Secondary Logo

Journal Logo


Proposed Decannulation Criteria for COVID-19 Patients

Giménez, Gloria C. BA; Müller-Thies, Marlene BA; Prado, Francisco J. MD; Bach, John R. MD

Author Information
American Journal of Physical Medicine & Rehabilitation: August 2021 - Volume 100 - Issue 8 - p 730-732
doi: 10.1097/PHM.0000000000001788


Recent publications comprehensively review many rehabilitation issues for recovering COVID-19 patients; however, criteria for decannulation are a glaring omission.1 Patients recovering from COVID-19 who remain ventilator dependent for weeks typically undergo tracheotomies.2,3 Foreign bodies like translaryngeal and tracheostomy tubes greatly increase the risk of “ventilator-associated pneumonias.” This is really an inaccurate designation for invasive interface-associated pneumonia because patients get pneumonia from the interface, not from the ventilator. Patients with tracheostomies have as much pathogenic bacteria in their lungs as patients with acute pneumonias because there is no antibiotic delivery to pathogenic bacteria on the tubes.4 Therefore, the sooner the tubes are removed, the better. Although there are publications that consider indications for placing tracheostomy tubes,2 and 8%–30% of COVID-19 patients in critical care undergo tracheotomies,3,5 there are no guidelines for decannulation6 and use of mechanical insufflation-exsufflation (MIE) has not been reported for this population.

There is only one publication about a patient with advanced adult respiratory distress syndrome (ARDS) secondary to COVID-19 who was weaned using average volume-assured pressure support, as is similar for other causes of ARDS treated with traditional noninvasive ventilation.7 Although different pathophysiological mechanisms would be involved in COVID-19 ARDS,8 long-term lung damage observed in the convalescent stage is common to other ARDS.9,10

Normally, patients are not extubated or decannulated until they are weaned from ventilatory support. However, patients with ventilatory pump failure who pass no ventilator weaning parameters or spontaneous breathing trials can also be routinely extubated or decannulated of tracheostomy tubes to continuous noninvasive ventilatory support (CNVS) if they have no ventilator-free breathing ability at all or to up to CNVS if they can breathe for at least a few minutes on their own.11–13

Decannulation criteria for ventilator unweanable patients to CNVS include having normal oxyhemoglobin saturation (O2 sat) in ambient air. Typically, however, COVID-19 patients recover the ability to breathe before their pulmonary dysfunction and blood oxygenation normalizes.9,10 Even when weaned from ventilatory support and O2 dependent, however, many patients are not decannulated in long-term acute care centers and ventilator units for fear of liability.14 There is also a lack of criteria for decannulation and lack of knowledge of NVS and MIE alternatives to invasive ventilation. In addition, although MIE has proven to be critical for the extubation and decannulation of ventilator unweanable patients to CNVS,11–13 there is only one publication reporting it to increase successful extubation rates in lung disease patients.15

Patients surviving severe COVID-19 infections are often extubated to up to continuous tracheostomy mechanical ventilatory support and have both lung damage9,10 and respiratory muscle weakness with ventilatory pump failure for months. There is no rehabilitation literature that considers criteria for decannulation or decannulation protocols for patients still dependent on oxygen supplementation with ventilatory pump failure as well. Although criteria for extubating unweanable patients include having normal O2 sat in ambient air, COVID-19 patients pose a challenge because the ability to breathe usually returns before ambient air O2 sat baselines return to normal. An occasional exception to this is the patient whose lung health returns but who remains ventilator dependent because of critical care neuromyopathy.16

Patients can be decannulated if too weak to breathe but with healthy lungs or with unhealthy lungs but strong enough to breathe but not if unhealthy and too weak. Clearly, even ventilator unweanable patients with healthy lungs can be decannulated to CNVS and MIE. We report preparation for possible decannulation of an apparently unweanable patient and his decannulation when still hypoxic by following the proposed decannulation criteria in Table 1. This study conforms to all Case Report guidelines and reports the required information accordingly (see Checklist, Supplemental Digital Content 1,

TABLE 1 - Proposed decannulation criteria for COVID-19 patients
Afebrile since discontinuation of antibiotics
White blood cell count normal off of antibiotics for at least 10 d
Alert and cooperative
No multiorgan failure or severe cardiovascular involvement
O2 requirement decreasing enough to maintain at least 95% of Spo 2
Vital capacity >1.5 l
Cough peak flows >300 l/m
Chest x-rays and computed tomography scans stable or improving


