WEDNESDAY, Oct. 28, 2026 (HealthDay News) -- Several melanoma risk prediction models show moderate discriminatory performance, supporting their potential use for risk stratification, according to a study published online Oct. 21 in JAMA Dermatology.Gillian Reyes-Marcelino, M.P.H., from The University of Sydney, and colleagues examined the discrimination and calibration of six published melanoma risk prediction models in a population-based study using data from 267,357 residents of New South Wales, Australia, from the Sax Institute's 45 and Up Study cohort with linked cancer data. A total of 45,061 participants completed the 2020 follow-up; the analysis sample included 43,235 participants, with 946 invasive and 1,406 in situ incident melanomas.The researchers observed variation in discrimination assessed using the area under the receiver operating characteristic curve at 10 years, from 0.58 to 0.70. The number of melanomas diagnosed in the cohort was underestimated by most risk models; the two models with the best calibration had calibration of −0.16 and 0.59 (slopes, 0.74 and 0.72, respectively)."This independent evaluation of predictive performance identified several melanoma risk prediction models with moderate discriminatory performance largely comparable to their original studies, highlighting their potential use in risk stratification to guide prevention and early detection strategies," the authors write. "Some recalibration may be needed to improve estimation of absolute risks at the population level."One author is coauthor of one of the risk tools.Abstract/Full Text (subscription or payment may be required).Sign up for our weekly HealthDay newsletter