Prediction models for postoperative atrial fibrillation in lung cancer surgery show moderate discriminative performance

BackgroundPostoperative atrial fibrillation (POAF) is a common and clinically significant complication following lung cancer surgery, associated with increased morbidity and mortality. Although numerous prediction models have been developed to estimate POAF risk, their overall performance and methodological quality remain unclear.MethodsA systematic review and meta-analysis were conducted in accordance with the PRISMA 2020 guidelines, and the protocol was registered with PROSPERO (CRD42025115874). Chinese and English databases were searched from their inception until 30 May 2024. Studies that developed or validated prediction models for postoperative atrial fibrillation (POAF) in patients with surgically treated lung cancer were included. Data were extracted using the CHARMS checklist and the risk of bias was assessed using PROBAST. A random-effects meta-analysis was performed to pool the discriminative performance of the eligible models, using the area under the curve (AUC).ResultsSix studies were included. Most models were developed using logistic regression, with age, sex, cardiovascular comorbidities and surgical factors being the most common predictors. Reported area under the curve (AUC) values ranged from 0.72 to 0.89. The pooled AUC was 0.79 (95% CI: 0.71–0.87), which indicates good overall discrimination. However, substantial heterogeneity was observed (I2 = 98.7%). Subgroup analysis with consistent outcome definitions showed reduced heterogeneity. All studies were judged to have a high overall risk of bias.ConclusionsCurrent POAF prediction models for lung cancer patients show acceptable discriminative ability but are limited by methodological weaknesses and lack of external validation, restricting their clinical applicability.Systematic Review Registrationhttps://www.crd.york.ac.uk/PROSPERO/view/CRD420251158742, identifier CRD420251158742.