STM Article Repository

Background: Eye colour can be a predictive factor for skin cancer. Several iris colour classification scales have been described, but are not widely accepted. Further investigation of the relationship between the iris and skin cancer requires a valid and reliable instrument for iris colour classification.
Objective: Development and validation of an easy and clinically reliable scale for iris colour classification.
Design: This is a descriptive case series study.
Participants: We analysed 211 irises digital photographs of patients with primary skin cancer taken with identical digital camera settings.
Methods: Three researchers independently rated the irises using a new scale based on three descriptive parameters (freckles, collarette, and periphery) and with two validated scales for comparison. For each parameter, we calculated the inter-ratter concordance (percentage of agreement; Cohen’s kappa), the validity (Spearman bivariate correlation coefficients) and the internal consistency (Cronbach’s alpha).
Results: The three ratters had almost perfect concordance for all three scales and for each parameter of the novel scale. Internal consistency analysis revealed an acceptable Cronbach’s alpha coefficient of 0.734 for the new scale total score.
Conclusions: The novel scale is a reliable and valid instrument for iris colour classification. It is easy to apply, interpret, and specifically relate to skin cancer risk.