Nomograms predicting the overall survival and cancer-specific survival of patients with stage IIIC1 cervical cancer

Abstract Background To explore the factors that affect prognosis of overall survival (OS) and cancer-specific (CSS) patients with stage IIIC1 cervical cancer establish nomogram models to predict this prognosis. Methods Data from in Surveil-lance, Epidemiology, End Results (SEER) programme meeting inclusion criteria were classified into a training group, validation data obtained First Affiliated Hospital Anhui Medical University 2010 2019. The incidence, Kaplan-Meier curves, OS CSS group evaluated. Nomograms established according results univariate multivariate Cox regression models. Harrell’s C-index, calibration plots, receiver operating characteristic (ROC) curves decision-curve analysis (DCA) calculated validate prediction incidence pelvic lymph node metastasis, high-risk factor for cancer, decreased slightly over time. Eight independent prognostic variables identified OS, including age, race, marriage status, histology, extension range, tumour size, radiotherapy surgery, but only seven CSS, status excluded. based on results. C-indexes nomograms 0.687 0.692, respectively, using random sampling SEER sets 0.701 0.735, external sets. AUCs 0.708 0.705 0.750 sets, respectively. In addition, 0.707 0.709 when validated 0.788 0.785 Calibration plots almost identical actual observations. DCA also indicated value two Conclusions OS. same predicted exception status. Both had good predictive clinical application after validation. Notably, size largest contribution nomograms.