998 Radiographic Features in Septic Arthritis To the Editor: We found the radiographic features in septic arthritis described by Master et al. (1) extremely interesting. We would however question the interpretation of their Figures 4 and 5 . The position of the scapula in the two radiographs is not comparable. Thus, although the humeral head appears to have subluxed superiorly, its relationship to the glenoid remains unchanged. This is illustrated in our Figures 1 and 2 that show a normal shoulder with slight alteration of patient’s position. ISABELLAS. LIYANAGE, MRCP, FRCR, DMRD Consultant Radiologist SUNIL P. LIYANAGE, MRCP, DCH Consultant Rheumatologist Canadian Red Cross Memorial Hospital Taplow, England REFERENCE I . Master R, Weisman MH, ArmbusterTG,Slivka J, Resnick D, Goergen TG: Septic arthritis of the glenohumeral joint. Arthritis Rheum 201500-1506, 1977 Author’s Response The Doctors Liyanage are correct in stressing that the relative position of the humeral head t o the acromium is influenced by the attitude of the shoulder during radiography. We attempted in all instances to obtain comparable radiographs in our patients with septic arthritis of the glenohumeral joint but in some cases, because of patient discomfort, this was not possible. Thus, Figures 4 and 5 in our article are exposed in different positions. This certainly could account for an apparent change in the position of the humeral head. I n our experience, such change in patient position accounts for slight modification (up to 5 mm) of the position of the humeral head with respect to the inferior aspect of the tip of the acromium (the tip of this structure is Figures I and 2. These radiographs of the same normal shoulder demonstrate apparent subluxation due to change in position of patient. Nore rhat relationship of humeral head to glenoid is unchanged.