Aim: To conduct a radiographic assessment of the occurrence of the acromioclavicular joint subluxation in patients with long-standing rheumatoid arthritis.
Methods: A cohort of 74 patients with seropositive rheumatoid arthritis was followed prospectively for 15 years. At the 15-year visit, plain radiographs of both shoulders (n = 148) were obtained using a standard method. Acromioclavicular joint destruction was evaluated using Larsen's method, and radiographic subluxation was measured on the radiographs using two different methods. In one of these methods (method A), the distance between the upper edge of the coracoid process and the upper surface of the clavicle was measured. The other method (method B) involved measurement of the alignment of the upper joint margins of the acromion and clavicle.
Results and conclusion: Two of the 148 acromioclavicular joints fulfilled criteria for subluxation with method B. Neither measurement A nor measurement B differed significantly between unaffected joints (Larsen's grade 0 or 1) and affected joints (Larsen's grade > or = 2). In addition, neither measurement was correlated with the stage of acromioclavicular joint destruction (r < 0.10 for both measurements). Measurements A and B were significantly correlated to each other (r = 0.23; 95% CI, 0.07 to 0.38). The two subluxations (one upward and one downward) occurred among the nine joints with the most severe destructive lesions (Larsen's grade 5). In conclusion, acromioclavicular joint subluxation is rare in rheumatoid arthritis, occurring only when destruction of the joint is severe. The method B used in this study may prove useful for both population studies and routine clinical work.