Weaning predictors do not predict extubation failure in simple-to-wean patients

Abstract

Background: Predictor indexes are often included in weaning protocols and may help the intensive care unit (ICU) staff to reach expected weaning outcome in patients on mechanical ventilation.

Objective: The objective of this study is to evaluate the potential of weaning predictors during extubation.

Design: This is a prospective clinical study.

Settings: The study was conducted in 3 medical-surgical ICUs.

Patients: Five hundred consecutive unselected patients ventilated for more than 48 hours were included.

Methods and measurements: All patients were extubated after 30 minutes of successful spontaneous breathing trial and followed up for 48 hours. The protocol evaluated hemodynamics, ventilation parameters, arterial blood gases, and the weaning indexes frequency to tidal volume ratio; compliance, respiratory rate, oxygenation, and pressure; maximal inspiratory pressure; maximal expiratory pressure; Pao(2)/fraction of inspired oxygen; respiratory frequency; and tidal volume during mechanical ventilation and in the 1st and 30th minute of spontaneous breathing trial.

Results: Reintubation rate was 22.8%, and intensive care mortality was higher in the reintubation group (10% vs 31%; P < .0001). The areas under the receiver operating characteristic curve showed that tests did not discriminate which patients could tolerate extubation.

Conclusion: Usual weaning indexes are poor predictors for extubation outcome in the overall ICU population.