Development and Reliability of the Joint Tissue Activity and Damage Examination for Quantitation of Structural Abnormalities by Musculoskeletal Ultrasound in Hemophilic Joints

Abstract

Objectives: Musculoskeletal ultrasound (US) is used increasingly to examine hemophilic arthropathy. However, quantitative algorithms to document findings are lacking. We developed and sought to validate a protocol quantifying hemophilic joint abnormalities.

Methods: Thirty-one patients with hemophilia were examined serially for 2 years with musculoskeletal US (≈600 joint examinations and ≈6000 images). Based on the spectrum of pathologies, a quantitative algorithm, named Joint Tissue Activity and Damage Examination (JADE), was developed for soft tissue and osteochondral measurements, including power Doppler, using nominal group techniques. To study intra- and inter-rater reliability, 8 musculoskeletal US-experienced hemophilia providers performed anatomic landmark recognition and tissue measurements on 86 images with arthropathic changes, with repetition 1 month later. Twenty-three musculoskeletal US-inexperienced providers performed similar assessments. Inter-operator reliability was established by 6 musculoskeletal US-experienced hemophilia providers, each acquiring images and JADE assessments of 3 hemophilic arthropathic joints. A radiologist and musculoskeletal sonographer functioned as adjudicators. The statistical analysis was performed with the intraclass correlation coefficient (ICC), Fleiss κ, and Cohen κ where appropriate.

Results: The musculoskeletal US-experienced providers showed excellent intra-and inter-rater reliability for tissue measurements (ICCs, 0.94-0.96). Agreement was good to excellent for landmark recognition (Fleiss κ, 0.87-0.94). Inter-operator reliability was excellent for measurements and landmark recognition (ICC, 0.90; Fleiss κ, 1.0). Agreement with adjudicators was mostly good to excellent. Musculoskeletal US-inexperienced providers showed excellent inter-rater reliability for measurements (ICC, 0.96) and moderate agreement for landmark recognition (Fleiss κ, 0.58).

Conclusions: The JADE protocol appears feasible for quantifying hemophilic intra-articular abnormalities. Musculoskeletal US-trained hemophilia providers showed high intra-rater, inter-rater, and inter-operator reliability, supporting JADE as a protocol for clinical management and research.

Keywords: arthropathy; hemophilia; inter-operator reliability; inter-rater reliability; musculoskeletal ultrasound; reliability.