Background/Purpose: The clinical decision-making process in pediatric arthritis lacks an objective, reliable bedside imaging tool. Although increasingly used for children, there is a need for reliable ultrasound (US) scanning protocols and scoring systems to provide objective assessments of various inflammatory lesions. The aim of this study was to develop a scanning protocol and assess the reliability of B-Mode and Doppler scoring systems for the pediatric ankle.

Methods: As part of the Childhood Arthritis and Rheumatology Research Alliance (CARRA) US group, 19 pediatric rheumatologists developed through a comprehensive literature review a set of standard views to assess inflammatory lesions of the synovial recesses as well as tendons of the pediatric ankle. The views included in the final acquisition protocol reached at least 80% agreement following a nominal group technique. Subsequently, a B-mode and Doppler scoring system was developed through a comprehensive literature review, then a consensus process. Three rounds of scoring of still-images were followed by one practical exercise. Agreement between raters determined using two-way single score intraclass correlation coefficients (ICC).

Results: Following published standardized guidelines, we examined the available evidence for using sonographic views to assess presence of synovitis in the tibiotalar (TTJ), subtalar (STJ) and talonavicular joints (TNJ), as well as tenosynovitis of the anterior, medial and lateral tendon compartments of the ankle. We identified 37 possible views to assess the presence of ankle synovitis and tenosynovitis in children.  We further considered different positions of the ankle (extension, neutral and flexion).  Following a practical exercise, a total of 9 views were chosen for each B-mode and Doppler in a nominal group technique (Table 1).  The scanning protocol was feasible with a scanning time of 12-15 minutes. Scoring protocols for each structure based upon anatomic variability among the joints and tendons of interest were developed. An initial scoring exercise included 124 images.  There was good to excellent agreement for B-mode on the anterior TTJ: medial and lateral views ICC 0.69 (CI 0.61-0.75) and midline view ICC 0.75 (CI 0.71-0.79).  The ICC for the TNJ was fair at 0.51 (CI 0.40-0.61). Views assessing the medial and lateral aspects of the STJ had good to fair reliability 0.67 (CI 0.57-0.75) and 0.51 (CI 0.43-0.59) respectively. ICC for tendon images was 0.49 (CI 0.37-0.59). The observations from this scoring exercise are elucidating which modifications are necessary to reach good to excellent reliability of the scoring systems. Additional scoring exercises and consensus rounds will be necessary to refine the individual scores.

Conclusion: A novel pediatric ankle US scanning protocol and scoring system for the assessment of arthritis were successfully developed. Preliminary results from practical exercises indicate overall good reliability but also illustrate the challenges in certain views especially the STJ, TNJ and tendons.  With further refinement and validation, this scoring system may serve as an assessment tool and outcome measure for both clinical and research applications

BM: B-mode, PD: Power Doppler, TTJ: Tibiotalar Joint, STJ: subtalar joint. Note: It is recommended to slide the probe medially to laterally and capture the area of maximal distention.

« Back to ACR Convergence 2020

ACR Meeting Abstracts - https://acrabstracts.org/abstract/novel-ultrasound-image-acquisition-protocol-and-scoring-system-for-the-pediatric-ankle/