Long-term home mechanical ventilation using a noninvasive ventilator via tracheotomy in patients with myasthenia gravis: a case report and literature review

Abstract

Neuromuscular diseases (NMD) are indications for long-term home mechanical ventilation (HMV). Noninvasive ventilation is preferred to HMV. However, invasive mechanical ventilation (IMV) is more appropriate if the patient has uncontrollable airway secretions, the possibility of aspiration, failure to wean, or severe weakness of the respiratory muscles. But if the patient undergoes multiple intubation or tracheotomy, it will be more painful and unbearable. For some end-stage NMD patients who need long-term tracheostomy, HMV using noninvasive ventilator via tracheotomy may be a conservative care option. An 87-year-old male with myasthenia gravis underwent repeated IMV and failed to wean. We used a noninvasive ventilator connected to a tracheostomy tube for mechanical ventilation. One and a half years later, the patient weaned successfully. However, there was a lack of evidence-based medicine and standardized guidelines in such areas as indications, contraindications, and ventilator parameter setting. For the systematic review, a literature search was performed in PubMed, Embase, Cochrane, and CNKI (China National Knowledge Infrastructure) to identify reported cases of using noninvasive ventilator in patients undergoing tracheostomy. A total of 72 cases who performed ventilation via tracheotomy tube were identified. The main diagnoses included NMD, chronic obstructive pulmonary disease (COPD), pneumonia, and congenital central hypoventilation syndrome (CCHS). Indications included dysfunctional ventilatory weaning response (DVWR), apnea and cyanosis. Clinical outcome was as follows: 33 patients were weaned, and 24 patients underwent HMV. A total of 288 cases who performed ventilation through the mask after blocking the tracheostomy tube were identified. The primary diagnoses included COPD, NMD, thoracic restriction, spinal cord injured (SCI), and CCHS. Indications included DVWR, apnea and cyanosis, routine weaning. Clinical outcome was as follows: successful tracheostomy tube decannulations were performed in 254 patients and failed in 33 patients. So, in patients requiring HMV, selection of noninvasive ventilation (NIV) or IMV should be individualized. Tracheostomy preservation should be considered in some patients with advanced NMD if there is respiratory muscle weakness or the risk of aspiration. And attempts can be made to use a noninvasive ventilator because of its advantages of portability, ease of operation, and low cost. Noninvasive ventilators can be used in patients with tracheotomy, whether direct connection tracheotomy or mask ventilation after the tube is capped, especially in weaning and tracheostomy tube decannulation.

Introduction

In patients with myasthenia gravis (MG), respiratory muscle weakness can cause dyspnea and respiratory failure. In addition to the treatment of the primary disease, respiratory support is an important therapy. ¹ Neuromuscular diseases (NMD), including MG, are indications for long-term home mechanical ventilation (HMV). Among patients performing HMV, the most common diagnosis is NMD, which accounts for about 16.3% ² and MG accounts for about 2.6% in another report. ³ In patients receiving HMV, about 72% received noninvasive ventilation (NIV), and about 18% received invasive mechanical ventilation (IMV) via tracheotomy. ⁴ Therefore, NIV is preferred to HMV. However, IMV is more appropriate if the patient has uncontrollable airway secretions, the possibility of aspiration due to dysphagia, failure to wean, or severe weakness of the respiratory muscles. The advantage of NIV is that it does not require the creation of an artificial airway, thereby reducing the risk of tracheal injury and preserving language ability. The use of IMV has also been associated with increased discomfort, risk of infection, and costs of equipment and care. ⁵ On the other hand, invasive ventilation via tracheostomy is associated with a reduced risk of aspiration, may be more comfortable when required for more than 20 h per day, and may allow long-term survival for those patients choosing to continue lifelong mechanical ventilator support. ⁶ Therefore, the decision between IMV and NIV must be individualized. ⁵

Especially, for some elderly patients with advanced MG, both need long-term HMV, and may not have a long life expectancy. However, if the patient undergoes multiple intubation or tracheostomy to create an artificial airway, it will be more painful and intolerable. Moreover, the cost of ventilators for IMV is also much higher than that of noninvasive ventilators, ⁷ and the operation and maintenance are much more complicated. Therefore, for some long-term tracheostomy patients, is mechanical ventilation via tracheotomy with a noninvasive ventilator a cost-effective treatment option? We report a case of MG who underwent NIV, followed by repeated IMV, and failed weaning. We performed IMV using a noninvasive ventilator connected to a tracheostomy tube via a leaky valve. The patient was well tolerated and currently weaned.