A 66-yr-old man was well until having fever and symptoms of COVID-19 and was admitted on March 1, 2020, intubated from March 10–31 and underwent tracheotomy for continuous tracheostomy mechanical ventilation from April 1, 2020. He weaned to tracheostomy collar breathing on May 26, 2020. He had required antibiotics on and off for 2 mos through May 26 because of recurrent pneumonias. From May 26 through June 2, he breathed via a tracheostomy collar on 10 l/min of O2 with Spo2 98%. Chest tubes for multiple pneumothoraces and effusions were removed on June 1. Refusing transfer to long-term care, he was discharged directly home to the care of his wife and four children.

On June 3, we found him on 10 l/O2. His venous CO2 levels were mildly elevated, and with deep insufflations, his O2 sat increased, so hypercapnia was suspected. In ambient air, his Spo2 quickly plummeted less than 70%. We switched him to a cuffed fenestrated size 6 Shiley tube, the cuff for effective use of MIE to clear airway secretions and the fenestration to facilitate breathing and using nasal and mouthpiece NVS via the upper airways. His cough flows were ineffective, so MIE was used via the tube at 15 to −50 cm H2O pressures every 1–2 hrs when awake with the cuff inflated. The 15 cm H2O positive pressure was the same that he had used for sleep tracheostomy mechanical ventilation. With chest tubes and pneumothoraces, it is prudent to limit insufflation pressures to those used for respiratory support. Breathing via the upper airway between tube changes with the ostomy covered was not possible because of stridor. A portable ventilator was ordered anticipating that he might require NVS after decannulation increased his dead space breathing.

We next saw him on June 25. After 3 wks of frequent MIE via the tracheostomy tube, his vital capacity (VC) was 900 ml, and he was now able to breathe via the upper airway without stridor. He was practicing nasal NVS at 18–20 cm H2O pressure assist control ventilation. People with VC of less than 1000 ml usually require nasal NVS for sleep after decannulation. The O2 sat breathing in ambient air was 86% and 97% on 4 liters of O2/min. He continued tracheostomy collar breathing and MIE via the tube but now at 25 to −50 cm H2O for decreases in O2 sat baseline apparently caused by airway secretions.

Five days later, on June 30, with a VC of 1000 ml, end-tidal CO2 34 mm Hg, O2 sat was normal on 3 liters of O2 but 87% in ambient air. After MIE sessions, the O2 sat increased to 93% in ambient air. The #6 Shiley tube was changed to a #4 cuffless fenestrated Shiley, which, when capped, would permit him to breathe or use nasal or mouthpiece NVS via the upper airway, with less obstruction, in case he fatigued. Antibiotics had been discontinued for 8 days. White blood cell count was normal. He mastered nasal and mouthpiece NVS via a 15-mm angled mouthpiece.

On July 10, his VC had suddenly increased to 2140 ml sitting and 1940 ml supine and cough peak flows were 330 l/m. He no longer needed NVS or MIE. However, he was still hypoxic, requiring 2 liters of O2/min for normal O2 sat. Nevertheless, with improvements in his interstitial lung disease and infiltrates documented on computed tomography scans, afebrile for 2 wks after discontinuation of antibiotics, and satisfying all other Table 1 criteria, he was decannulated. Despite 2 mos of antibiotics for repeated and persistent lung infections, he had no further infections nor antibiotic use after decannulation. His ostomy closed in 2 days. By telephone, we learned that his O2 sat became a steady 95% in ambient air by July 21, 2020, and he only used 2 liters of O2 during daily home physical therapy sessions for strengthening by graded exercise intensity, transfers, reverticalization, reambulation, and gait training with oximetry monitoring.


