Electrical Impedance Tomography-derived Flow Index During Spontaneous Breathing Trial Stratifies the Risk of Reintubation Within 48 h After Extubation

Overview

Accurate prediction of readiness to liberate patients from mechanical ventilation remains challenging. Conventional indices such as the rapid shallow breathing index (RSBI) and maximal inspiratory pressure (MIP) often miss early signs of injurious breathing patterns or regional ventilation asynchrony that can lead to extubation failure. Electrical impedance tomography (EIT) provides continuous, non-invasive imaging of regional lung ventilation. We developed a novel EIT-derived Flow Index (FI) which integrates the magnitude of inspiratory effort with the temporal synchrony of lung filling. This prospective, multicenter observational study aimed to (1) validate the predictive value of FI during spontaneous breathing trials (SBT) compared with conventional weaning indices, and (2) compare the predictive ability of EFI with traditional weaning indices(RSBI,MIP,P0.1).

This multicenter observational study was conducted in three tertiary ICUs in China. Adult patients (≥18 years) receiving invasive mechanical ventilation for ≥48 hours and meeting readiness criteria underwent a standardized 30-minute pressure support SBT (PS 8 cmH₂O, PEEP 5 cmH₂O, FiO₂ ≤0.5) with continuous EIT monitoring. EFI was calculated breath by breath from the global impedance-time signal. Conventional weaning indices (RSBI, MIP, P0.1) and physiological variables were recorded. Primary endpoint: SBT success (completion of 30-minute SBT without predefined failure criteria). Key secondary endpoint: reintubation within 48 hours after extubation. Other secondary endpoints: ventilator-free days at day 7 (VFD-7), ICU mortality. Predictive performance was assessed using ROC analysis. The EFI cutoff value was derived from the primary endpoint analysis and carried forward for post-extubation risk stratification.

Conditions