Background/Purpose:

Although ultrasound (US) shows great potential for aiding assessment of LS disease activity, its use has been limited because both image acquisition and interpretation are operator dependent. Our multi- disciplinary group (LOCUS, LOcalized scleroderma Clinical and Ultrasound Study group) has worked to develop an US scoring measure (U-DA) to help standardize sonographic interpretation of jLS patients (Pediatr Rheumatol 2010;8:14). The U-DA evaluates echogenicity and vascularity (color Doppler signal) differences in each tissue layer of the lesion compared with the corresponding normal tissue layer. A score of 0 represents no difference n the lesion compared with the normal site. 

Objective: To assess the reliability of an ultrasound scoring measure (U-DA) for jLS.

Methods:

LOCUS conducted a 3-day workshop meeting in 2009 on acquiring and interpreting US scans from jLS patients, which was attended by 12 radiologists and sonographers from 5 pediatric rheumatology centers. The group had developed a preliminary ultrasound scoring measure in 2007, and this preliminary measure was reviewed in conjunction with jLS US images showing the range of sonographic differences that had been observed up to the time of the meeting. This review led to modification of U-DA; definitions for scoring levels were then finalized. A tutorial on U-DA scoring was conducted, followed by individual scoring of two jLS US scans and collective review of scoring of these scans. Eleven attendees then scored a randomly-ordered set of 16 jLS scans, with 10 attendees rescoring the same set in a different random order on a second day. Kendall coefficients of concordance were calculated to determine intra-and inter-rater reliability, and scoring of each U-DA was separately analyzed to evaluate for potential issues. 

Results:

Raters showed moderate to high intra-rater reliability for scoring total echogenicity (Kendall’s coefficient 0.77 to 0.92) and vascularity (Kendall’s coefficient 0.64-0.92), where total refers to sum of scores from each identifiable tissue layer (dermis, hypodermis, deep tissue). A moderate level of inter-rater reliability was found for scoring total echogenicity (Kendall’s coefficient 0.64, 0.56) and vascularity (0.58, 0.57).  A high level of agreement was observed for dermis scoring (for example, majority of raters agreed on dermis vascularity score for 15/16 scans), with a much lower level of agreement observed for deep tissue scoring (majority agreement for 10/16 scans on deep tissue vascularity score). 

Table 1: Intra- and Inter-rater reliability of U-DA Total Echogenicity, Total Vascularity

U-DA Scoring Parameter	Kendall’s Coefficient	P value
Total Echo: Inter-rater, 1st reading	0.64	<0.0001
Total Echo: Inter-rater, 2nd reading	0.56	<0.0001
Total Echo: Intra-rater reliability range	0.77-0.92	0.118 to <0.0001
Total Vasc: Inter-rater, 1st reading	0.58	<0.0001
Total Vasc: Inter-rater, 2nd reading	0.57	<0.0001
Total Vasc: Intra-rater reliability range	0.64-0.92	0.1374 to <0.0001

Conclusion:   The U-DA was found to have a moderate level of intra- and inter-rater reliability for total echogenicity and vascularity. Among the different tissue layers, deep tissue layer showed the lowest concordance. More training in deep tissue layer evaluation may further improve the reliability of scoring the U-DA.

---

« Back to ACR/ARHP Annual Meeting 2012

ACR Meeting Abstracts - https://acrabstracts.org/abstract/reliability-of-an-ultrasound-scoring-measure-for-juvenile-localized-scleroderma-jls/