A pilot study of a new test to predict extubation failure

INTRODUCTION To determine whether subjecting patients to 100 ml of additional dead space after a 120-minute weaning trial could predict readiness for extubation. METHODS This was a prospective, non-randomised pilot study in an intensive care unit at a university hospital with 14 beds. It included all non-tracheostomised patients with improvement of the underlying cause of acute respiratory failure, and those with no need for vasoactive or sedative drugs were eligible. Patients fulfilling the Consensus Conference on Weaning extubation criteria after 120 minutes spontaneous breathing (n = 152) were included. To the endotracheal tube, 100 cc dead space was added for 30 minutes. Patients tolerating the test were extubated; those not tolerating it received six hours of supplementary ventilation before extubation. The measurements taken and main results were: arterial pressure, heart rate, respiratory rate, oxygen saturation, end-tidal carbon dioxide and signs of respiratory insufficiency were recorded every five minutes; and arterial blood gases were measured at the beginning and end of the test. Extubation failure was defined as the need for mechanical and non-invasive ventilation within 48 hours of extubation. RESULTS Twenty-two patients (14.5%) experienced extubation failure. Only intercostal retraction was independently associated with extubation failure. The sensitivity (40.9%) and specificity (97.7%) yield a probability of extubation failure of 75.1% for patients not tolerating the test versus 9.3% for those tolerating it. CONCLUSIONS Observing intercostal retraction after adding dead space may help detect susceptibility to extubation failure. The ideal amount of dead space remains to be determined. TRIAL REGISTRATION Current Controlled Trials ISRCTN76206152.