Background/Purpose: Musculoskeletal ultrasound (MSUS) is increasingly being utilized in children. In order to provide objective assessments of arthritis, reliable scoring systems are needed. Yet, given the unique pediatric anatomy, adult scoring systems cannot simply be adapted for children. The aim of this study was to develop and assess reliability of a B-Mode and Doppler scoring system for arthritis in the pediatric knee.

Methods: As part of the Childhood Arthritis and Rheumatology Research Alliance (CARRA) ultrasound group, six pediatric rheumatologists developed a set of standard views of the knee relevant for the assessment of arthritis in children. This was done through a comprehensive literature review followed by a practical exercise and consensus process. Similarly, a B-mode and Doppler scoring system was developed through both literature review and consensus process and subsequently validated by several practical exercises on images that were obtained based on the previously agreed standard views. Agreement between raters was determined using two-way single score intraclass correlation coefficients (ICC).

Results: A total of 21 possible views/images to assess the presence of arthritis in children were identified. After evaluation of the proposed views during a practical exercise, a total of 3 views were chosen for each B-mode and Doppler in a subsequent consensus process, all at 30 degrees of flexion: a suprapatellar view, a medial and a lateral parapatellar view. An initial scoring exercise included 218 images. ICC for all B-mode images was 0.66 (CI 0.58-0.73). ICC for all Doppler images was 0.53 (CI 0.37-0.67). There was excellent agreement for B-mode on suprapatellar images with ICC 0.78 (CI 0.70-0.85) and fair to good agreement for parapatellar images: medial ICC 0.52 (CI 0.37-0.67), lateral ICC 0.60 (CI 0.46-0.73). The ICC for Doppler was fair at 0.39 (CI 0.22-0.58), 0.57 (CI 0.38-0.75) and 0.54 (CI 0.32-0.73) for suprapatellar, medial and lateral parapatellar views respectively. After revision of the scoring system, 90 images were assessed in a second round. Results improved with the ICC for B-mode in the suprapatellar view now at 0.82 (CI 0.70-0.91, excellent). The ICC for Doppler on the suprapatellar view improved to 0.49 (CI 0.25-0.78 fair/good) and on the lateral parapatellar view to 0.64 (CI 0.37-0.87, good).

Conclusion: A novel B-Mode and Doppler scoring system for assessing arthritis of the pediatric knee was successfully developed through a consensus process. Preliminary results from practical exercises indicate overall good reliability but also illustrate the challenges in certain views especially with regards to Doppler scoring. With further refinement and validation, this scoring system could serve as a clinical tool and outcome measure for scientific purposes.

« Back to 2017 ACR/ARHP Annual Meeting

ACR Meeting Abstracts - https://acrabstracts.org/abstract/a-novel-ultrasound-scoring-system-for-the-pediatric-knee/