Postoperative pulmonary complications after high-altitude thoracic surgery requiring one-lung ventilation: a single-centre retrospective cohort study

Abstract

Abstract Background Postoperative pulmonary complications (PPCs) are common adverse outcomes after thoracic surgery and one-lung ventilation (OLV). Their reported incidence varies with endpoint definition and ascertainment. Lower inspired oxygen partial pressure may alter perioperative oxygenation at high altitude, but direct evidence from thoracic OLV cohorts remains limited. We described PPC occurrence and evaluated associated patient and perioperative factors in a high-altitude thoracic OLV cohort. Methods This single-centre retrospective cohort included adults undergoing thoracic surgery with general anaesthesia and OLV at Tibet Autonomous Region People’s Hospital (Lhasa, approximately 3,650 m). The primary outcome was an in-hospital composite outcome comprising postoperative pneumonia, imaging-confirmed atelectasis, acute respiratory distress syndrome (ARDS), postoperative respiratory failure, initial postoperative ventilatory support lasting more than 48 h, or reintubation, with events assessed within the first 7 postoperative days. The primary multivariable logistic regression analysis used multiple imputation by chained equations and included 14 covariates. Complete-case and no-BMI multiple-imputation models were sensitivity analyses. Marginally standardised risk ratios (RRs) and average marginal effects (AMEs) complemented odds ratios (ORs). Results Among 492 patients, 211 developed PPCs (42.9%), mainly atelectasis (141, 28.7%) and pneumonia (113, 23.0%). In the primary model, ASA physical status at least III (OR 2.37, 95% confidence interval [CI] 1.32–4.26; P = 0.004), preoperative anaemia (OR 2.74, 95% CI 1.56–4.81; P < 0.001), surgical duration per 1-hour increase (OR 1.40, 95% CI 1.17–1.68; P < 0.001), and bronchial blocker use (OR 2.81, 95% CI 1.33–5.90; P = 0.007) remained associated with PPCs after multivariable adjustment. Corresponding RRs for these four factors were 1.45, 1.52, 1.16, and 1.51, and AMEs were + 18.1, + 20.6, + 6.8, and + 21.0% points, respectively. The directions and statistical significance of these associations were consistent in sensitivity analyses. Conclusions In this single-centre high-altitude thoracic OLV cohort, PPCs occurred in 42.9% of patients and were predominantly atelectasis and pneumonia. After multivariable adjustment, ASA physical status ≥ III, preoperative anaemia, longer surgical duration, and bronchial blocker use remained associated with PPCs. The bronchial-blocker estimate is susceptible to confounding by indication. Without a low-altitude comparator or granular ventilation data, the findings should be interpreted as observational, hypothesis-generating signals rather than causal effects of altitude or lung-isolation devices. External validation is required.