This article is only available in the PDF format. Download the PDF to view the article, as well as its associated figures and tables.
We fully acknowledge that the chance-corrected intergrader agreement coefficient (ie, generalized κ) of .31 for the photonumeric scale does not demonstrate a high level of agreement among graders when assessing varying degrees of photodamaged skin. However, what we have shown is that a grading scale anchored by representative photographs can significantly improve the level of intergrader agreement achieved with a conventional scale. Whether or not this is an "acceptable" level of agreement is not the issue. An improvement is still an improvement, regardless of the size of the increment; until a superior method is devised, we feel the photonumeric scale represents a preferable alternative.Regarding intraobserver agreement (repeatability), the methods employed to test and summarize our data were appropriate and not unduly influenced by limited power or nonnormality.The speculation that the level of intergrader agreement from a photonumeric scale would be especially unsuited when used by less experienced workers is
Griffiths CEM, Hamilton TA. A Photonumeric Scale for the Assessment of Cutaneous Photodamage-Reply. Arch Dermatol. 1992;128(10):1407. doi:10.1001/archderm.1992.01680200119027