Abstract: Objective To systematically identify independent risk factors for postoperative pleural effusion (PE) following video-assisted thoracoscopic surgery (VATS), develop a risk prediction model, and evaluate the protective effect of preoperative paravertebral block (PVB).Methods This study integrated multicenter retrospective and prospective cohorts, including 2 141 patients who underwent thoracoscopic surgery in the Department of Thoracic Surgery at Tongji Hospital, Wuhan, between August 25, 2024 and December 31, 2024.Postoperative pleural effusion was assessed using chest CT, and relevant biomarkers were measured.Multivariable ordinal Logistic regression was used to identify independent predictors of postoperative PE.To explore the association between different analgesic modalities including PVB, intercostal nerve block (INB), and local infiltration analgesia (LIA) and inflammatory or metabolic markers were further analyzed.Results Multivariable analysis identified older age, lower postoperative albumin, surgical type, extent of intraoperative lymph node dissection, and analgesic modality as independent predictors of postoperative PE(P＜0.05).PVB provided significant protective effects compared with LIA or INB.A nomogram incorporating these factors demonstrated good discrimination and calibration in both the training cohort (AUC=0.704) and the validation cohort (AUC=0.692).Further analyses indicated that PVB significantly reduced postoperative elevations in C-reactive protein and fasting plasma glucose(P＜0.05).Conclusions Perioperative PVB independently reduces the risk of postoperative PE after VATS, potentially via anti-inflammatory and antioxidative mechanisms.The nomogram effectively stratifies patient risk, supporting individualized perioperative management.

Key words: video-assisted thoracoscopic surgery, pleural effusion, perioperative analgesia, paravertebral block, observational study