Perioperative Factors Related to Postoperative Mechanical Ventilation Duration after Thoracic Surgery Requiring One-Lung Ventilation: A Retrospective Cohort Study

Abstract

Background: One-lung ventilation (OLV) is essential for thoracic surgery but may contribute to perioperative gas-exchange impairment and delayed respiratory recovery. Postoperative mechanical ventilation may serve as a pragmatic marker of early respiratory recovery and ICU resource use after thoracic procedures.

Aim: To explore the relationship between a simple set of clinically available perioperative variables and postoperative mechanical ventilation duration after elective thoracic surgery requiring OLV.

Methods: This single-center retrospective observational cohort study included adult patients undergoing elective thoracic surgery requiring OLV. The primary outcome was the time from ICU admission to successful extubation, defined as the absence of reintubation within 48 hours. The multivariable linear regression model of log-transformed ventilation duration included ARISCAT score, OLV duration, intraoperative P/F ratio, and surgical approach.

Results: Ninety-seven patients had complete data for the main multivariable analysis, and no patient required reintubation within 48 hours after extubation. The overall model was statistically significant (F (4,92) = 3.85, R² = 0.1435, adjusted R² = 0.1063, p = 0.0061). OLV duration showed the strongest hypothesis-generating trend: each additional 30 minutes of OLV corresponded to an estimated 8.4% longer postoperative ventilation duration (95% CI, -1.1% to +18.9%, p = 0.082). ARISCAT score, P/F ratio, and thoracotomy versus VATS showed no statistically significant relationship with ventilation duration.

Conclusions: In this retrospective cohort, commonly available perioperative variables explained only a small proportion of the variability in postoperative ventilation duration, and no individual variable demonstrated an independent statistically significant relationship with the outcome. The non-significant trend observed for OLV duration should be considered hypothesis-generating and evaluated in larger cohorts.