Predicting cancer-specific mortality in patients with parotid gland carcinoma by competing risk nomogram.

Abstract

The aim of the present study was to establish a competing risk nomogram to predict parotid gland cancer-specific mortality (PGC-SM). Seven thousand nine hundred and sixty-two patients extracted from SEER database were randomly categorized into training and validation sets. The competing risk model was used to identify factors associated with PGC-SM. The nomogram was evaluated via concordance indexes (C-indexes), calibration plots, and decision curve analysis (DCA). Male, elderly, white, widowed, larger tumor, no surgery, advanced tumor grade, lymph node (LN) metastasis, adenocarcinoma (ADC), and higher TNM stage were associated with higher incidence of PGC-SM. Calibration plots showed that the nomogram was well calibrated. C-indexes for nomogram were 0.84 (95% CI: 0.81-0.86) and 0.84 (95% CI: 0.82-0.86) in training and validation sets, respectively. DCA demonstrated the clinical usefulness of nomogram. The competing risk nomogram shows high performance in predicting PGC-SM, which might enable clinicians formulate suitable treatment protocols for patients with parotid gland carcinoma (PGC).