The Inter-rater Reliability of Shoulder Arthroscopy

Abstract

The purpose of this study was to determine the inter-rater reliability of orthopaedic shoulder surgeons in evaluating the intra-articular structures involved in real-time diagnostic shoulder arthroscopy by use of a videotape model. Twenty patients (nineteen male patients and one female patient) diagnosed with recurrent anterior shoulder instability with a mean age of 27.9 years (range, 15 to 44 years) consented to participate in the study. Standardized diagnostic shoulder arthroscopies (modified Snyder protocol) were done by a single experienced surgeon, and the procedures were videotaped. The patients' clinical information, radiographs, and videotape of the arthroscopy were sent to 6 experienced shoulder surgeons. The surgeons reviewed the clinical information, assessed the intra-articular structures shown on the videotape, and recorded their diagnoses on a standardized data collection form. The primary outcome was the median overall percent agreement for the video review surgeons by structure evaluated. The median values for overall agreement for the video review surgeons were as follows: anterior labrum, 90%; inferior labrum, 75%; superior labrum, 60%; posterior labrum, 65%; superior glenohumeral ligament, 50%; middle glenohumeral ligament, 50%; anterior inferior glenohumeral ligament, 25%; glenoid surface, 35%; detection of Hill-Sachs lesion, 85%; biceps tendon, 70%; supraspinatus tendon, 85%; infraspinatus tendon, 70%; and subscapularis tendon, 80%. The inter-rater reliability for orthopaedic shoulder surgeons' arthroscopic assessment of intra-articular anatomy in patients with anterior shoulder instability varied by structure examined. It was very good (>80%) for the anterior labrum and supraspinatus tendon and in detecting a Hill-Sachs lesion, poor (<40%) for the glenoid and anterior inferior glenohumeral ligament, and intermediate for all other structures examined. Level V, diagnostic study, expert opinion.