Objective: To develop a simple but reproducible radiological scoring system for the hip in ankylosing spondylitis (AS).
Methods: A consensus approach was used to develop a grading system whereby hip radiographs of 470 patients with AS were scored from 0 to 4 (normal, suspicious, mild, moderate, and severe), producing the Bath Ankylosing Spondylitis Radiology Hip Index ( BASRI-hip or BASRI-h). The system was designed to be specific for AS radiographic change at the hip and includes circumferential joint space narrowing, osteophytes, erosions, and protrusio acetabulae. This score can then be added to the BASRI-spine, forming the grading system BASRI-total, scored 2-16. Radiographs of 134 patients were used to test reproducibility. Blinded radiographs of 100 non-AS patients were included randomly to assess disease specificity. Sensitivity to change was determined using 438 radiographs from 122 patients, over 219 time intervals.
Results: Inter and intraobserver variation for the right and left hip showed 83-84% and 94-96% complete agreement, respectively. Disease specificity was 0.76. If non-AS patients with a total hip replacement were excluded, disease specificity was 0.83. Sensitivity to change became apparent at one year (p < 0.001). Grading the hips using a single radiograph takes seconds.
Conclusion: BASRI-h is a reliable method for grading hip radiographic change in AS. It is disease-specific, sensitive to change, simple to use, and rapid to perform.