Inter-observer and Intra-observer Reliability of Clinical Assessment of Knee Osteoarthritis

Background: Due to inconsistency in literature, we investigated both inter- and intra-observer reliability of clinical tests used in the assessment of osteoarthritis (OA) of the knee. Methods: Following a period of standardisation of the clinical test procedures, using symptomatic knee OA participants, inter-observer reliability between two clinicians at the same clinical visit, and intra-observer reliability of assessments 2 to 8 weeks apart was determined using 25 (mean age 63 years) and 71 (mean age 64 years) participants respectively. Clinical assessment included assessment of bony enlargement (yes/no), joint crepitus (yes/no), muscle atrophy (yes/no), effusion (5-point scale), joint tenderness (yes/no) and knee range of movement (ROM) (degrees). Intra-class correlation coefficients (ICC), kappa agreement (К) and weighted kappa (Кω) using SPSS software version 22 and Bland and Altman plots were used to determine inter- and intra-observer levels of agreement. Results: Inter-observer agreement was moderate for patellar tap (К=0.52), patellofemoral (К=0.53) and anserine tenderness (К=0.48); good for bony enlargement (К=0.61), quadriceps atrophy (К=0.74), knee joint crepitus (К=0.64) and bulge sign (Кω=0.78) and excellent for tibiofemoral tenderness (К=0.85), knee flexion (ICC=0.97) and extension (ICC=0.88) ROM. Intra-observer agreement was good for patellar tap (К=0.73), knee joint crepitus (К=0.73), bulge sign (Кω =0.83), patellofemoral (К=0.66), tibiofemoral joint (К=0.61) and anserine tenderness (К=0.73) and excellent for bony enlargement (К=0.98), quadriceps atrophy (К=0.85) and knee flexion (ICC=0.995) and extension (ICC=0.96) ROM. Conclusion: Among men and women with symptomatic knee OA, for the majority of clinical signs of knee OA, the reliability of clinical examination of the knee was good.