A Pilot Study to Assess the Reliability of Digital Image-Based PASI Scores Across Patient Skin Tones and Provider Training Levels

Abstract Introduction The ability to perform psoriasis skin assessments remotely through digital image-based psoriasis area and severity index (DIB-PASI) would be a valuable tool for psoriasis clinical trials. An ideal teledermatological assessment would be robust across patients of diverse skin tones as well as across assessors of varying experience levels. In this pilot study, we evaluated the reliability of face-to-face (FTF) versus DIB-PASI scores determined by trained clinical assessors with a spectrum of experience and with patients of different skin tones. Methods Fourteen subjects of varying skin tones with moderate-to-severe plaque psoriasis were treated with adalimumab. In-person PASI assessments and digital photography were performed in the clinic at weeks 0, 12, and 24. Photographs were reviewed by four independent assessors to derive a digital image-based PASI score. The concordance of face-to-face PASI (FTF-PASI) and DIB-PASI were analyzed across patient and assessor factors. Results Overall concordance between FTF-PASI and DIB-PASI was high (ICC 0.82, p < 0.0001), with good agreement across individual assessors. When analyzed by PASI score component or body region, digital assessors also demonstrated good agreement with the FTF assessor. Similarly, DIB-PASI showed high concordance with FTF-PASI for patients with light skin tones and patients with medium-to-dark skin tones, and across clinical training levels. Conclusion Overall, PASI scores derived from digital images showed good agreement with those determined in person. Importantly, these remote assessments were reliable for both light and medium-to-dark skin tones, and robust to training level of the assessor. The findings from this pilot study lay the foundation for expanding teledermatology-based clinical trials for patients with psoriasis and enabling accurate, remote monitoring of disease severity and therapy response.http://deepblue.lib.umich.edu/bitstream/2027.42/173939/1/13555_2022_Article_750.pd