BACKGROUND: Assessment of range of motion (ROM) and muscle strength is fundamental in the clinical diagnosis of hip osteoarthritis (OA) but reproducibility of these measurements has mostly involved clinicians from secondary care and has rarely reported agreement parameters. Therefore, the primary objective of the study was to determine the inter-rater reproducibility of ROM and muscle strength measurements. Furthermore, the reliability of the overall assessment of clinical hip OA was evaluated. Reporting is in accordance with proposed guidelines for the reporting of reliability and agreement studies (GRRAS).
METHODS: In a university hospital, four blinded raters independently examined patients with unilateral hip OA; two hospital orthopaedists independently examined 48 (24 men) patients and two primary care chiropractors examined 61 patients (29 men). ROM was measured in degrees (deg.) with a standard two-arm goniometer and muscle strength in Newton (N) using a hand-held dynamometer. Reproducibility is reported as agreement and reliability between paired raters of the same profession. Agreement is reported as limits of agreement (LoA) and reliability is reported with intraclass correlation coefficients (ICC). Reliability of the overall assessment of clinical OA is reported as weighted kappa.
RESULTS: Between orthopaedists, agreement for ROM ranged from LoA [-28-12 deg.] for internal rotation to [-8-13 deg.] for extension. ICC ranged between 0.53 and 0.73, highest for flexion. For muscle strength between orthopaedists, LoA ranged from [-65-47N] for external rotation to [-10 -59N] for flexion. ICC ranged between 0.52 and 0.85, highest for abduction. Between chiropractors, agreement for ROM ranged from LoA [-25-30 deg.] for internal rotation to [-13-21 deg.] for flexion. ICC ranged between 0.14 and 0.79, highest for flexion. For muscle strength between chiropractors, LoA ranged between [-80-20N] for external rotation to [-146-55N] for abduction. ICC ranged between 0.38 and 0.81, highest for flexion. Weighted kappa for the overall assessment of clinical hip OA was 0.52 between orthopaedists and 0.65 between chiropractors.
CONCLUSIONS: Reproducibility of goniometric and dynamometric measurements of ROM and muscle strength in patients with hip OA is poor between experienced orthopaedists and between experienced chiropractors. Orthopaedists and chiropractors can to a moderate degree differentiate between hips with or without osteoarthritis.