Several early studies have demonstrated that noninvasive ventilator is safe and effective not only in providing noninvasive pressure support for patients with respiratory failure,^8,9 but also in providing invasive ventilation support for stable patients with tracheostomy tubes. ¹⁰ There have been previous reports of patients using noninvasive ventilators via tracheotomy,^11,12 and with the advances in ventilator technology, the functions of these ventilators have become more comprehensive and integrated. In recent years, ‘portable ventilators’ with both noninvasive and invasive modes are on the market, such as Stellar 150, Astral 150, LTV series, PB 540, iVent 101, and T-Bird series. However, there were few relevant randomized controlled trials (RCTs) that can be retrieved, and there was a lack of evidence-based medicine and standardized guidelines in such areas as indications, contraindications, ventilator parameter setting, indications for weaning, and timing of switching to traditional invasive ventilator. Despite the success of this case, we only referred to previous guidelines for NIV or IMV individually and lacked proven and reliable theoretical support. So, we reviewed some literatures on the use of noninvasive ventilators in patients undergoing tracheostomy, briefly analyzing the reasons, objectives, and outcomes. Limited by space, this article focuses on issues related to the selection of interfaces in HMV for NMD patients and puts forward some reflections and perspectives about noninvasive ventilators.

Case report

The patient was an 87-year-old male who was previously healthy. Two years ago, the patient developed dyspnea and dysphagia and was diagnosed with MG. In addition to the treatment of the primary illness, the patient purchased a noninvasive ventilator for home NIV in order to relieve his dyspnea. As the disease progressed, because of respiratory failure, he was treated with tracheotomy and IMV for 2 weeks in hospital and was discharged after remission. The tracheostomy and nasogastric tubes were retained because of dysphagia and risk of aspiration. During this period, when the patient presented with dyspnea, the tracheostomy tube was blocked, the cuff was deflated, and NIV was performed using an oronasal mask. But the patient had poor tolerability due to ventilation more than 15 h per day and worried about the risk of facial pressure ulcers. Three months later, the patient was hospitalized again for myasthenic crisis and was treated with IMV for 2 weeks. When the patient’s spontaneous breathing and hemodynamics were stable, the ventilator parameters were pressure support ventilation (PSV) mode, positive end-expiratory pressure (PEEP) 4 cmH₂O, and the oxygenation index (PaO₂/FiO₂) was 250 mmHg; the weaning process was started. He was able to tolerate spontaneous breathing test (SBT) ¹³ for 60 min, but failed at least five times when the ventilator was actually disconnected. Moreover, after each attempt of weaning, the patient was extremely nervous and irritable, accompanied by chest tightness. Blood gas analysis (BGA) showed PO₂ 69 mmHg, PCO₂ 48 mmHg. There was no improvement in symptoms after conventional oxygen therapy, and PSV was resumed.

Despite repeated psychological counseling and respiratory rehabilitation, the patient did not wean properly. Through analysis and tracheoscopy, we concluded that the patient did not have an inadequate treatment of primary disease or airway obstruction. Possible causes were mild dysfunctional ventilatory weaning response (DVWR) due to respiratory muscle weakness and fear and lack of self-confidence. ¹⁴

On the 23rd day of this hospitalization, we held a family meeting to discuss future treatment options, including (1) the patient had incurable primary disease and the possibility of another myasthenic crisis; (2) the patient had dysphagia, poor airway clearance, was not suitable for NIV, and had potential of IMV in the future. ¹⁵ Therefore, this patient was more suitable to retain tracheotomy.^16–18 (3) The patient acceptance of partial decline in quality of life, such as speech and daily activities; (4) the financial capacity of the family is limited; (5) the patient do not require high FiO₂ support, and at that time, portable ventilators for tracheotomy patients could not be procured locally for a variety of reasons. But the patient had a previously used noninvasive ventilator exactly. After discussion, we decided to use the noninvasive ventilator connected to a tracheostomy tube through a leak valve for ‘conservative’ mechanical ventilation. Written informed consent was obtained from the patient.

The ventilator type used was a bi-level positive airway pressure (BiPAP) noninvasive ventilator equipped with original humidifier. The connection method is shown in Figure 1. The tracheostomy tube type used was a suction-type tracheostomy tube; the cuff was deflated during ventilation and inflated when drinking and eating to prevent aspiration. Since the patient was conscious, he was instructed to close his mouth as much as possible during mechanical ventilation to reduce air leakage from the upper airway. Ventilator parameter setting was as follows: S/T mode, backup respiratory rate 16 breaths·per minute, inspiratory time percentage 30%, inspiratory positive airway pressure (IPAP) 18 cmH₂O, expiratory positive airway pressure (EPAP) 4 cmH₂O.

