Prognostic modeling of overall survival and analysis of K-M survival curves in patients with primary colon cancer: A SEER-based study.

Abstract

This study aimed to establish a validated prognostic survival column line chart by analyzing data from patients with colon cancer (CC) in the SEER database. The nomogram proposed in this study was based on the retrospective data of patients diagnosed with CC in the SEER database from 1975 to 2015. Randomly divided into training and validation sets, the nomogram was constructed using the Cox model, and the discriminatory power of the nomogram and its predictive accuracy were determined using the consistency index and associated calibration curves. In a multifactorial analysis of the main cohort, the independent factors for survival were age, sex, race, tumor stage, and tumor grade, all of which were included in the nomogram and were prognostic factors for patients with CC (P < .05). The calibration curve of the survival probability showed good agreement between the prediction of the nomogram and the actual observation. The validation calibration curve showed good correlation and agreement between predicted and observed values. Multifactorial analysis showed that the factors affecting the prognosis of patients with CC included age, sex, race, tumor-node-metastasis stage, and tumor pathological stage. The nomogram prediction model proposed in this study has high accuracy and can provide more accurate prognostic prediction and relevant reference values for assessing the postoperative survival of CC patients and guiding clinical decision-making.