Interobserver reliability of the Kellgren-Lawrence classification of degenerative knee osteoarthritis among resident physicians from the University of the Philippines - Philippine General Hospital

ABSTRACT

@#<p style="text-align justify;"><strong>Objectives:</strong> To determine the interobserver reliability of the Kellgren-Lawrence classification among selected residents from departments forming the University of the Philippines - Philippine General Hospital (UP-PGH) Osteoarthritis Multidisciplinary Clinic (OAMDC).</p><p style="text-align justify;"><strong>Methods:</strong> From each department, 3 resident physicians (n = 9) were randomly chosen and tasked to categorize 20 knee anteroposterior and lateral radiographs into KLC grades. Inter-observer reliability was assessed using Fleiss's kappa coefficient (?).</p><p style="text-align justify;"><strong>Results:</strong> Results show that there was 31.90% (? = 0.3190 ± 0.0228, p-value < 0.05) agreement beyond chance in KLC grading of the radiograph series among all participating residents. Sub-group analyses of interobserver reliability in terms of departmental affiliation noted a range of agreement beyond chance in KLC grading, from 10.52% (? = 0.1052 ± 0.0779, p-value < 0.05) to 56.38% (? = 0.5638 ± 0.0844, p-value < 0.05).</p><p style="text-align justify;"><strong>Conclusion:</strong> The findings reveal significant variability of agreement beyond chance in KLC grading, both within and among residents of UP-PGH OAMDC-participating clinical departments, which may reflect differences in training or competency and/or the documented limitations of the KLC system. Further investigation to improve diagnostic and severity assessment accuracy and uniformity in the institution is therefore warranted.</p>