Figure 1.

Mechanical ventilation of a noninvasive ventilator through a leak valve (arrow) connected to a tracheostomy tube.

The patient was observed continuously after the beginning of mechanical ventilation, without symptoms such as chest tightness and dyspnea. In addition, although the noninvasive ventilator provided air leak compensation, our observations indicated that there was no leakage-related alarm on the ventilator during ventilation. The ventilation protocol was well synchronized, and the tidal volume was relatively constant. Vital signs were stable: respiratory rate: 15–18 breaths·per minute, tidal volume: 600 ml, and BGA: SpO₂ > 95%, PaO₂ > 80 mmHg, PaCO₂ < 45 mmHg. The duration of mechanical ventilation was 3–4 h in the morning, 3–4 h in the afternoon, and 8–10 h at night. After 2 weeks, the patient was evaluated. BGA and blood biochemical tests were normal. Ventilator pressure was downregulated to IPAP 16 cmH₂O and EPAP 3 cmH₂O. The patient was well tolerated. Thereafter, evaluations were performed every 2 weeks to gradually reduce the ventilatory pressure or shorten the duration of ventilation. When the pressure is adjusted to IPAP 10 cmH₂O and EPAP 3 cmH₂O, the ‘plateau’ is entered (if the pressure IPAP is less than 10 cmH₂O, the patient feels hypoventilation). The patient was discharged after 10 weeks. Caregivers were trained in airway cleaning techniques before discharge; ¹⁹ the patient’s family purchased a sputum aspirator, pulse oximeter, and a self-inflating bag to facilitate emergency use or monitor oxygen saturation. ²⁰ The patient performed HMV at home with good compliance and tolerance. With the passage of time, the ventilation time of the patient gradually decreased, until one and a half years later, the patient was able to continue without mechanical ventilation for 8 weeks, and finally weaned successfully. Ventilator parameters, clinical manifestations, and physiological indexes of the patient are shown in Table 1.

Table 1.

Ventilator parameters, clinical manifestations, and physiological indexes of the patient.

Index	Initial	2 weeks	4 weeks	6 weeks	8 weeks	10 weeks (until weaning)
IPAP, cmH2O	18	16	14	12	10	10
EPAP, cmH2O	4	3	3	3	3	3
TV, ml	600	580	550	520	500	500
MV, l/min	7.2	6.96	6.6	6.24	6	6
Air leakage, ml/min	0	0	0	0	0	0
RR, breaths·per minute	16	16	15	15	14	13
Backup RR, breaths·per minute	12	12	12	12	12	12
I:E	1:2.5	1:2.5	1:2.5	1:2.5	1:2.5	1:2.5
Rise time, ms	250	500	500	500	500	500
Daily ventilation time, h	15	13	10	10	9	0–9
SPO2,%	98	99	100	100	100	100
pH	7.45	7.45	7.44	7.45	7.44	7.42
PO2, mmHg	85	87	88	87	91	93
PCO2, mmHg	48	45	43	38	37	38
HCO3−, mmol/l	25	24	23	24	23	24
HR, beats per·minute	95	90	88	85	75	70
BP, mmHg	125/88	120/85	120/80	118/82	120/85	115/82
Spontaneous respiratory rate while offline, breaths·per minute	18	18	17	16	15	15
Dyspnea	Weak	No	No	No	No	No

BP, blood pressure; EPAP, expiratory positive airway pressure; HCO3−, bicarbonate; HR, heart rate; I:E, inspiratory:expiratory ratio; IPAP, inspiratory positive airway pressure; MV, minute ventilation; PCO₂, carbon dioxide partial pressure; PO₂, oxygen partial pressure; RR, respiratory rate; SPO₂, pulse oxygen saturation; TV, tidal volume.

During the whole period of HMV, there was no infection at the tracheotomy site, and there was a small amount of granuloma formation around the incision, considering common complications of tracheotomy. ²¹ Because of timely chest wall vibration and effective airway clearance, no ventilator-associated pneumonia occurred. And no ventilator-associated lung injury occurred throughout the treatment period. One ventilator was replaced due to malfunction of the key of the ventilator. There were no other malfunctions and abnormal alarms, and no major operational errors. The patient’s feedback during follow-up was that this type of ventilation was acceptable as conservative treatment. In addition, due to fear of intubation and unacceptability of mask NIV, the patient is willing to keep the tracheotomy until the end of life.

