31 Evaluation of Intra- and Inter-User Reliability in Quantitative Scar Assessments

Abstract Introduction To adequately evaluate the efficacy of current and emerging anti-scar technologies, scar assessment must be carried out in a systematic, objective manner using non-invasive instruments with low potential for user bias. In addition, for these assessments to be compatible with normal clinical workflow, they must be robust, user independent and rapid. Methods In this IRB-approved study, scar height, texture, color and biomechanics were evaluated using non-invasive, quantitative instruments. One scar site per subject (n = 15) was marked for analysis and assessed, in triplicate, by three independent investigators to evaluate inter- and intra-user variability. Scar color was assessed using digital image analysis, commercially available spectroscopy equipment for skin, the Vancouver Scar Scale (VSS) and the Patient and Observer Scar Assessment Scale (POSAS). Biomechanical analysis was performed using three commercially available non-invasive instruments along with VSS and POSAS. Scar height and texture were assessed using a 3D scanner, conventional molding/casting combined with digital image analysis along with VSS and POSAS. Intraclass correlation coefficients (ICC) were calculated to assess intra and inter-user reliability with the quantitative instruments and kappa reliability statistics were performed to assess inter-user reliability with VSS/POSAS. Results Inter-user evaluation of scar color was significantly more reliable with spectroscopy equipment vs. digital photograph analysis (0.9798 and 0.6148, respectively) and more reliable than POSAS and VSS (0.2063 and 0.5994, respectively). All evaluations with VSS and POSAS had fair to moderate inter-user reliability. Evaluation of scar height/texture had greater intra-user reliability with the molding technique (0.8191) vs. the 3D scanner (0.6098); however, both had poor inter-user reliability. Biomechanical analyses using quantitative instruments had poor to moderate inter-user reliability (0.50–0.75) based on the type of instrument and property quantified. Intra-user reliability was significantly better (moderate to good) and was dependent on investigator experience. Conclusions Quantitative analyses of color can be more reliably assessed using a commercially available instrument versus digital image analysis or scar scales. Evaluation of scar biomechanics and height/texture are more heavily dependent on the assessor and require significant levels of training to achieve acceptable levels