Online prediction tools for melanoma survival: A comparison.

Abstract

Breslow thickness, patient age and ulceration are the three most valuable clinical and pathological predictors of melanoma survival. A readily available reliable online tool that accurately considers these and other predictors could be valuable for clinicians managing melanoma patients. To compare online melanoma survival prediction tools that request user input of clinical and pathological features. Search engines were used to identify available predictive nomograms. For each, clinical and pathological predictors were compared. Three tools were identified. The American Joint Committee on Cancer tool inappropriately rated thin tumors as higher risk than intermediate tumors. The University of Louisville tool was found to have six shortcomings: a requirement for sentinel node biopsy, unavailable input of thin melanoma or patients over 70 years of age and less reliable hazard ratio calculations for age, ulceration, and tumor thickness. The LifeMath.net tool was found to appropriately consider tumor thickness, ulceration, age, sex, site and tumor subtype in predicting survival. The authors did not have access to the base data used to compile various prediction tools. The LifeMath.net prediction tool is the most reliable for clinicians in counselling patients with newly diagnosed primary cutaneous melanoma regarding their survival prospects.