RESUMO
OBJECTIVE: To determine the validity of a hardness sensor to objectively assess skin induration in patients with systemic sclerosis, and to compare the hardness sensor with the modified Rodnan skin score (MRSS) and a durometer. METHODS: The skin induration was measured in two assessments: a Latin square experiment to examine the hardness sensor's intraobserver and interobserver reliability; and a longitudinal cohort to evaluate the distribution of hardness sensor measurements, the correlation between hardness sensor, durometer and MRSS, and the sensitivity to change in skin hardness. Other outcome data collected included the health assessment questionnaire (HAQ) disability index and Keitel function test (KTF) score. RESULTS: The reliability of the hardness sensor was excellent, with high intraobserver and interobserver intraclass correlation coefficients (0.97; 0.96), which was higher than MRSS (0.86; 0.74). Interobserver reproducibility of hardness sensor was only poor in abdomen (0.38), yet for durometer it was poor in face (0.11) and abdomen (0.33). The hardness sensor score provided a greater dynamic evaluation range than MRSS. Total hardness sensor score correlated well with MRSS (r=0.90, p<0.001), total durometer score (r=0.95, p<0.001), HAQ disability index (r=0.70, p<0.001) and KTF score (r=0.66, p<0.001). Change in hardness sensor score also correlated with change in MRSS (r=0.78, p<0.001), total durometer score (r=0.85, p<0.001), HAQ disability index (r=0.76, p<0.001) and KTF score (r=0.67, p<0.001). CONCLUSION: The hardness sensor showed greater reproducibility and accuracy than MRSS, and more application sites than durometer; it can also reflect patients' self-assessments and function test outcomes.