RESUMO
OBJECTIVE: Evaluate interobserver and intraobserver reliability of hip arthroscopic classifications for labral tears. DESIGN: Retrospective diagnostic study of nonconsecutive patients. SETTING: Institutional study. PATIENTS: From a database of 278 hip arthroscopy videos for treatment of femoroacetabular impingement, 70 videos were chosen by simple random sampling. Exclusion criteria included presence of radiological arthrosis (Tonnis > 2), previous hip surgery, inadequate lesion palpation, poor image quality, and refusal to participate in the study. The final sample included 60 videos. INTERVENTIONS: Four hip surgeons evaluated the videos twice at 1-month intervals and classified the lesions according to Lage, Seldes, and Beck classifications for hip labral tears. MAIN OUTCOME MEASURES: Interobserver and intraobserver reliability with the percent of agreement and weighted Cohen kappa values. RESULTS: Patients had a mean age of 33 years (SD, 7; range, 18-47 years), and 32 (53%) were men. Femoroacetabular impingement types included combined (CAM and pincer) in 31 (52%), CAM in 27 (45%), and pincer in 2 (3%). For interobserver reliability, the average weighted kappa values were 0.68, 0.65, and 0.78 for the Lage, Seldes, and Beck classifications, respectively. For intraobserver reliability, the mean weighted kappa values were 0.87, 0.64, and 0.93 for the Lage, Seldes, and Beck classifications, respectively. CONCLUSIONS: Beck classification had the highest average values for interobserver and intraobserver agreements. Lage, Seldes, and Beck scores for acetabular labrum tears showed substantial interobserver agreement. In the intraobserver evaluation, the Seldes system presented substantial agreement, whereas Lage and Beck classifications were considered excellent agreement.
