Background/Purpose:

LS commonly causes severe morbidity for the growing child. Optimal therapy is not known and the lack of an agreed-upon standard for assessing disease state and monitoring treatment response has hindered treatment comparison. To work towards conducting comparative effectiveness studies, an LS-focused Childhood Arthritis and Rheumatology Research Alliance subgroup developed a clinical disease activity measure (LS Activity Score, Arthritis Care Res 2012 DOI: 10.1002/acr.21687), based upon LoSSI and LOCUS studies (J Rheumatol 2008;35:640; Arthritis Rheum 2011;63 Suppl:S955).  The LS Activity Score scores for 7 parameters (erythema, violaceous color, waxy or white lesion, skin thickness of lesion edge, lesion warmth, new lesion, change in lesion size) in affected anatomic sites, with body divided into 19 sites.

Objective: To evaluate the validity and reliability of the LS Activity Score. 

Methods:  

A two-day workshop meeting was conducted in which 13 pediatric rheumatologists and a pediatric dermatologist reviewed LS clinical measures and conducted a reliability of scoring study.  The 14 raters evaluated 13 juvenile LS (jLS) patient volunteers with the LS Activity Score in a random order twice in one day. For each patient, raters were told which anatomic sites to assess (1-2/patient); new lesion and lesion size change were not scored because they require prior patient evaluation. Raters scored Physician Global Assessment (PGA, 0-100 mm) of disease activity (DA) and PGA of disease damage (DD) for each patient based upon evaluated anatomic sites. We hypothesized the LS Activity Score would have moderate to high correlation with PGA-DA (convergent validity) and low correlation with PGA-DD (divergent validity). Spearman’s rho was calculated to assess level of correlation, and intraclass correlation coefficients (ICC) calculated to assess intra- and inter-rater reliability (0.20-0.39 considered low, 0.40-0.59 moderate, >0.60 high).

Results:

Mean age of jLS patients was 13.2 years, most common subtype was linear scleroderma (4 limb, 4 head). Mean patient LS Activity Scores ranged from 0.50 to 9.28 (4-52% of maximum possible score), mean PGA-DA from 4.0 to 59.6. There was a high correlation between LS Activity score and PGA-DA (ρ= 0.94), and low with PGA-DD (ρ= -0.121).  Raters showed moderate inter-rater reliability for LS Activity Score and PGA-DA, with nearly all showing moderate to high intra-rater reliability for these scores (Table). Among the different parameters, violaceous color and lesion warmth had the lowest inter-rater reliability.

Table: Inter rater and intra-rater reliability of LS Activity score

Domains	Round 1 ICC (95% CI)	Round 2 ICC (95% CI)
LS Activity Score:
Inter-rater reliability	0.564 (.38, .79)	0.628 (.43, .83)
Intra-rater reliability (median)		0.804 [range 0.464-0.910]
PGA-DA:
Inter-rater reliability	0.566 (.38, .79)	0.516 (.33, .76)
Intra-rater reliability (median)		0.734 [range 0.369-0.955]
LS Activity Score parameters: inter-rater reliability
Erythema	0.392 (.25, .59)	0.483 (.33, .67)
Violaceous color	0.187 (.09, .36)	0.140 (.06, .29)
Waxy or white lesion	0.429 (.29, .63)	0.518 (.37, .70)
Lesion warmth	0.197 (.10, .37)	0.148 (.07, .30)
Skin Thickness of lesion edge	0.407 (.26, .60)	0.307 (.18, .50)

Conclusion:    We demonstrate initial construct validity for LS Activity score, and found moderate inter-rater reliability. Additional studies are needed to fully evaluate this measure’s validity. More training for the violaceous color and lesion warmth parameters may improve the measure’s reliability.

---

« Back to 2012 ACR/ARHP Annual Meeting

ACR Meeting Abstracts - https://acrabstracts.org/abstract/initial-evaluation-of-a-localized-scleroderma-ls-clinical-activity-measure/