to remark on various radiological parameters of FAI and dysplasia, make a diagnosis, and recommended treatment. Free-Marginal Multirater Kappa analysis was performed to determine multi-rater agreement. RESULTS: The 'crossover sign', 'ischial spine sign' and 'lateral centre-edge angle' were found to have the highest inter-rater agreement. For the whole group only fair agreement was seen for the diagnosis of type of FAI (κ = 0.3) and dysplasia (κ = 0.3) after plain radiographs with supplementary investigations being only marginally helpful in improving the diagnosis of type of FAI (κ = 0.4). In contrast, high-volume hip preservation surgeons had a substantial improvement in diagnosis after supplementary investigations (type of FAI κ = 0.75, and dysplasia (κ = 0.65). CONCLUSIONS: Our study shows that agreement on diagnosis and treatment on the basis of plain radiographs is low amongst general hip surgeons. This improved with surgeon volume and experience, suggesting that protocols for diagnosis of these conditions need to be altered as assessment appears to be rather subjective.
PURPOSE: To evaluate the degree of inter-rater agreement in identifying important radiographic features of femoroacetabular impingement (FAI) and hip dysplasia. In addition, the effect of supplementary investigations (CT and/or MRI) and clinical experience of the observer was determined. METHODS: During the British Hip Society meeting 2013 participants were asked to remark on various radiological parameters of FAI and dysplasia, make a diagnosis, and recommended treatment. Free-Marginal Multirater Kappa analysis was performed to determine multi-rater agreement. RESULTS: The 'crossover sign', 'ischial spine sign' and 'lateral centre-edge angle' were found to have the highest inter-rater agreement. For the whole group only fair agreement was seen for the diagnosis of type of FAI (κ = 0.3) and dysplasia (κ = 0.3) after plain radiographs with supplementary investigations being only marginally helpful in improving the diagnosis of type of FAI (κ = 0.4). In contrast, high-volume hip preservation surgeons had a substantial improvement in diagnosis after supplementary investigations (type of FAI κ = 0.75, and dysplasia (κ = 0.65). CONCLUSIONS: Our study shows that agreement on diagnosis and treatment on the basis of plain radiographs is low amongst general hip surgeons. This improved with surgeon volume and experience, suggesting that protocols for diagnosis of these conditions need to be altered as assessment appears to be rather subjective.