Respiratory Implications of Pediatric Neuromuscular Disease

Manually Assisted Cough

A manually assisted cough or quad cough involves the application of chest-wall or abdominal thrusts by a caregiver in synchrony with a patient's cough to augment or replace expiratory muscle activity. The effectiveness of this therapy depends heavily on the skill of the caregiver to apply adequate pressure and also to be able to synchronize the thrust with the patient's cough effort.³⁸ Often, manually assisted cough is combined with breath-stacking to enhance cough peak flow over that achieved with manually assisted cough alone.^15,39,40 Whereas breath-stacking can be accomplished with glossopharyngeal breathing or a manual resuscitation bag with or without a one-way valve, it can also be performed with a bi-level pressure generator or mechanical ventilator. Use of a manual resuscitation bag typically requires the assistance of a second caregiver to deliver the breath while the first caregiver applies abdominal or thoracic expiratory thrusts. Among 52 adults with DMD, manual insufflation was equivalent to use of a mechanical ventilator in increasing cough peak flow over unassisted cough peak flow values.⁴¹ There are, however, limits to the effectiveness of manually assisted cough and manually assisted cough with breath-stacking. Chest-wall distortion and scoliosis can reduce the usefulness of manually assisted cough, and the frequency at which manually assisted cough is required during acute illnesses can cause caregiver fatigue.

Certain physiologic measurements can also predict which patients would benefit from manually assisted cough augmentation techniques. Measurements of cough peak flow along with P_Emax and vital capacity (VC) were collected in 179 adolescents and adults with NMD at baseline, after breath-stacking, and after breath-stacking plus manually assisted cough.¹⁵ There was an inverse relationship between improvement in cough peak flow with any cough augmentation technique and the baseline VC: although there was no VC upper limit of effectiveness for manually assisted cough, improvements with cough augmentation decreased linearly with increasing VC and P_Emax. Thus, application of manually assisted cough would be unlikely to enhance cough peak flow if the baseline VC were >1,910 mL or if the P_Emax were >34 cm H₂O. There was also a VC below which manually assisted cough without breath-stacking (558 mL) or manually assisted cough plus breath-stacking (304 mL) would be unlikely to produce an effective cough peak flow. Similarly, if the P_Emax were <14 cm H₂O, it would be improbable for a patient to achieve an effective cough peak flow with manually assisted cough techniques. In these latter situations, use of a mechanical in-exsufflator was recommended to avoid complications of secretion retention.¹⁵

Mechanical Insufflation-Exsufflation

If baseline lung function,¹⁵ severe chest-wall distortion, or an unstable chest wall precludes the use of manually assisted cough, MI-E can be used to produce an effective cough peak flow. MI-E involves passive lung expansion with the use of a positive-pressure insufflation followed rapidly by exsufflation with negative pressure to produce an expiratory flow velocity high enough to shear secretions from the airway wall and move them toward the airway opening, where they can be expectorated or suctioned.⁴² The MI-E device can be used via face mask or mouthpiece or attached directly to a tracheostomy or endotracheal tube.⁴³

Insufflation and exsufflation pressures and duration of pressure application are set independently according to patient comfort and effectiveness. In a review of MI-E use in 62 subjects with NMD ranging in age from 3 months to 28.6 y, Miske et al⁴³ found no correlation between the pressures used and either subject age or the type of underlying NMD. Using MI-E with a lung model set to simulate normal adult lung mechanics, Gómez-Merino et al⁴⁴ assessed the effects of altering insufflation and exsufflation pressures and insufflation to exsufflation times on resulting lung volumes, pressures, and flows. Whereas they noted that MI-E settings in practice must be individualized, they observed that in their lung model, effective cough peak flows >2.7 L/s did not occur with pressure spans <30 cm H₂O. Further, increasing insufflation times resulted in significantly greater exsufflation flows, whereas lengthening exsufflation time did not increase exsufflation flow. Subsequently, the same group varied compliance and resistance of the lung model and showed that alteration in respiratory system mechanics, including decreased compliance or increased resistance, would require increased insufflation and exsufflation pressures to achieve adequate pre-cough volumes and cough peak flows.⁴⁵ The authors cautioned that in such situations, patients would require higher MI-E pressures to achieve adequate cough peak flows. Underlining that point, Miske et al⁴³ reported that 39% of subjects were instructed to increase pressure settings during periods of illness to remove secretions more effectively.

