Age-related vascularization and ossification of joints in children: an international pilot study to test multi-observer ultrasound reliability

OMERACT paediatric ultrasound subtask force

Research output: Contribution to journalArticle

Abstract

OBJECTIVE: To determine the intra- and inter-observer reliability of ultrasound (US)-detected age-related joint vascularization and ossification grading in healthy children.

METHODS: Following standardized image acquisition and machine setting protocols, 10 international US experts examined four joints (wrist, second metacarpophalangeal joint, knee, and ankle) in 12 healthy children (divided into four age groups: 2-4, 5-8, 9-12, and 13-16 years). Grey-scale was used to detect the ossification grade, and power Doppler (PD) US was used to detect physiological vascularization. Ossification was graded from grade 0 (no ossification) to grade 3 (complete ossification). A positive PD signal was defined as any PD signal inside the joint. Kappa statistics were applied for intra- and inter-observer reliability.

RESULTS: According to the specific joint and age, up to four solitary PD signals (mean, 1.5) were detected within each joint area with predominant localization of the physiological vascularization in specific anatomic positions: fat pad, epiphysis, physis, and short bone cartilage. The kappa values for ossification grading were 0.87 (range, 0.85-0.91) and 0.58 for intra- and inter-observer reliability, respectively. The bias-adjusted kappa values for intra- and inter-observer reliability were 0.71 (range, 0.44-1.00) and 0.69, respectively.

CONCLUSION: Detection of normal findings (i.e., grading of physiological ossification during skeletal maturation and identification of physiological vessels) can be highly reliable by using clear definitions and a standardized acquisition protocol. These data will permit development of a reliable and standardized US approach for evaluating paediatric joint pathologies. This article is protected by copyright. All rights reserved.

    Fingerprint

Cite this