Background Lung-protective ventilation strategies (LPVS) are widely used to reduce postoperative pulmonary complications (PPCs) in patients undergoing one-lung ventilation. The relationship between the partial pressure of carbon dioxide in end-expiratory gas (PetCO₂) and postoperative outcomes remains unclear. This study investigated the association of PetCO₂ with PPCs in patients undergoing thoracoscopic lobectomy with LPVS. Methods This retrospective, observational, single-centre study included patients who underwent thoracoscopic lung lobectomy with LPVS between 31 December 2020 and 31 December 2024. The primary outcome was the incidence of PPCs during hospitalisation; the secondary outcome was postoperative length of stay. General and intraoperative characteristics were compared between patients who developed and those who did not develop PPCs. Multivariable logistic regression was performed to determine independent risk factor of PPCs. Subsequently, a predictive nomogram for PPC events was developed in R. The performance of nomogram model was comprehensively evaluated through ROC curve, calibration plot, and the Hosmer–Lemeshow goodness-of-fit test. Results Among the 383 patients included, 74 (19.3%) developed PPCs. The PPCs group had significantly lower PetCO₂ levels (t = 3.007, p = 0.003) and longer (t = 2.102, p = 0.036). Receiver operating characteristic analysis identified 40.9 mmHg as the optimal PetCO₂ threshold (sensitivity: 80.0%, specificity: 43.4%). Multivariable regression showed that PetCO₂ ≥ 40.9 mmHg (p = 0.040; odds ratio [OR] = 0.516), age (p < 0.001; OR = 1.032), American Society of Anaesthesiologists Physical Status Classification grade III (p = 0.029; OR = 1.876), forced expiratory volume in 1 s/forced vital capacity (p = 0.015; OR = 0.979), and estimated blood loss (p = 0.030; OR = 1.723 per 100mL) were independent risk factor of PPCs. The predictive nomogram model demonstrated an area under the ROC curve of 0.790 (95% CI:0.711–0.827). Calibration assessment revealed a slope approximating unity, and the Hosmer–Lemeshow goodness-of-fit test indicated an adequate model fit (χ² = 5.668, P = 0.196). Conclusions The incidence of PPCs in patients undergoing thoracoscopic lung lobectomy with LPVS was 19.3%. PetCO₂ ≥ 40.9 mmHg was independently associated with a lower risk of PPCs. The constructed nomogram exhibited favorable predictive accuracy for PPCs in patients undergoing thoracoscopic lung lobectomy with LPVS.

Association of intraoperative end-tidal CO₂ levels with postoperative pulmonary complications in thoracoscopic lobectomy: a retrospective cohort study

期刊论文
出版模式：

开放获取
发表日期：

2026年6月1日

查看全文

Zhenming Zhu,
Linjing Sun,
Yongtao Gao

摘要

Background

Lung-protective ventilation strategies (LPVS) are widely used to reduce postoperative pulmonary complications (PPCs) in patients undergoing one-lung ventilation. The relationship between the partial pressure of carbon dioxide in end-expiratory gas (P_etCO₂) and postoperative outcomes remains unclear. This study investigated the association of P_etCO₂ with PPCs in patients undergoing thoracoscopic lobectomy with LPVS.

Methods

This retrospective, observational, single-centre study included patients who underwent thoracoscopic lung lobectomy with LPVS between 31 December 2020 and 31 December 2024. The primary outcome was the incidence of PPCs during hospitalisation; the secondary outcome was postoperative length of stay. General and intraoperative characteristics were compared between patients who developed and those who did not develop PPCs. Multivariable logistic regression was performed to determine independent risk factor of PPCs. Subsequently, a predictive nomogram for PPC events was developed in R. The performance of nomogram model was comprehensively evaluated through ROC curve, calibration plot, and the Hosmer–Lemeshow goodness-of-fit test.

Results

Among the 383 patients included, 74 (19.3%) developed PPCs. The PPCs group had significantly lower P_etCO₂ levels (t = 3.007, p = 0.003) and longer (t = 2.102, p = 0.036). Receiver operating characteristic analysis identified 40.9 mmHg as the optimal P_etCO₂ threshold (sensitivity: 80.0%, specificity: 43.4%). Multivariable regression showed that P_etCO₂ ≥ 40.9 mmHg (p = 0.040; odds ratio [OR] = 0.516), age (p < 0.001; OR = 1.032), American Society of Anaesthesiologists Physical Status Classification grade III (p = 0.029; OR = 1.876), forced expiratory volume in 1 s/forced vital capacity (p = 0.015; OR = 0.979), and estimated blood loss (p = 0.030; OR = 1.723 per 100mL) were independent risk factor of PPCs. The predictive nomogram model demonstrated an area under the ROC curve of 0.790 (95% CI:0.711–0.827). Calibration assessment revealed a slope approximating unity, and the Hosmer–Lemeshow goodness-of-fit test indicated an adequate model fit (χ² = 5.668, P = 0.196).

Conclusions

The incidence of PPCs in patients undergoing thoracoscopic lung lobectomy with LPVS was 19.3%. P_etCO₂ ≥ 40.9 mmHg was independently associated with a lower risk of PPCs. The constructed nomogram exhibited favorable predictive accuracy for PPCs in patients undergoing thoracoscopic lung lobectomy with LPVS.