Literature review method

To explore the use of noninvasive ventilators in patients undergoing tracheostomy, we searched the complete articles of relevant references in PubMed, Embase, and Cochrane, including as far as possible all available cases of using noninvasive ventilator in patients with tracheotomy. Because there were few such documents through pre-search, we also searched CNKI (China National Knowledge Infrastructure) and selected those with English abstracts. The search terms were ‘noninvasive ventilator’ and ‘tracheotomy’. The search string was: (((‘noninvasive’[All Fields] OR ‘noninvasively’[All Fields] OR ‘noninvasiveness’[All Fields]) AND (‘ventilated’[All Fields] OR ‘ventilates’[All Fields] OR ‘ventilating’[All Fields] OR ‘ventilation’[MeSH Terms] OR ‘ventilation’[All Fields] OR ‘ventilate’[All Fields] OR ‘ventilations’[All Fields] OR ‘ventilator s’[All Fields] OR ‘ventilators, mechanical’[MeSH Terms] OR (‘ventilators’[All Fields] AND ‘mechanical’[All Fields]) OR ‘mechanical ventilators’[All Fields] OR ‘ventilator’[All Fields] OR ‘ventilators’[All Fields] OR ‘ventillation’[All Fields])) OR (‘non-invasive’[All Fields] AND (‘ventilated’[All Fields] OR ‘ventilates’[All Fields] OR ‘ventilating’[All Fields] OR ‘ventilation’[MeSH Terms] OR ‘ventilation’[All Fields] OR ‘ventilate’[All Fields] OR ‘ventilations’[All Fields] OR ‘ventilator s’[All Fields] OR ‘ventilators, mechanical’[MeSH Terms] OR (‘ventilators’[All Fields] AND ‘mechanical’[All Fields]) OR ‘mechanical ventilators’[All Fields] OR ‘ventilator’[All Fields] OR ‘ventilators’[All Fields] OR ‘ventillation’[All Fields])) OR ((‘biosynthesis’[MeSH Subheading] OR ‘biosynthesis’[All Fields] OR ‘bi’[All Fields]) AND (‘level’[All Fields] OR ‘levels’[All Fields]) AND (‘ventilated’[All Fields] OR ‘ventilates’[All Fields] OR ‘ventilating’[All Fields] OR ‘ventilation’[MeSH Terms] OR ‘ventilation’[All Fields] OR ‘ventilate’[All Fields] OR ‘ventilations’[All Fields] OR ‘ventilator s’[All Fields] OR ‘ventilators, mechanical’[MeSH Terms] OR (‘ventilators’[All Fields] AND ‘mechanical’[All Fields]) OR ‘mechanical ventilators’[All Fields] OR ‘ventilator’[All Fields] OR ‘ventilators’[All Fields] OR ‘ventillation’[All Fields]))) AND (‘tracheostomy’[MeSH Terms] OR ‘tracheostomy’[All Fields] OR ‘tracheostomies’ [All  Fields]  OR (‘tracheotomy’[MeSH Terms] OR ‘tracheotomy’[All Fields] OR ‘tracheotomies’[All Fields])). Inclusion criteria were as follows: (1) type of study: case report, case series, clinical study, clinical randomized control study; species: human; language: English and Chinese; published from January 1981 to November 2022; (2) subjects: patients with tracheotomy, regardless of primary disease, age, and gender; (3) intervention: mechanical ventilation using a noninvasive ventilator; (4) outcomes: clinical efficacy and duration of clinical treatment. Exclusion criteria were as follows: (1) reviews, guidelines; (2) duplicate articles; (3) full text not available; (4) no English-language abstracts in Chinese literature; (5) cases of using portable ventilators equipped with invasive ventilation mode or equipped with active exhalation port circuit; (6) incomplete treatment data. The data of the reported cases were summarized and tabulated. Retrieved: 27 November 2022.

Results

A total of 1250 English and 204 Chinese articles were retrieved. A total of 47 were selected for full article retrieval, and of these, 16 articles complying with the inclusion criteria were found. Of these, eight reported ventilation using noninvasive ventilator via tracheostomy tube. Eight other studies have reported ventilation through the mask after blocking the tracheostomy tube (Figure 2).

Figure 2.

Diagram of study selection.

