The Effect of Preoperative Preparation with Helium-Oxygen Mixture on the Incidence of Pulmonary Complications in Patients with COPD and Lung Cancer

Aim of the study. To study the effect of helium-oxygen mixture on predictors of postoperative pulmonary complications in cancer patients with chronic obstructive pulmonary disease (COPD). Materials and methods. A single-center prospective clinical study with historical control included 208 patients, among them 104 patients received helium-oxygen mixture inhalations (70% helium and 30% oxygen) in the preoperative period (He group) and 104 patients were the historical control group (Ctrl group). Given the risk of bias associated with confounders, we conducted a one-to-one matching analysis based on pseudorandomization to adjust for the unbalanced baseline characteristics of the groups. We used logistic regression to develop the pseudo-randomization estimates. We used the nearest neighbor matching 1:1 with a caliper of 0.1 to achieve better similarity among matched pairs. After pseudo-randomization, we included 87 patients in each group, ensuring adequate balance across all covariates. A multivariate logistic regression analysis was performed to identify factors associated with the risk of postoperative pulmonary complications. Results. In the He group, there was a statistically significant improvement in a number of functional parameters compared to the Ctrl group. FEV1, FVC, mod. Tiffno index values in the He group increased significantly ( р = 0,0009; р = 0,0115; р = 014, respectively), gas exchange parameters (PaO₂, PaCO₂, pH, SpO₂) improved ( р = 0.0006; р = 0.004; р = 0.0097; р = 0.001, respectively). Hypoxia tolerance tests also showed significantly greater values in the He group (Stange test, р = 0.016; Sabrazes (Gench) post-exhalation breath-holding test, р = 0.024). Analysis of postoperative parameters showed significant advantage of the He group over the control group in terms of critically important clinical outcomes. At the final stage of stepwise selection, three independent predictors were included in the risk model for postoperative pulmonary complications: SpO₂, the breath-holding test, and the duration of postoperative mechanical ventilation. The quality of the model was high, with a correctly classified case rate of 92.2%, a Hosmer–Lemeshow goodness-of-fit statistic of р = 0.933, and a total model significance of p