Evidence-Based Dermatology
April 2014

Risk Prediction Models for Incident Primary Cutaneous MelanomaA Systematic Review

Author Affiliations
  • 1Cancer Epidemiology and Services Research, Sydney School of Public Health, The University of Sydney, Sydney, New South Wales, Australia
  • 2Sydney School of Public Health, The University of Sydney, Sydney, New South Wales, Australia

Copyright 2014 American Medical Association. All Rights Reserved. Applicable FARS/DFARS Restrictions Apply to Government Use.

JAMA Dermatol. 2014;150(4):434-444. doi:10.1001/jamadermatol.2013.8890

Importance  Currently, there is no comprehensive assessment of melanoma risk prediction models.

Objective  To systematically review published studies reporting multivariable risk prediction models for incident primary cutaneous melanoma for adults.

Evidence Review  EMBASE, MEDLINE, PREMEDLINE, and Cochrane databases were searched to April 30, 2013. Eligible studies were hand searched and citation tracked. Two independent reviewers extracted information.

Findings  Nineteen studies reporting 28 melanoma prediction models were included. The number of predictors in the final models ranged from 2 to 13; the most common were nevi, skin type, freckle density, age, hair color, and sunburn history. There was limited reporting and substantial variation among the studies in model development and performance. Discrimination (the ability of the model to differentiate between patients with and without melanoma) was reported in 9 studies and ranged from fair to very good (area under the receiver operating characteristic curve, 0.62-0.86). Few studies assessed internal or external validity of the models or their use in clinical and public health practice. Of the published melanoma risk prediction models, the risk prediction tool developed by Fears and colleagues, which was designed for the US population, appears to be the most clinically useful and may also assist in identifying high-risk groups for melanoma prevention strategies.

Conclusions and Relevance  Few melanoma risk prediction models have been comprehensively developed and assessed. More external validation and prospective evaluation will help translate melanoma risk prediction models into useful tools for clinical and public health practice.