Electrical impedance tomography assessment of bronchoalveolar lavage-associated lung derecruitment and its related factors: a single-center retrospective study.

BACKGROUND: Bronchoalveolar lavage (BAL) may cause regional lung derecruitment during bronchoscopy. However, this effect may be underestimated by conventional oxygenation monitoring, especially under general anesthesia and mechanical ventilation. This study aimed to assess BAL-associated lung derecruitment using electrical impedance tomography (EIT) and to explore related clinical, radiological, and physiological factors. METHODS: This single-center retrospective observational study included adult patients who underwent BAL with periprocedural EIT monitoring between March 2024 and October 2025. EIT data were analyzed at three time points: before BAL, immediately after BAL, and after completion of the procedure. The primary outcome was residual end-expiratory lung impedance loss after BAL. Secondary outcomes included changes in defect score, global inhomogeneity index, center of ventilation, regional ventilation delay, respiratory mechanics, and oxygen saturation. Linear regression was used to identify factors associated with residual EELI loss. RESULTS: Among 405 screened patients, 286 met the inclusion criteria, and 277 were finally analyzed. BAL was associated with a significant decrease in EELI and increases in defect score and global inhomogeneity index, indicating lung derecruitment and more heterogeneous ventilation distribution. Driving pressure increased and dynamic compliance decreased after BAL, while SpO2 showed no significant change. In multivariable analysis, preoperative atelectasis was independently associated with greater residual EELI loss, with an adjusted β coefficient of 1.71 (95% CI 0.41-3.00, P = 0.010). Baseline defect score was also independently associated with residual EELI loss, with an adjusted β coefficient of 0.51 (95% CI 0.07-0.95, P = 0.025). In contrast, total BAL instilled volume, number of BAL cycles, and bronchoscope insertion duration were not significantly associated with residual EELI loss. CONCLUSIONS: BAL was associated with immediate lung derecruitment and worsening of regional ventilation distribution under general anesthesia and mechanical ventilation, despite stable SpO2. Preoperative atelectasis and baseline regional ventilation defects were independently related to incomplete lung volume recovery after BAL. EIT may help identify patients at higher risk of BAL-associated lung derecruitment and support individualized respiratory management during bronchoscopy.