Abstract
Factors influencing the prognosis of patients with terminal-stage cancer remain poorly understood. In this study, we examined these factors and developed a visual model to predict patient survival. Data were collected from patients with terminal-stage cancer treated at the Air Force Hospital of the People's Liberation Army Eastern Theater Command between 2011 and 2020 were collected. Patients were categorized into the training and validation cohorts. Clinical and laboratory characteristics were collected for analysis and prognostic factors were identified to construct a predictive model, develop a nomogram in the training set (n = 193) and verify it in the validation set (n = 85). Our findings revealed that survival predictions for terminal-stage cancer were not associated with common factors such as tumor type, stage, patient age at diagnosis, or Eastern Cooperative Oncology Group performance status score. Instead, factors such as willingness to receive treatment, dyspnea, serum urea, serum albumin, and neutrophil count proved to be critical. These factors were used to create a highly accurate and reliable nomogram. A comprehensive analysis of prognostic factors in patients with terminal-stage cancer resulted in the development of a practical nomogram model for clinical application.