Construction and Validation of a Prediction Model for Postoperative Fatigue Syndrome in Chinese Patients with Lung Cancer

Background: Postoperative fatigue syndrome (POFS) is prevalent in patients with lung cancer after surgery but often overlooked clinically, affecting patient care and recovery. Predictive models for assessing the risk and severity of postoperative fatigue in persons diagnosed with lung cancer are lacking. Objective: To develop a predictive model for POFS in patients with lung cancer to address under-recognition and its impact on recovery. Methods: Data from 203 lung cancer surgery patients were analyzed through univariate analysis to compare the relevant factors between 2 groups . Least absolute shrinkage and selection operator regression were used to screen potential key predictors. Multivariate regression analysis was used to identify independent influencing factors and build a nomogram. Receiver operating characteristic curve, calibration curve, and decision curve analysis (DCA) were used to evaluate the discrimination, accuracy, and clinical usability of the prediction model, with internal validation by the Bootstrap method. Results: Of the 203 patients, 57.1% developed POFS. The prediction model included 5 significant predictors: sleep quality, pain, activated partial thromboplastin time, forced vital capacity, and forced expiratory volume in 1 second/forced vital capacity ratio. The nomogram based on this model achieved an area under the receiver operating characteristic curve of 0.870, indicating good accuracy, with strong predictive power in internal validation. DCA showed clinical utility when the probability of POFS was above approximately 13%. Conclusions: We found a high prevalence of POFS in survivors with lung cancer and successfully constructed a comprehensive nomogram with 5 factors.