Exploring reliability of scar rating scales using photographs of burns from children aged up to 15 years.

Assessing burn scars from photographs is a common practice given the growing trend to support health service delivery via electronic media (eg, email, videoconferencing). Scar rating scales, originally designed for in-person assessment, have been used to rate scars from photographic images. Evidence for the reliability of this practice is lacking. Five raters completed three scar rating scales (Patient and Observer Scar Scale, Manchester Scar Scale, modified Vancouver Scar Scale), both in-person and using photographs on 12 participants (seven male, five female) with 18 scar areas (3 × 3 cm). Interrater reliability for the scar parameters of vascularity, color, contour, pliability, and overall opinion achieved intraclass correlation coefficient values of between 0.71 and 0.87 (in-person) and 0.72 and 0.77 (using photographs) for multiple raters. The level of agreement between in-person and photographic assessment was below acceptable levels, which brings into question construct validity when scar rating scales are used in a way for which they were not designed. Reliability estimates in this study were likely reduced by the underrepresentation of scars in the more severe range. This limitation needs to be addressed in future research. Advances are required in the development and refinement of burn scar rating scales, specifically for photographic use, given their routine use in clinical care.