Individualization of pressure and time settings for MI-E is important not only because of the effects of alteration of lung mechanics on cough peak flow, but also because of inherent differences in machines. A bench study that compared 5 devices for their accuracy in delivering set pressures and set times to a lung model disclosed discrepancies in most MI-E models tested.⁴⁶ The investigators also found inconsistencies in the pressures delivered and their duration between different devices of the same model of in-exsufflator. Furthermore, the performance of each of the models was affected differently by alterations in lung mechanics and imposed leaks. The authors cautioned that the different MI-E devices are not interchangeable and that settings should be targeted for each patient with the actual machine to be used.

Because recommendations for MI-E settings in infants and children who require tracheostomies are lacking and patient data are sparse, Striegl et al⁴⁷ used a lung model simulating normal lung mechanics of a 6- and 10-kg infant and assessed the effects of altering MI-E pressures and duration of insufflation and exsufflation delivered via 3.0-, 3.5-, and 4.0-mm inner diameter tracheostomy interfaces. With all tracheostomy tubes studied, an insufflation time of 1 s was required for pressure equilibration across the tube. Additionally, an inspiratory volume >70% of the calculated VC was achieved at all insufflation pressures with an insufflation time of 1 s. Maximum expiratory flow, a surrogate for cough peak flow, increased with increasing insufflation time, insufflation pressure, and exsufflation pressure, but not with increasing exsufflation time. Expiratory pressure had a greater effect on maximum expiratory flow than did inspiratory pressure, and the authors reasoned that secretion clearance could be enhanced with greater pressure differentials between insufflation and exsufflation by using asymmetric settings (eg, +20/−30 cm H₂O). They also emphasized that insufflation pressures as low as 20 cm H₂O achieved adequate pre-cough lung volumes. The authors cautioned, however, that primary safety concerns regarding the use of MI-E in infants include the risk of barotrauma, central airway collapse with high exsufflation pressures, and loss of FRC with prolonged exsufflation times.

The risk for barotrauma with MI-E use is real, but complications like pneumothorax are extremely rarely reported in adults,⁴⁸ and none have been reported in children. The author has, however, cared for a 16-y-old boy with DMD who experienced recurrent pneumothoraces with the use of MI-E and positive-pressure ventilation via tracheostomy. Rupture of a tympanic membrane was also described in a 19-y-old patient with DMD who used both MI-E and positive-pressure NIV.⁴⁹ Treatment included both temporary cessation of use of MI-E and a reduction in inflating pressures of his noninvasive ventilator. Dynamic airway collapse has been reported in adults with bulbar dysfunction during the exsufflation cycle of MI-E use,^22,50 and use of a coordinated abdominal thrust has been advocated as a means to avoid airway collapse during exsufflation.²² Using transnasal flexible laryngoscopy to study the laryngeal response to MI-E use in healthy young adults, Andersen et al⁵¹ found that application of a negative pressure during exsufflation produced muscular and/or reflex responses that caused hypopharyngeal constriction. These investigators also found that the epiglottis often retroflexed during insufflation, causing intermittent airway obstruction and potentially limiting the effectiveness of insufflation. This certainly would be a concern when using MI-E in a subject who also has significant laryngomalacia. Other reported complications of MI-E use, like cardiac dysrhythmias, aggravation of gastroesophageal reflux, hemoptysis, or abdominal distention and discomfort, are uncommon or have never been reported in children with NMDs.^42,43 The intra-abdominal pressure generated with MI-E use was 51 ± 32% of the set insufflation pressure among 13 children with NMDs, and lower than the measured gastric pressure during spontaneous cough in 1 of the 3 subjects who could also generate a spontaneous cough.⁵² These data suggest that MI-E use should be safe in the immediate postoperative period, even for children who require upper-abdominal surgery (ie, gastrostomy tube placement).

