Development and validation of a melanoma risk score based on pooled data from 16 case-control studies

Area of Science:

• Dermatology
• Epidemiology
• Biostatistics

Background:

• Cutaneous melanoma risk assessment is crucial for public health.
• A self-assessment tool requires a robust and validated risk algorithm.

Purpose of the Study:

• To develop and validate a melanoma risk algorithm for public self-assessment.
• To identify key factors contributing to melanoma risk.

Main Methods:

• A logistic random coefficients model was applied to pooled data from 16 case-control studies.
• Risk categories were defined using predicted melanoma odds and validated through cross-validation.
• Performance was assessed using an independent UK case-control dataset.

Main Results:

• The algorithm demonstrated robust threshold estimates confirmed by cross-validation.
• The model effectively discriminated between melanoma cases and controls (AUC = 0.75).
• Significant differences in risk group distribution were observed between cases and controls.

Conclusions:

• A composite score for relative melanoma risk has been developed and validated.
• This validated score can serve as a valuable tool for public melanoma risk awareness.