A total of 72 cases who performed ventilation via tracheotomy tube were identified and tabulated (Table 2). Cases were mainly from Asia and the South Americas, with an age range of 4–90 years, 41 males and 31 females. The main diagnoses included NMD 31.9% (23/72), chronic obstructive pulmonary disease (COPD), 26.4% (19/72); pneumonia, 15.3% (11/72); and congenital central hypoventilation syndrome (CCHS), 2.8% (2/72). Indications included DVWR, 97.2% (70/72); apnea and cyanosis, 2.8% (2/72). The reasons for choosing this type of ventilation included the heavy economic burden and psychological pressure of patients receiving invasive ventilation. Initial purposes of using this method included weaning, 65.2% (47/72); respiratory support, 34.8% (25/72). Clinical outcome was as follows: 33 patients were weaned, and 24 patients underwent HMV. Fifteen patients died, mainly due to severe primary disease, terminal disease, malignant tumor, and inability to wean. No deaths due to ventilator failure were found. Among patients undergoing HMV, NMD accounted for 79% (19/24), CCHS for 8.3% (2/24), and COPD for 8.3% (2/24). The duration of ventilation was 8.5 days–4 years. All ventilators are single-circuit with a passive exhalation port (air leak valve).

Table 2.

Reported cases of ventilation using a noninvasive ventilator via tracheotomy.

References	Year	Country	Language	N	Mean age (years)	Sex	Diagnosis	Indications	Objective	Considering resource constraints or heavy economic burden	Mean duration of NIV via tracheotomy	Clinical outcome	Ventilator circuit	Tracheostomy connection
Ye et al. 22	2019	China	Chinese	2	54	2 M 1 F	1 pneumonia, 1 COPD	DVWR	Respiratory assist	Yes	Not supplied	2 HMV	Single-circuit	Passive air leak valve
Preutthipan et al. 11	2015	Thailand	English	2	4.5	2 F	2 CCHS	Apnea and cyanosis	Respiratory assist	Yes	4 y	2 HMV	Single-circuit	Passive air leak valve
Ibrahim et al. 12	2012	Brazil	English	26	73	11 M 15 F	7 COPD,4 ALS, 1 vascular disease, 2 multiple myeloma, 3 PO, 1 Parkinson, 1 pneumonia, 1 porphyria,1 CKF,3 stroke, 1 Alzheimer, 1 muscular dystrophy	DVWR	Weaning	Yes	53.5 d	14 discharge,12 death	Single-circuit	Not supplied
Ma et al. 23	2012	China	Chinese	3	66.3	2 M 1 F	1 COPD, 1 ALS, 1 intracerebral hemorrhage	DVWR	Respiratory assist	Yes	Not supplied	3 HMV	Single-circuit	Passive air leak valve
Li et al. 24	2011	China	Chinese	10	72.7	6 M 4 F	4 COPD, 4 pneumonia, 2 septic shock	DVWR	Weaning	Yes	Not supplied	8 discharge, 2 death	Single-circuit	Passive air leak valve
Shan et al. 25	2011	China	Chinese	2	63	1 M 1 F	1 ALS, 1 lung cancer	DVWR	Respiratory assist	Yes	Not supplied	1 HMV,1 death	Single-circuit	Passive air leak valve
Xiang et al. 26	2009	China	Chinese	16	59	12 M 4F	16 ALS	DVWR	Respiratory assist	Yes	39 m	16 HMV	Single-circuit	Passive air leak valve
Ma et al. 27	2008	China	Chinese	11	77	8 M 3 F	5 pneumonia, 6 COPD	DVWR	Weaning	Yes	23.5 d	11 discharge	Single-circuit	Passive air leak valve

ALS, amyotrophic lateral sclerosis; ARDS, acute respiratory distress syndrome; CCHS, congenital central hypoventilation syndrome; CKF, chronic kidney failure; COPD, chronic obstructive pulmonary disease; d, day; DVWR, dysfunctional ventilatory weaning response; F, female; HMV, home mechanical ventilation; M, male; m, month; NIV, noninvasive ventilation; PO, postoperative following major surgery; y, year.

