Self-assessment of scleroderma skin thickness: Development and validation of the pastul questionnaire

Spierings, J.; Ong, V.; Denton, C.

Background: Evaluation of skin is central to both clinical practice and trials in systemic sclerosis (SSc). This is generally done with the modified Rodnan Skin Score (mRSS). Remote consultations are now widely implemented in response to the COVID-19 pandemic, which has inevitably limited evaluation of skin. To monitor skin during this pandemic and to further explore ways to assess skin, we developed the PASTUL (Patient self-Assessment of Skin Thickness in Upper Limb) questionnaire. Objectives: This study evaluated feasibility and validity of PASTUL in SSc. Methods: The PASTUL questionnaire uses a simple self-assessed grading of skin as normal, mild, moderate, or severely thickened at eight sites of upper limb corresponding to mRSS. Assessed grades were converted to an integer scale [0, 1, 2, 3]. Detailed instructions for patients were provided. Scleroderma Skin PRO (SSPRO) and Scleroderma Health Assessment Disability Index (SHAQ-DI) were also completed. For comparison, physician assessed mRSS was performed in a subgroup of patients. Construct validity was evaluated by examining the correlation between PASTUL, mRSS, SSPRO and SHAQ-DI using Pearson's correlation coefficient. Content validity was evaluated by scoring relevance, clarity and practical difficulty. Test-retest reliability was estimated using intraclass correlation coefficient (ICC). Results: 130 patients were invited of which 104 (80%) completed the questionnaires. The mRSS was undertaken in 78 patients (n=42, 54% limited cutaneous SSc (lcSSc)). The PASTUL was completed by patients (86%) or by a partner/ friend (14%). Mean PASTUL score was 11 (SD 7), mean HAQ-DI 1.41 (SD 0.77) and mean SSPRO 48.3 (SD 27.0). PASTUL strongly correlated with total SSPRO and SSPRO subdomain physical limitations (r=0.60 and 0.62, respectively) (Figure 1A). Correlations between PASTUL and mRSS and mRSS upper limbs were moderate (r=0.56 and 0.58, respectively) (Figure 1B). An overview of all correlations is provided in Table 1. Correlation between PASTUL and mRSS was stronger in lcSSc compared to diffuse cutaneous SSc patients (r=0.53 vs 0.43) and when assessed by a partner/friend compared to patients themselves (r=0.90 vs 0.54). The PASTUL demonstrated excellent test-retest reliability (ICC of 0.93, p<0.001) and good content validity. Conclusion: The significant correlation of PASTUL scores with total SSPRO and physical limitation scores and moderate correlation with mRSS support the potential of PASTUL for remote evaluation of skin thickness in virtual clinical settings. Future studies may explore sensitivity to change and utility in clinical trials.