A Cochrane review⁵³ concluded that existing data are insufficient to recommend MI-E use for patients with NMD, principally because most studies report short-term effects of a single MI-E treatment, and there are not long-term studies examining survival, length of hospital stay, quality of life, or serious adverse effects related to MI-E use. However, there have been several small non-randomized trials that address some of these issues. Vianello et al⁵⁴ compared the short-term outcomes of 11 consecutive subjects with NMD admitted to the ICU during an acute lower-respiratory-tract infection and treated with MI-E in addition to conventional chest physiotherapy with outcomes of 16 historical matched controls who received chest physiotherapy alone. Treatment failure, defined as need for cricothyroid minitracheostomy or endotracheal intubation, was significantly less among the group receiving MI-E compared with controls. There were, however, no differences in duration of hospitalization, time spent receiving mechanical ventilation, or proportion of subjects who required bronchoscopically assisted removal of secretions between the 2 groups. Chatwin and Simonds⁵⁵ assessed the addition of MI-E to chest physiotherapy followed by manually assisted cough during an acute respiratory illness in 8 subjects 4–44 y old. The addition of MI-E to the airway clearance regimen significantly shortened the duration of airway clearance sessions, but subjects felt more fatigued after MI-E use than when no MI-E was employed, perhaps because of long exsufflation times.

Two studies from Western Australia surveyed families of children with NMD who used an MI-E device at home.^56,57 The first, involving 10 subjects 1.4–18.1 y old, also involved a retrospective chart review.⁵⁶ Among 7 of the subjects for whom there was adequate duration of observation before and after MI-E use, there was a significant reduction in number of days in the hospital at 6 and 12 months after institution of MI-E compared with the same time frame before its use. There was also a nearly statistically significant reduction in number of ICU days after MI-E use but no significant reduction in number of hospitalizations. Nevertheless, 9 of 10 parents reported that MI-E use enhanced or improved their child's quality of life and emphasized their perception that MI-E use often allowed the child to stay at home for treatments during acute respiratory illnesses. The second survey and database review involved 37 subjects 1–59 y of age (mean 19.8 y) who used an MI-E device for 0.1–4.0 y.⁵⁷ Of the subjects or parents, >90% were satisfied with home use of the MI-E, and 88% agreed that the device improved their/their child's quality of life. Home use of an MI-E device was significantly associated with fewer emergency room visits (relative risk = 1.76, 95% CI 1.10–2.84), but there were no statistically significant reductions in hospitalizations or length of hospital stay resulting from MI-E use during the observation period. Echoing the generally positive attitude of children and their families toward home MI-E use, Moran et al⁵⁸ interviewed 8 parents and 3 children with NMD who had used an MI-E at home for ≥3 y. They expressed feelings of improved ability to manage the child's life and to have fewer lifestyle disruptions because of their ability to intervene during crises without having to go to the hospital. There were minimal negative impacts on patients and families, mostly related to their viewing the use of the MI-E as an acknowledgment of progression of the child's disease. Most negative sentiments were offset by parents' sense of empowerment that the device gave them.

In summary, MI-E effectively increases cough peak flow and is especially helpful when other cough augmentation techniques are technically not practicable. Whether MI-E is superior to other techniques of airway clearance or not when outcomes like frequency of hospitalization, length of stay, and overall survival are considered remains to be answered. Additionally, recognition of patient characteristics that would identify who would have less favorable responses to MI-E use are needed to optimize airway clearance regimens for children with NMDs.

Intrapulmonary Percussive Ventilation

Intrapulmonary percussive ventilation (IPV) provides low-amplitude bursts of air at frequencies in the range of 50–550 cycles/min, with pressures of 5–35 cm H₂O superimposed on the patient's own breathing frequency. The percussions of gas are applied to the airways using a mouthpiece, face mask, or artificial airway (endotracheal or tracheostomy tube). They are delivered continuously through a sliding Venturi device called a Phasitron, powered with compressed gas at pressures of 20–40 psi. The oscillations cause airway walls to vibrate in synchrony, loosening secretions. The device not only provides a mechanical means for mobilization of secretions, but it also enables simultaneous delivery of medications like bronchodilators or hypertonic saline via entrainment through the Phasitron. The intrapulmonary percussive ventilator was developed in 1979⁵⁹ and received FDA clearance in 1993⁶⁰; its first use in children occurred in the early 1990s.^23,61,62