A total of 288 cases who performed ventilation through the mask after blocking the tracheostomy tube were identified (Table 3). Cases were from Asia, Europe, and the North Americas, with an age range of 16–73 years. Diagnostic data for 66 patients in two articles were not specified. In other 222 patients, the primary diagnoses included COPD, 33.3% (74/222); NMD, 15.3% (34/222); thoracic restriction, 10.4% (23/222); spinal cord injured (SCI), 9.9% (22/222); postoperative, 4.9% (11/222); and CCHS, 2.25% (5/222). Indications for using this method included DVWR, 89.2% (257/288); apnea and cyanosis, 1.7% (5/288); routine weaning, 9.0% (26/288). The reasons for choosing this type of ventilation included noninvasive ventilator provides effective respiratory support; clearance of respiratory secretions can be accomplished by retained tracheotomy; if NIV fails, the tracheotomy can be reopened immediately and easily. Inclusion criteria for patients choosing this ventilatory modality were mostly based on passing the spontaneous breathing trial (SBT)^13,28–30 of conventional mechanical ventilation. Ventilation using this method success with clinical stability was characterized by (1) no fever (temperature <38°C); (2) stable hemodynamics (heart rate < 120 breaths·per minute, mean arterial blood pressure >100 mmHg), and blood gases (SaO₂ > 90%, PCO₂ < 60 mmHg, pH > 7.35 during spontaneous breathing); (3) conscious and cooperative patient; (4) effective cough and no indication for bronchoscopy to remove secretion; ²⁸ (5) the IPAP was kept at no more than 25 cmH₂O. The EPAP was kept at 4 cmH₂O. ²⁹ At this time, attempts are made to decannulate or wean. Clinical outcome was as follows: successful tracheostomy tube decannulations were performed in 254 (88.1%) patients and failed in 33 (11.9%) patients. One patient died due to recurrence of pneumonia. Among successful decannulation patients, 158 (62.2%) patients continued sequential NIV. No deaths due to ventilator failure were found.

Table 3.

Reported cases of ventilation using a noninvasive ventilator in patients with capped tracheotomy tube.

References	Year	Country	Language	N	Mean age (years)	Sex	Diagnosis	Indications	Objective	Patient state at initiation of NIV or initial NIV criteria	Mean duration of NIV after tube is capped	Clinical outcome	Type of ventilator	Interface
Bach et al. 31	1996	USA	English	37	42.9 ± 18.4	Not supplied	22 SCI, 11 NMD, 2 GBS, 1 OHS, 1 partial lung resection	DVWR	Tracheostomy tube removal	PaO2 > 60 mm Hg, SaO2 > 92%	3–9 d	32 succeeded(17 sequencing NIV) 5 failed	MI-E or portable volume ventilator	Mouthpiece or nasal
Schönhofer et al. 28	2002	Germany	English	114	55.8	Not supplied	39 COPD, 23 thoracic restriction, 20 NMD, 14 with other causes	DVWR	Weaning	in COPD: PCO2>50 mmHg; in others: PCO2 >45 mmHg	Not supplied	96 discharged patients continued to use NIV at night	Not supplied	Not supplied
Vagiakis et al. 32	2010	Greece	English	1	16	F	CCHS	Apnea and cyanosis	Tracheostomy tube removal	PaCO2 35 mmHg, PaO2 102 mmHg	Sustained	Sequencing long-term NIV	BiPAP	Not supplied
Duan et al. 29	2012	China	English	15	73 ± 7	10 M 5 F	7 Exacerbation of chronic pulmonary disorders, 5 PO, 3 Pneumonia	DVWR	Weaning	PaO2 ⩾ 60 mm Hg, FiO2 ⩽ 0.5, PaO2/FiO2 ⩾ 150, PEEP ⩽ 5 cmH2O	9 d ±	14 weaned 1 dead	BiPAP	Facemask
Pu et al. 33	2015	China	English	26	70.82 ± 7.02	18 M 8 F	COPD, cerebral vascular accident, ARDS, pneumonia a	IMV via tracheostomy	Weaning	PaO2 79.61 ± 8.86 mmHg, PaCO2 51.08 ± 7.85 mmHg	3.4 ± 0.23 d	21 succeeded 5 failed	Not supplied	Nasal or full-face mask
Sancho et al. 30	2016	Spain	English	40	64.74 ± 13.78	21 M 19 F	Hypertension, diabetes mellitus, congestive heart failure, COPD, pneumonia, PO a	DVWR	Weaning	PaO2 94.87 ± 27.76 mmHg PaCO2 50.32 ± 10.33 mmHg	Sustained	40 weaned and sequencing long-term NIV	Not supplied	Nasal or oronasal mask
Paglietti et al. 34	2019	Italy	English	4	10.25	1 M 3 F	4 CCHS	Apnea and cyanosis	Tracheostomy tube removal	PaCO2 ⩽ 45 mmHg, SaO2 ⩾ 96%	Sustained	4 succeeded and sequencing long-term NIV	Not supplied	Nasal or oronasal mask
Ceriana et al. 35	2019	Italy	English	51	68.2 ± 9.9	24 M 22 F	28 COPD, 6 bronchiectasis, 3 NMD, 3 ARDS, 6 PO	DVWR	Weaning	SaO2 ⩾ 90%	7.2 ± 2.3 d	46 weaned 5 failed	Bi-level ventilator	NIV mask

