Interobserver Reliabilities of 2 Classifications for Arthroscopic Classification of the Pulvinar Tissue
W. Michael Pullen, Kinsley J. Pierre, Nicole Simunovic, Nicole S. Pham, Stephen Aoki, Olufemi R. Ayeni, Thomas Byrd, Dominic Carreira, T. Sean Lynch, Richard C. Mather, Ivan Ho-Bun Wong, Marc R. SafranBackground:
The pulvinar is located adjacent to the medial wall of the acetabulum and consists of fat and fibrous connective tissue. Patients experiencing hip pain who undergo hip arthroscopy may exhibit varied appearances of pulvinar pathology. The precise significance of these variations remains unknown.
Purpose:
To evaluate the interobserver reliability of 2 arthroscopic grading systems for the pulvinar to allow for future study.
Study Design:
Cohort study (Diagnosis); Level of evidence, 3.
Methods:
In total, 300 intraoperative arthroscopic photographs of the pulvinar were reviewed by 9 experienced hip preservation surgeons. Images were reviewed for quality and classified using 2 systems: system A (descriptive) and system B (Likert). System A consisted of 6 descriptive options: normal, mild synovitis, inflamed/hemorrhage, hyperplasia, fibrotic, and other (eg, pigmented villonodular synovitis, synovial chondromatosis). Alternatively, system B used a 4-point Likert scale: normal, mild synovitis, moderate synovitis, and severe synovitis. A power analysis conducted prior to the study revealed that a minimum sample size of 251 would result in a 95% confidence interval with a margin of 0.1 when the κ statistic reached at least 0.60. Statistical examination was performed to provide a Fleiss κ score with a 95% confidence interval. Agreement was classified as poor (<0), slight (0-0.2), fair (0.21-0.4), moderate (0.41-0.6), substantial (0.61-0.8), and almost perfect (>0.8).
Results:
A total of 300 intraoperative arthroscopic photographs of the pulvinar were reviewed and rated. System A demonstrated an interobserver reliability of κ = 0.28 (95% CI, 0.27-0.29), and system B demonstrated an interobserver reliability of κ = 0.37 (0.36-0.38). In total, 46 images were noted to be suboptimal by the evaluating surgeons. When these were removed, interobserver reliability of system A improved to κ = 0.29 (0.28-0.30), whereas system B improved to κ = 0.38 (0.36-0.39).
Conclusion:
Interobserver reliability for the pulvinar, when viewed arthroscopically, was fair for a Likert-based rating system and fair for a descriptive-based classification. Understanding and developing these classification systems may allow further investigation to create treatment algorithms for pulvinar pathology. These findings may establish a framework for understanding whether variations in pulvinar pathology correlate with patient symptoms and post–hip arthroscopy outcomes.