Perioperative predictors of recurrence after decortication for tuberculous empyema

Abstract Background Chronic tuberculous empyema is a serious complication of tuberculosis, and recurrence sometimes follows surgical decortication. How routine perioperative and treatment-related variables shape this risk remains unclear. Methods We retrospectively reviewed patients with microbiologically proven tuberculous empyema who underwent video-assisted thoracoscopic surgery or open decortication at a single centre. Demographic data, comorbidities, inflammatory and nutritional indices, coagulation markers, operative time, blood loss, surgical approach and total duration of preoperative anti-tuberculous therapy (ATT) were collected. Recurrence(referring to imaging-defined recurrence) was defined as new pleural effusion, cavity or atelectasis on follow-up imaging. Logistic regression identified independent predictors and model performance was evaluated by area under the curve, calibration and decision-curve analysis. Results Baseline demographic, comorbidity and microbiological characteristics had limited ability to discriminate recurrence. Higher preoperative C-reactive protein, greater intraoperative blood loss and thoracotomy (vs. video-assisted thoracoscopic surgery) were independently associated with recurrence, whereas a longer total ATT course (≥ 3 months) reduced risk. A simple model using these four factors achieved an area under the curve of about 0.70 with acceptable calibration and clinical utility. Notably, because VATS was non-randomly applied and reserved for less complex disease, its apparent protective signal likely reflects underlying severity bias rather than a proven causal surgical advantage. Conclusions In tuberculous empyema, recurrence after decortication appears to be driven more by systemic inflammation, operative difficulty and adequacy of ATT than by baseline patient profile, and can be estimated with readily available perioperative variables.