ARDS, acute respiratory distress syndrome; BiPAP, bi-level positive airway pressure; CCHS, congenital central hypoventilation syndrome; COPD, chronic obstructive pulmonary disease; d, day; DVWR, dysfunctional ventilatory weaning response; F, female; FiO₂, fraction of inspired oxygen; GBS, Guillain-Barre syndrome; h, hour; IMV, invasive mechanical ventilation; M, male; m, month; MI-E, mechanical insufflation-exsufflation; NIV, noninvasive ventilation; NMD, neuromuscular diseases; OHS, obesity hypoventilation syndrome; PaCO₂, partial pressure of carbon dioxide; PEEP, positive end-expiratory pressure; PaO₂, oxygen partial pressure; PO, postoperative following major surgery; SaO₂, oxygen saturation; SCI, spinal cord injured; y, year.

^aThe number of cases is not supplied.

Discussion

Type and advantage of noninvasive ventilators

There is no consensus in the literature regarding the terminology used to categorize devices for NIV support. Some scholars have classified these ventilators into three categories: bi-level ventilators, intermediate ventilators, and critical care ventilators. ³⁶ This classification is more suited to the clinical applications of respiratory and critical care medicine. Bi-level ventilators have a single-limb configuration with exhalation port, whereas critical care ventilators have dual-limb circuits with an expiratory valve, and intermediate ventilators may have either configuration. The intermediate ventilator is also known as the subacute ventilator. ³⁷ With advances in ventilator technology, it is becoming increasingly difficult to clearly classify each type of device. Originally designed specifically for NIV, the bi-level ventilators now have an invasive mode of positive pressure delivery. In contrast, the critical care ventilators originally designed for IMV now have a noninvasive positive delivery mode. ³⁶ Nowadays, devices specifically designed for NIV and critical care ventilators with NIV modes offer advantages over their predecessors. Moreover, bi-level ventilators demonstrated comparable performance capabilities with regard to trigger, pressurization rate, inspiratory work, cycling, and response to ventilatory demands when compared with critical care ventilators. ³⁸ In particular, the bi-level ventilator is also superior to the critical care ventilator in the compensation of leakage. ³⁹ NIV algorithms can help to reduce some asynchronies caused by leaks, but these algorithms are not necessarily able to eliminate them completely. This may be due to differences in the manufacturer’s approach to improving the patient-ventilator asynchrony. Nevertheless, studies have shown that some noninvasive ventilators are better synchronized than critical care ventilators because of algorithmic problems, and that some critical care ventilators may not work in the presence of large air leaks. ⁴⁰ In our case, the patient’s entire airway is open, including deflating the cuff during ventilation; when invasive ventilation was performed using a bi-level ventilator, the synchrony of the patient-ventilator was good, and there was no leak alarm. These may be related to the optimization of leak compensation algorithms in noninvasive ventilators.

Circuits of noninvasive ventilators

Currently, the circuits used in noninvasive ventilators include single-limb circuit with a passive exhalation port (some manufacturers call it ‘the leak valve’), active exhalation port in a dual-limb circuit, active exhalation port in a coaxial circuit, active exhalation port in single-limb circuit. ⁴¹ For single-limb circuit, the inspiratory pathways share the same limb with the expiratory pathways, so there is a risk of re-inhaling CO₂. To avoid this, the ventilator delivers fresh air continuously at high enough speeds during expiration to flush the dead spaces of the circuit. And airflow can be adjusted to compensate for unexpected leaks during the provision of preset pressure. In addition, the single-limb circuit involves fewer tube routes, is easier for caregivers to operate, and is less costly compared with the dual-limb circuit. ⁴² A study has shown that passive exhalation ports have been shown to be effective in reducing CO₂ re-breathing during long-term ventilation through tracheostomy and have similar effects on patient symptoms and pressure on caregivers to conventional active expiratory valves. ⁴³ Thus, a single-limb circuit with exhalation port can be considered for long-term patients undergoing tracheostomy. In addition, FiO₂ depends to a large extent on the type of circuit used with the active exhalation valve. The active exhalation valve circuit provides higher FiO₂ at lower oxygen flow rates than the passive exhalation valve circuit. ⁴¹ However, the patient in our case was not primarily hypoxic, and the patient’s dyspnea was relieved by the use of a single-limb circuit with a passive exhalation port providing some pressure support. It has been shown that during home IMV, the use of passive exhalation circuit is no less important than the use of active exhalation circuit in ensuring good respiratory gas exchange and maintaining safety. ⁴⁴ But the physiological mechanisms still have to be studied in the further.

