Patients with high-level cervical SCIs for whom mechanical ventilation cannot yet be discontinued either will or will not pass an SBT, depending on the recurrence of some ventilator-free breathing ability. They usually remain cannulated for prolonged periods of time, often indefinitely. They remain at increased risk of airway damage from suctioning procedures, complications associated with the tracheostomy, airway microbial colonization, pulmonary infection, and recurrent ICU admission. Because they cannot breathe, a capping trial may seem inappropriate or even dangerous for the simple reason of interrupting invasive mechanical ventilation. Often, the presumed permanent dependence on the artificial airway is accepted by physicians because of the continuous necessity for removal of bronchial secretions. It is rarely appreciated that these artificial airways cause and increase secretions and impair the ability to cough. Many tracheostomized patients, however, choose to be decannulated or to forgo tracheostomy if noninvasive techniques are available [10].

The question then arises regarding whether and how to remove the tracheostomy tube for these patients. In the 1990s, Bach and coworkers [11, 12] showed that patients with high-level cervical SCIs without ventilator-free breathing ability could be extubated and decannulated by converting them from invasive to noninvasive ventilation using facial interfaces during the night and mouthpiece ventilation during the day. It was extremely important that patients were able to manage their bronchial secretions, which were the main reason for oxygen desaturation, atelectasis, retention pneumonia, and hospitalization. When using assisted coughing techniques following aggressive lung volume recruitment (LVR), a cough peak flow (CPF) of > 160 l/min could usually be achieved, and this proved enough to safely permit the removal of the tracheostomy tube. Bach et al. proposed continuing LVR techniques following decannulation every day to prevent airway secretion encumbrance. Patients could use air (or breath) stacking or glossopharyngeal breathing (GPB) to augment lung volume, improve voice volume, improve respiratory compliance, facilitate noninvasive mechanical ventilation, and improve CPFs. In addition, they could regain ventilator-free breathing ability by using GPB as a breathing method. McKim and coworkers [8] demonstrated that, using LVR and assisted coughing techniques, CPFs measured after decannulation were some 30–40 l/min greater than during coughing with the capped cannula in place using these techniques, probably because of the surplus airway resistance of the endotracheal part of the tube. They suggested that, in cannulated patients, these findings might lower the CPF threshold of 160 l/min with this flow difference as an indication for safe decannulation.

For the high cervical SCI patient who cannot pass an SBT, discontinuing invasive mechanical ventilation often seems impossible. In the ICU it may go unnoticed that one of the main reasons for failure to discontinue invasive ventilation is airway secretion encumbrance. During their training, intensivists have been taught to rely on artificial airways to invasively remove airway secretions and they have seldom been trained in using LVR techniques, which are necessary to effectively resolve retention of bronchial secretions. Hence, they may be reluctant to remove the tracheostomy tube because they think the patients depend on them for mucus removal. Of note, contrary to what is seldom considered, these patients often can be successfully liberated from their artificial airways [13]. In these patients, the aims are to discontinue invasive ventilation, proceed with mechanical ventilation noninvasively, and remove the artificial airway. The latter goal cannot be achieved without effective airway clearance techniques [13, 14].

For SCI patients in the ICU, early treatment with LVR techniques is recommended. For an effective tution of air stacking or GPB, the patient should be awake, alert, cooperative, and highly motivated to achieve the maximum insufflation capacities necessary to enhance CPFs for mucus removal. Training in these techniques is time consuming and requires a dedicated and patient teacher. In the early post-injury phase it is more prudent to use mechanical insufflation-exsufflation (MI-E) with a cough-assist device, which can also be operated on a cuffed tracheostomy tube for unresponsive patients. Only after lungs and airways have been completely cleared of secretions, which in our experience can never be achieved with invasive suctioning procedures only, training in noninvasive mechanical ventilation may be initiated. Starting invasive LVR techniques early in the course of recovery may prevent pulmonary complications and can shorten ICU stay and length of hospital admission [14].