We present a successful case of weaning a COVID-19 patient with ventilatory failure and hypoxemia as a model for how to provide care for this new population of patients. Although physical medicine interventions, like NVS and MIE, that is, applying pressures to the body or airways,17 can be used to safely extubate11,12 or decannulate ventilator-dependent patients who have normal O2 sat in ambient air13,18,19 or hypoxic patients who are strong enough to breathe to normal CO2, tube removal for hypoxic patients still ventilator dependent is not advised. Thus, the strategies for this patient were to normalize his ambient air O2 sat by clearing airway secretions with the use of MIE via the tube and subsequently via the upper airway, therefore, to clear atelectasis and lung infection, and to train him in nasal and mouthpiece NVS in the event that his lungs become healthy so that he can be decannulated before he is strong enough to breathe as we do for ventilator unweanable patients with neuromuscular conditions. Most COVID-19 patients wean before their lungs become healthy.5

To improve lung health in preparation for decannulation, MIE was used via the tube with the cuff inflated at the positive inspiratory pressure he was using for tracheostomy mechanical ventilation, 15 to −60 cm H2O exsufflation pressure. Full negative pressure can be used because it would not be expected to cause or exacerbate any lung injury or barotrauma.

In the current SARS-CoV-2 pandemic, despite the initial fear of an increased risk from viral aerosolization, O2 therapy with a high-flow nasal cannula and high-flow nasal prong therapy are used in early stages of disease as well as for any subsequent extubation.20 Another physical medicine intervention that has been reported to decrease intubation rates and renormalize O2 sat is cold laser therapy.21

Decannulating ventilator unweanable patients to CNVS and MIE facilitates ventilator weaning,22 and it reduces airway secretion difficulties because very often it is the tracheostomy tube that causes most of the secretions and hampers their expulsion by coughing. Furthermore, although incidence of critical care neuromyopathies for patients with ARDS is well known, it is unclear how much COVID-19 increases their incidence.16 Such patients can remain too weak to breathe long after their lungs are healthy enough for normal O2 sat in ambient air but can be extubated to CNVS and MIE without resort to tracheotomy. Unfortunately, patients who undergo tracheotomies often unnecessarily retain them when spontaneous ventilation cannot be achieved, although ventilatory support can be provided noninvasively.7,23–25

Use of MIE via the tracheostomy tube and noninvasive interfaces has not been reported for this patient population. However, used at reduced insufflation settings, MIE still expelled secretions, and both increased O2 sat acutely as well as baseline O2 sat over time. Despite his many blebs, MIE caused no further barotrauma. Using it every 2 hrs, his VC and ability to ventilate his lungs improved and O2 requirement decreased to 3 l/min at decannulation. According to the World Health Organization, it seems safe to release patients from isolation based on clinical criteria that require a minimum time in isolation of 13 days. It is important to note that the clinical criteria require that patients’ symptoms have been resolved for at least 3 days before release from isolation, with a minimum time in isolation of 13 days since symptom onset. After 20 days of severe illness, COVID-19 patients are not thought to be contagious, and with regular level 3 personal protective equipment, transmission risk is very low. However isolation and negative pressure room might still be recommended. We just do not know for sure yet.


In summary, it is useful to monitor VC, cough peak flows, O2 sat, and CO2 and recommended to use MIE via invasive airway tubes and via the upper airways for COVID-19 patients and for others too weak to breathe and cough unaided.11–13 Mechanical cough assistance and NVS have also proven to be very useful strategies for various causes of ventilatory pump failure, even in patients with no ventilator-free breathing ability. As has been mentioned in the current COVID-19 pandemic, after a period, generally not more than 20 days and with adequate personal protection measures, the risk of infection by viral aerosolization is minor, and therefore, use of noninvasive therapeutic measures can be safe. Use of NVS and MIE, with appropriate precautions, can help wean these patients. Thus, it is important for critical care staff to gain experience and develop skill in offering NVS and MIE to facilitate the extubation and decannulation of these and other critical care patients.