Interfaces of noninvasive ventilators

The noninvasive ventilator interfaces include oronasal mask (also called full-face mask), nasal mask (including nasal pillow), total face mask,^45–47 and helmet,^48,49 each of which has its own advantages and disadvantages^5,26,28,33, we are familiar with these characteristics and will not dwell on them here. Another method, which is less used, is noninvasive ventilator connected to tracheotomy for ventilation. It is essentially the use of a noninvasive ventilator to perform the functions of IMV. Hill ⁵⁰ has asserted that noninvasive ventilators can be used for PSV- and PEEP-related ventilation in patients with tracheostomy. Ibrahim et al. ¹² proposed that the use of a noninvasive portable ventilator connected to tracheostomy tube may be an alternative for patients who cannot wean and discharge from hospital. In addition, from the economic aspect, there are also scholars who recognize this method. ⁵¹ But the relatively conservative view is that, because of the many aspects involved, this choice should be made after discussion by a multidisciplinary team, ⁵² and that NIV should take precedence over IMV whenever possible to avoid the risk of ventilator and catheter-related complications. ⁴⁶ Obviously, these concerns are primarily about the complications of tracheotomy. In fact, the results of our literature search suggest that there are indeed fewer cases of mechanical ventilation using this method. Nonetheless, the attempts of these patients have given rise to several insights, such as the replacement of the invasive ventilator for weaning, respiratory support, and the reduction of the patient’s financial burden. In addition, NIV with a mask after temporary blockage of the tracheotomy is a special method. This method combines the characteristics of both invasive ventilation and NIV. It can be helpful for ventilator-dependent patients in weaning and tracheostomy patients in decannulation.

Tracheostomy ventilation in NMD patients

Then, what thoughts can be brought about by this case of MG that we have reported? Previous guidelines suggest that HMV can be performed noninvasively or by tracheostomy. ⁵³ So, which NMD patients are more suitable for tracheotomy when performing HMV? Some researchers have confirmed that although some patients meet the indications for NIV, these patients are more suitable for invasive ventilation if they have excessive airway secretions or dysphagia or are unable to wean due to severe weakness of the respiratory muscles.^5,54 According to Muir et al., ¹⁶ although NIV is widely used, tracheostomy is still mandatory in some cases of respiratory distress, such as advanced stage of NMD with bulbar signs, and should be considered as a potential resource. Furthermore, although NIV is a first-line treatment for ventilatory support in NMD patients, the positive impact of NIV treatment on the prognosis of patients may be limited if there is a significant decline in throat function and it is difficult to maintain airway clearance. ¹⁷ Seneviratne et al. ¹⁵ believe that NIV can reverse respiratory muscle fatigue caused by myasthenic crisis, but when first tried, it is necessary to clarify to the patient whether invasive ventilation is an acceptable option if NIV does not provide adequate support. The above point of view shows that for NMD patients who need long-term HMV, if there is respiratory muscle weakness or bulbar signs or the possibility of aspiration, then tracheotomy is still an irreplaceable option, at least an important alternative. And, in fact, research in recent years has shown favorable survival outcomes for patients receiving tracheostomy ventilation and implies that patients can be discharged back home from hospital with appropriate packages of care. ⁵⁵ The cases we reported also support this view. However, it still needs to be confirmed by RCTs with a large sample.

Therefore, in patients requiring HMV, selection of NIV or IMV should be individualized. Tracheostomy preservation should be considered in some patients with advanced NMD if there is respiratory muscle weakness or the risk of aspiration. And attempts can be made to use a noninvasive ventilator because of its advantages of portability, ease of operation, and low cost. In addition, through our review, we have found that noninvasive ventilators can be used in patients with tracheotomy, whether direct connection tracheotomy or mask ventilation after the tube is capped, especially in weaning and tracheostomy tube decannulation. We look forward to more clinical trials and basic research on the use of noninvasive ventilators in patients undergoing tracheostomy.