Abstract

Arthrographic distension of the glenohumeral joint was adopted as a mainstream treatment for frozen shoulder before any randomised controlled trials were performed. Interpretation of the effectiveness of this procedure rests mostly on data from cohort ­studies of which there are few of high quality. Papers reporting long-term results have either excluded ­diabetic patients or failed to report patient orientated outcomes. The authors present a long-term prospective cohort study of 51 patients (12 diabetics and 39 non-diabetics), with 53 frozen shoulders, who had an arthrographic distension performed by a single radiologist as a primary intervention. Oxford shoulder score (OSS), visual analogue pain score (VAS), and range of movement (ROM) were recorded pre-distension, at 2 days and 1 month post-distension. OSS and VAS were recorded again at a mean of 14 months post distension (range : 8-26 months). OSS improved from a pre-distension mean of 22.3 by 16.9 points at final follow-up (p < 0.001, 2 tailed paired samples t-test) whilst VAS improved from a mean pre-distension value of 7.1 by -3.5 (p < 0.001). ROM improved by a mean of 39.3 degrees in flexion, 55.2 degrees in abduction and 19.5 degrees in external rotation at one month (p < 0.001 for all). The outcome in diabetic ­patients was the same as in non-diabetic patients. ­Arthrographic distension is a safe and effective treatment for frozen shoulder ; it is also effective in ­diabetic patients. It gives long-term improvement. The authors believe that the low number of patients requiring a secondary procedure makes arthrographic distension preferable to manipulation under anaesthesia.