Development and validation of prognostic nomogram in pediatric melanoma: a population-based study

Abstract

The aim of the study is to use the Surveillance, Epidemiology, and End Results (SEER) database to develop a useful clinical nomogram that uses prognosis prediction for pediatric melanoma patients. We obtained clinical information on pediatric melanoma patients from the SEER database between 2000 and 2018. Each patient was split into a training cohort or a validation cohort at random. Results between various subgroups were compared using Kaplan–Meier analyses. We created a nomogram to calculate the probability of survival for pediatric patients with melanoma. The performance of nomograms was assessed using calibration and discrimination. To assess the clinical use of this newly created model, decision curve analysis was also performed. In this study, a total of 890 eligible patients were chosen at random and allocated to 70% of training cohorts (n = 623) and 30% of validation cohorts (n = 267). After applying the chosen various components to create a nomogram, validated indexes showed that the nomogram had a strong capacity for discrimination. The training set’s and validation set’s C-index values were 0.817 and 0.832, respectively. The calibration plots demonstrated a strong correlation between the observation and the forecast. The model has a good clinical net benefit for pediatric melanoma patients, according to the clinical decision curve. In conclusion, we created an effective survival prediction model for pediatric melanoma. This nomogram is accurate and useful for clinical decision-making. Still, more external confirmation is required.