Feasibility and perceptions among hemophilia patients and doctors towards ultrasound for the assessment of subclinical abnormalities in hemophilic joints

This study explores the feasibility and perceptions of using ultrasound (US) for identifying subclinical abnormalities in hemophilic joints from the perspectives of both patients and medical practitioners. This was a single-center pilot follow-up study that included multidisciplinary medical professionals and severe hemophilia patients. The Hemophilia Early Arthropathy Detection with Ultrasound in China (HEAD-US-C) scale and Hemophilia Joint Health Score (HJHS) 2.1 were utilized to assess joint conditions. Inter-rater agreement for HEAD-US-C scoring on stored US images was evaluated. Data from 13 multidisciplinary doctors and 28 patients (112 joints: 40 elbows, 24 knees, 48 ankles) were analyzed. Among 40 joints without reported bleeding during follow-up, US total scores increased in 25.0% (10/40), whereas HJHS scores increased in 12.5% (5/40). Among the 78 joints with unchanged HJHS scores, US detected increased structural scores in 21.8% (17/78) of joints. Conversely, among the 66 joints with unchanged US scores, only 9.1% (6/66) exhibited HJHS score changes >0. At the joint level, HEAD-US-C scores were positively correlated with HJHS scores (repeated measures correlation coefficient = 0.757, 95% CI 0.648-0.835, P < .001). US was perceived as useful by 69.2% of doctors (38.5% very useful, 30.8% somewhat useful), and US-based visual explanations were perceived as useful by all patients. The overall linear weighted kappa for patient-level aggregated HEAD-US-C scores assigned by the 2 sonographers was 0.903 (95% CI 0.871-0.935). US assessment using the HEAD-US-C scale appeared feasible and was well accepted by doctors and patients. In this pilot study, US identified increased structural scores in some joints without corresponding HJHS changes, suggesting that it may provide complementary structural information to clinical joint assessment. The findings also showed a positive association between HEAD-US-C scores and HJHS scores and high inter-rater agreement for HEAD-US-C scoring on stored US images. However, this study was not designed to demonstrate superiority over HJHS, diagnostic performance, improved medication adherence, or complete reliability of the HEAD-US-C scoring system.