The mechanisms by which IPV clears the airways of secretions have not been well-studied, but the device is purported to dilate airways both mechanically via a constant distending pressure as well as through the positive-pressure oscillations, and pharmacologically when bronchodilators are administered. It must be cautioned, however, that drug delivery to the lung via IPV is inferior to that achieved by jet nebulization because IPV creates a smaller particle (about 0.2 μm vs 2 μm) to be nebulized and delivers the drug at much higher flows compared with jet nebulization.^63,64 On inspiration, the high-frequency oscillations cause pulmonary expansion and allow gas to fill distal lung units beyond accumulated airway secretions.⁶⁰ Additionally, like other high-frequency oscillatory techniques, IPV may augment mucus-air flow interactions.²⁵ By prolonging the inspiratory phase of the percussions, expiratory flow becomes greater than peak inspiratory flow, and an expiratory flow bias is created.^25,65 The asymmetry of air flow accompanying the oscillations is critical for mucus clearance: when artificial mucus was instilled in the airways of sheep, secretion clearance was enhanced only when expiratory flow exceeded inspiratory flow.⁶⁶ The effect was further enhanced by adding postural drainage to the maneuver.

Birnkrant et al²³ described the successful use of IPV in 4 patients who had persistent atelectasis or infiltrates unresponsive to conventional chest physiotherapy. Two children and an adult had neuromuscular disease, and a third child had recurrent aspiration associated with a repaired tracheoesophageal fistula. Three of the four experienced a dramatic response to the therapy, with resolution of hypoxemia within 48 h of starting the therapy after failing 5–9 d of chest physiotherapy with or without manually assisted cough. One boy with DMD experienced 2 episodes of third degree heart block associated with hypoxemia that was presumed to be the result of acute airway obstruction from mobilization of secretions from the peripheral to central airways. As a result, the authors cautioned that close observation of patients who are unable to clear secretions by coughing is required when using IPV.

The role of IPV in relieving atelectasis was evaluated retrospectively in hospitalized children and prospectively in a randomized, controlled trial comparing IPV with manual chest physiotherapy with postural drainage in intubated and mechanically ventilated children.⁶⁷ Children in the retrospective study ranged in age from 1 month to 15 y; the authors noted that 2 infants <3 kg experienced hypotension during IPV administration, so they restricted subsequent IPV use to children >3 kg. Those subjects in the prospective study who were treated with IPV experienced radiographic improvement of atelectasis, whereas those who received manual chest physiotherapy with postural drainage had no change in their chest radiographs. Additionally, the duration of treatment to resolution of the atelectasis was half as long in the IPV group compared with the chest physiotherapy/postural drainage group.

Two small trials of IPV use were conducted among subjects with NMD. Toussaint et al²⁵ examined short-term secretion clearance as determined by sputum weight in 8 young men with DMD and tracheostomies. The 3 subjects who did not produce large amounts of secretions at baseline experienced no significant increase in secretion production when using IPV in addition to forced expiratory technique and manually assisted cough (standard therapy) compared with standard therapy alone. The 5 young men who were mucus hypersecretors, producing >30 mL/d of mucus, increased secretion production by 69% when IPV was added to standard therapy compared with standard therapy alone. Reardon et al²⁴ compared IPV with incentive spirometry (IS) in preventing lower-respiratory tract infections in 18 children with NMD followed over a single respiratory virus season. None of the 9 subjects in the IPV group received antibiotics, whereas subjects in the IS group received 44 d of antibiotics. Similarly, 3 subjects in the IS group experienced 3 episodes of pneumonia or bacterial bronchitis, whereas none of those in the IPV group had a pulmonary infection. Furthermore, those who used IS required significantly more supplemental respiratory treatments than those who used IPV. One weakness of this study is that the comparison group used an airway clearance modality (IS) that requires inspiratory muscle strength to be effective.

High-Frequency Chest-Wall Compression

HFCWC most commonly is accomplished by an inflatable vest attached by hoses to an air pressure generator. The generator rapidly injects and withdraws small volumes of air into and out of the vest, inflating and deflating it to produce chest compressions at frequencies and pressures selected by the clinician. The frequencies typically used range between 5 and 25 Hz; the pressure is set in arbitrary units. No special positioning or breathing techniques are required during HFCWC therapy, making it attractive for children with neurological impairment or those unable to perform such techniques because of weakness.