1. Wang TJ, Chau B, Lui M, et al.: Physical medicine and rehabilitation and pulmonary rehabilitation for COVID-19. Am J Phys Med Rehabil 2020;99:769–74
2. Lamb CR, Desai NR, Angel L, et al.: Use of tracheostomy during the COVID-19 pandemic: American College of Chest Physicians/American Association for Bronchology and Interventional Pulmonology/Association of Interventional Pulmonology Program Directors Expert Panel Report. Chest 2020;158:1499–514
3. Queen Elizabeth Hospital Birmingham COVID-19 airway team: Safety and 30-day outcomes of tracheostomy for COVID-19: a prospective observational cohort study. Br J Anaesth 2020;125:872–9
4. Baram D, Hulse G, Palmer LB: Stable patients receiving prolonged mechanical ventilation have a high alveolar burden of bacteria. Chest 2005;127:1353–7
5. Cummings MJ, Baldwin MR, Abrams D, et al.: Epidemiology, clinical course, and outcomes of critically ill adults with COVID-19 in New York City: a prospective cohort study. Lancet 2020;395:1763–70
6. Stelfox HT, Crimi C, Berra L, et al.: Determinants of tracheostomy decannulation: an international survey. Crit Care 2008;12:R26
7. Mittal A, Forte M, Leonard R, et al.: Refractory acute respiratory distress syndrome secondary to COVID-19 successfully extubated to average volume-assured pressure support non-invasive ventilator. Cureus 2020;12:e7849
8. Gattinoni L, Chiumello D, Caironi P, et al.: COVID-19 pneumonia: different respiratory treatments for different phenotypes? Intensive Care Med 2020;46:1099–102
9. McDonald LT: Healing after COVID-19: are survivors at risk for pulmonary fibrosis? Am J Physiol Lung Cell Mol Physiol 2021;320:L257–65
10. Wang F, Kream RM, Stefano GB: Long-term respiratory and neurological sequelae of COVID-19. Med Sci Monit 2020;26:e928996
11. Bach JR, Sinquee DM, Saporito LR, et al.: Efficacy of mechanical insufflation-exsufflation in extubating unweanable subjects with restrictive pulmonary disorders. Respir Care 2015;60:477–83
12. Bach JR, Gonçalves MR, Hamdani I, et al.: Extubation of patients with neuromuscular weakness: a new management paradigm. Chest 2010;137:1033–9
13. Bach JR, Saporito LR, Shah HR, et al.: Decanulation of patients with severe respiratory muscle insufficiency: efficacy of mechanical insufflation-exsufflation. J Rehabil Med 2014;46:1037–41
14. Bach JR; New Jersey Department of Health and Senior Services: Study of ventilator dependent patient needs in New Jersey. DOH Report. June 2015
15. Gonçalves MR, Honrado T, Winck JC, et al.: Effects of mechanical insufflation-exsufflation in preventing respiratory failure after extubation: a randomized controlled trial. Crit Care 2012;16:R48
16. Frithiof R, Elham R, Eva K, et al.: Critical illness polyneuropathy and myopathy in COVID-19 patients: a prospective observational intensive care unit cross-sectional cohort study. Research Square 2020. doi:10.21203/
17. Bach JR, Takyi SL: Physical medicine interventions to avoid acute respiratory failure and invasive airway tubes. PM R 2015;7:871–7
18. Bach JR, Alba AS: Noninvasive options for ventilatory support of the traumatic high level quadriplegic patient. Chest 1990;98:613–9
19. Bach JR: New approaches in the rehabilitation of the traumatic high level quadriplegic. Am J Phys Med Rehabil 1991;70:13–9
20. Agarwal A, Basmaji J, Muttalib F, et al.: High-flow nasal cannula for acute hypoxemic respiratory failure in patients with COVID-19: systematic reviews of effectiveness and its risks of aerosolization, dispersion, and infection transmission. Can J Anaesth 2020;67:1217–48
21. Vetrici MA, Mokmeli S, Bohm AR, et al.: Evaluation of adjunctive photobiomodulation (PBMT) for COVID-19 pneumonia via clinical status and pulmonary severity indices in a preliminary trial. J Inflamm Res 2021;14:965–79
22. Bach JR, Saporito LR: Criteria for extubation and tracheostomy tube removal for patients with ventilatory failure: a different approach to weaning. Chest 1996;110:1566–71
23. Bach JR, Giménez GC, Chiou M: Mechanical in-exsufflation-expiratory flows as indication for tracheostomy tube decannulation: case studies. Am J Phys Med Rehabil 2019;98:e18–20
24. Bach JR: Inappropriate weaning and late onset ventilatory failure of individuals with traumatic spinal cord injury. Paraplegia 1993;31:430–8
25. Bach JR: Pulmonary rehabilitation considerations for Duchenne muscular dystrophy: the prolongation of life by respiratory muscle aids. Crit Rev Phys Rehabil Med 3:239–69

COVID-19; Pulmonary Rehabilitation; Tracheostomy Decannulation; NVS and MIE Protocols

Supplemental Digital Content

Copyright © 2021 Wolters Kluwer Health, Inc. All rights reserved.