Weaning critically ill adults from invasive mechanical ventilation: a national survey

To determine the stated practices of clinicians in weaning critically ill adults from invasive ventilation.

We conducted a cross-sectional, self-administered postal survey of Critical Care physicians and respiratory therapists (RTs) in leadership roles at Canadian teaching hospitals. We identified respondents using electronic mail and telephone correspondence. We used rigorous survey methodology to develop, test, and administer the questionnaire.

One hundred ten of 162 (67.9%) clinicians returned the survey with 99 respondents (55 physicians and 44 RTs) completing it either in-part or in-full. Approximately 95% of respondents acknowledged ever performing spontaneous breathing trials (SBTs) in clinical practice. Of these, 95.6% and 32% of respondents reported conducting daily and twice-daily screening to identify SBT candidates, at least sometimes. The three most common techniques to conduct SBTs included; pressure support (PS) with positive end-expiratory pressure (70.8%), continuous positive airway pressure (35.7%), and use of a T-piece (25.0%). PS ventilation was the weaning strategy used most frequently before SBTs. Most respondents (57.1%) considered continuous infusion of sedative-hypnotics to be a relative contraindication to tracheal extubation. However, concurrent administration of low dose vasopressors, inotropes, and analgesic boluses, or continuous analgesic infusions were considered acceptable amongst 60.8%, 73.2%, 78.4% and 58.8% of respondents, respectively. We did not observe regional variation in whether clinicians ever perform SBTs, the ventilatory modes used prior to an SBT nor in the use of PS and SBTs during the weaning process.

Pressure support and SBTs are common features of weaning in Canadian teaching hospitals. Compared to the published literature, our survey suggests that weaning practices have evolved over time and that practice variation may be greater on an international level compared to a national level.