HFCWC at frequencies >3 Hz cause cough-like shear forces on mucus and generate asymmetrical air-flow velocities that favor movement of secretions toward the mouth.^68,69 In addition, high-frequency oscillation of sputum in the range of 10 Hz causes a reduction in its viscosity.⁷⁰ Coughing clears the trachea and main bronchi of secretions, but it is not as effective in clearing secretions from peripheral airways.⁷¹ In contrast, HFCWC at 13 Hz has been shown in an animal model to clear radioactive tracer particles effectively from the periphery of the lung.⁷² As a result, HFCWC is considered a mucus-mobilization technique that is used in tandem with cough or other mucus extraction techniques to clear the airways of secretions.

The effectiveness of HFCWC has been demonstrated in children with cystic fibrosis,^73,74 but there are few studies demonstrating its effectiveness in children with NMD. Case reports support its use in critically ill infants and children with SMA, DMD, and cerebral palsy during acute illnesses.^75–78 In one 5-y-old with SMA II and ARDS following respiratory syncytial virus infection, HFCWC was used as the primary method of airway clearance and was well tolerated when IPV combined with MI-E caused episodes of hypoxemia from presumed alveolar de-recruitment.⁷⁷ Most case reports note how well-tolerated HFCWC is, but one 11-y-old with cerebral palsy and an ineffective cough developed acute hypoxemic respiratory failure after receiving a treatment with HFCWC, ostensibly from aspiration of secretions after they had been mobilized.⁷⁸ The authors cautioned that assistive coughing devices or techniques might be required in addition to use of HFCWC when patients have ineffective cough.

Small studies in children with a variety of neuromuscular or neurological diseases have compared health-care utilization before and after HFCWC use.^26–28,79 Seven subjects with quadriplegic cerebral palsy demonstrated a significant reduction in episodes of pneumonia and more effective airway suctioning, defined as suctioning attempts in which sputum was recovered, in the 12 months after use of HFCWC compared with the 12 months before.²⁸ Furthermore, none of the subjects experienced adverse events resulting from HFCWC use. Similarly, 15 children and young adults with familial dysautonomia experienced improved health outcomes in the year following HFCWC use compared with the year before, as reflected in fewer hospitalizations, episodes of pneumonia, days with antibiotics, courses of antibiotics, physician visits, and days of school absence.²⁷ Additionally, resting oxyhemoglobin saturation increased from 94% (interquartile range 89–96%) at baseline to 98% (interquartile range 98–98%) at 12 months (P = .004). Another observational study involving 22 neurologically impaired children also demonstrated a significant reduction in hospitalization rates from 45% before HFCWC use to 36 and 13% in the 1 and 2 y after its use.²⁶ A randomized, controlled trial involving 23 children with either cerebral palsy or NMD compared clinical outcomes, including duration of acute infections, adverse events, polysomnogram results, chest radiograph findings, and body mass index in a group who were treated with standard chest physiotherapy and another group who received HFCWC over 5 months.⁷⁹ There were no therapy-related adverse events. Although there was no difference between groups in oral or intravenous antibiotic use, there was a trend toward fewer hospitalizations for intravenous antibiotic therapy in the HFCWC group. That group also had higher maximum oxygen saturation, but there were no differences between groups in nocturnal saturations. There were also no differences either between groups or by the end of therapy in chest radiograph scores or body mass index. Notably, a post-study caregiver survey disclosed significantly better adherence to the prescribed thrice-daily therapy in the HFCWC group, with caregivers citing difficulty with the time required and positioning involved to provide chest physiotherapy.

Recently, Lechtzin et al⁸⁰ compared health-care claims of patients with NMD from 2 large commercial insurance claim databases before and after use of HFCWC. Of the 426 subjects in the study, 43.9% were 0–18 y old. For the group, total medical costs per member per month decreased 18.6% after HFCWC use (P = .002), in-patient admission costs decreased by 41.7% (P = .001), and costs related to treatment of pneumonia decreased by 18.1% (P = .02). Although the data did not allow the authors to determine either frequency of HFCWC use or adherence to therapy, they speculated that the reductions in health-care utilization and claim costs were the result of a proactive airway clearance regimen that included HFCWC.

