The Weaning Index during Spontaneous Breathing Trials for Predicting Extubation Failure in Critically Ill Patients: A Bicentric Prospective Study.

BACKGROUND

An increase in the Weaning Index (WI), defined as the product of end-tidal carbon dioxide (EtCO2) and respiratory rate (RR), has been observed in patients who fail spontaneous breathing trials (SBT). This study aimed to assess whether the WI recorded during SBT can predict extubation failure (EF) in critically ill patients.

METHODS

In this bicentric prospective study, ICU patients mechanically ventilated for at least 48 hours and who successfully tolerated a 30-minute SBT were included. EtCO2, RR, and WI were recorded at the beginning and end of the SBT. Changes in these parameters were analysed. Predictive performance for EF on day 7 was assessed, including areas under the receiver operating characteristic curves (AUCs).

RESULTS

Among the 114 patients, 23 (20.2%) experienced EF. Absolute changes (median [IQR]) of EtCO2, RR and WI were higher in failure group than success group respectively: 1.0 [-1.0-3.4] mmHg, 3 [1-5] breaths/min and 169.0 [94.0-232.0] mmHg x breaths/min versus -0.2 [-0.2-1.0] mmHg (p = 0.003), 0 [-1-1] breaths/min (p < 0.001) and -6.0 [-59.0-38.0] mmHg x breaths/min (p < 0.001). In multivariate adjusted models, increases in EtCO2, RR, and WI were independently associated with EF, respectives OR [95% CI]: 1.78 [1.30-2.44], 3.26 [1.77-6.01], and 1.03 [1.01-1.04]. ΔEtCO2 showed limited predictive ability for EF (AUC = 0.70 [95%CI: 0.61-0.78]), whereas ΔRR and ΔWI demonstrated excellent discrimination (AUC = 0.84 [95%CI: 0.76-0.90] and AUC = 0.91 [95%CI: 0.85-0.96]).

CONCLUSIONS

SBT-induced changes in RR and WI demonstrated excellent performance in predicting EF in critically ill patients. In contrast, changes in EtCO2 during the SBT provided limited predictive value.

REGISTRATION: The study was registered in the ClinicalTrials.gov registry: NCT05802745.