Accuracy of intra-articular injection of the glenohumeral joint: a modified anterior approach.

This study examined the accuracy of our modified anterior approach to the glenohumeral joint using arthrography. Two hundred fifty-six consecutive patients with adhesive capsulitis received the glenohumeral joint injection: a mixture of 1 mL of 40 mg of triamcinolone hexacetonide, 2 mL of 2% lidocaine, 3 mL of contrast medium (Ioxitalamate), and 4 mL of normal saline solution, for a total of 10 mL, through a modified anterior approach. Anteroposterior, lateral, and axial arthrography after each injection was performed to determine the accuracy and patterns of the injection. The injection was judged to be successful if some of the contrast medium was placed within the joint. There were 78 male and 178 female patients, and the mean age was 54 years (range, 28 to 87 years). In 232 of the 256 patients (90.6%), the injection was deemed to have reached the glenohumeral joint, and these cases were considered successful. Among these patients, the contrast medium was shown only in the joint in 189 (73.8%), whereas in 43 (16.8%) it was shown both in and out of the joint. Of the 24 misguided injections, 15 (5.9%) had anterior placement, 6 (2.3%) had posterior placement, and 2 (0.8%) had superior placement, whereas the material was located in all portions outside the joint in 1 (0.4%). This study showed that our modified anterior approach had an accuracy greater than 90% for the intra-articular injection of the glenohumeral joint without radiographic guidance.