1. Lifestyle, clinical and histological indices-based prediction models for survival in cancer patients: a city-wide prospective cohort study in China.
- Author
-
Sun, Ce, Xu, Huan, Wang, Suixiang, Li, Ke, Qin, Pengzhe, Liang, Boheng, and Xu, Lin
- Subjects
- *
PREDICTION models , *LONGITUDINAL method , *RECEIVER operating characteristic curves , *SLEEP duration , *CANCER patients - Abstract
Purpose: We developed a nomogram to predict 3-year, 5-year and 7-year cancer survival rates of cancer patients. Methods: This prospective cohort study included 20,491 surviving patients first diagnosed with cancer in Guangzhou from 2010 to 2019. They were divided into a training and a validation group. Lifestyle, clinical and histological parameters (LCH) were included in multivariable Cox regression. Akaike information criterion was used to select prediction factors for the nomogram. The discrimination and calibration of models were assessed by concordance index (C-index), area under time-dependent receiver operating characteristic curve (time-dependent AUC), and calibration plots. We used net reclassification index (NRI) and integrated discrimination improvement (IDI) to compare the clinical utility of LCH prediction model with the prediction model based on lifestyle factors (LF). Results: 13 prediction factors including age, sex, BMI, smoking status, physical activity, sleep duration, regular diet, tumor grading, TNM stage, multiple primary cancer and anatomical site were included in the LCH model. The LCH model showed satisfactory discrimination and calibration (C-index = 0.81 (95% CI 0.80–0.82) for training group and 0.80 (0.79–0.81) for validation group, both time-dependent AUC > 0.70). The LF model including smoking status, physical activity, sleep duration, regular diet, and BMI showed less satisfactory discrimination (C-index = 0.60 (95% CI 0.59–0.61) for training and 0.60 (0.58–0.62) for validation group). The LCH model had better accuracy and discriminative ability than the LF model, as indicated by positive NRI and IDI values. Conclusions: The LCH model shows good accuracy, clinical utility and precise prognosis prediction, and may serve as a tool to predict cancer survival of cancer patients. [ABSTRACT FROM AUTHOR]
- Published
- 2023
- Full Text
- View/download PDF