From the above data, it would seem reasonable to combine a mucus-mobilization device with mucus extraction, either by manually assisted cough or MI-E. It is often difficult, however, to get more than one piece of airway clearance equipment per patient approved by insurance companies, and the cost of the equipment can be greater than can be borne by a family (Table 1). Thus, additional data are critically needed to determine cost-effective strategies for airway clearance therapies for children with NMDs.

Medications That Alter Mucociliary or Cough Clearance

The characteristics of mucus and airway-lining fluid greatly affect how successfully either mucociliary clearance or cough clears the respiratory tract of secretions. Cough clearance is optimized when mucus has high viscosity and low tenacity, whereas mucus variables that favor effective mucociliary clearance are opposite to those that favor cough clearance.³¹ Thus, the more elastic the mucus, the better mucociliary clearance will be; the more viscous the mucus, the better cough clearance will be.

Agents that alter the properties of mucus have been divided into 4 categories (Table 2). Although several medications can alter the physical properties of mucus or the airway-lining fluid, all pharmacologic therapies aimed at enhancing airway clearance, including the use of bronchodilators, mucolytics like dornase alfa, and hyperosmolar agents like hypertonic saline, are considered off-label for children with NMD.^81,82 Given the lack of clinical trials outside of those involving children with cystic fibrosis or bronchiolitis, such therapies are not recommended in most clinical practice guidelines for patients with NMD.^13,81–83 Nevertheless, there are case reports on the use of dornase alfa for patients with neuromuscular weakness who developed atelectasis unresponsive to conventional therapies.^29,84

Use of inhaled dornase alfa (DNase) has been shown to improve and slow the decline of lung function in children with cystic fibrosis.^85,86 Patients with cystic fibrosis have purulent sputum with a large burden of necrotic neutrophils that release DNA into the sputum. DNase cleaves extracellular DNA and separates it from proteins. This reduces the concentration of highly polymerized DNA in sputum and allows endogenous proteolytic enzymes to lyse proteins, thereby decreasing the viscosity of sputum.^31,87 In contrast, adults with non-cystic fibrosis bronchiectasis who were randomized to receive DNase in a large controlled trial had worse outcomes compared with those who received placebo.⁸⁸ The reasons for the poorer results remain unclear, but such findings highlight the point that just because a particular intervention works in one population with specific characteristics, such results may not be generalizable to all populations who may have different characteristics. There are several case series and small trials in which DNase has been used to treat infants with atelectasis.³² Although the only drug-related adverse events reported were transient episodes of hypoxemia after the drug was instilled through an endotracheal tube, the drug still must be used with caution in infants and children with neuromuscular weakness given the lack of data regarding its safety and effectiveness in that population.

In recent years, inhaled hypertonic saline has been used in older children with cystic fibrosis to help preserve lung function and reduce the frequency of pulmonary exacerbations⁸⁹ and in infants with viral bronchiolitis to ameliorate lower airway obstruction and shorten the duration of hospitalization.⁹⁰ There are no published case reports or series on the use of inhaled hypertonic saline in patients with NMD. Perhaps because of pediatricians' familiarity with using nebulized hypertonic saline in children with cystic fibrosis and in infants with bronchiolitis, anecdotally it has been increasingly prescribed during acute respiratory illnesses in patients with NMD.

Hypertonic saline probably has several different mechanisms of action.^31,91 It is routinely used as an expectorant to aid patients in producing sputum for examination.⁹² This is probably the result of its osmotic effect on increasing the depth of the airway surface fluid layer, which also improves mucociliary clearance.⁹³ It may also have mucolytic properties, by disrupting ionic bonds within the mucus gel, thereby reducing entanglements and cross-linking.⁹⁴ Like DNase, hypertonic saline can also dissociate white cell DNA from mucoproteins, making the latter more amenable to digestion by proteolytic enzymes.⁹¹ Inhaled hypertonic saline may also have anti-inflammatory properties, including an increase in the levels of 2 compounds that protect against oxidant injury, glutathione and thiocyanate, in the airway surface liquid⁹⁵ and possibly a reduction in interleukin-8, as demonstrated in bronchoalveolar lavage fluid of patients with cystic fibrosis.⁹⁶

In contrast, inhaled hypertonic saline may have some deleterious effects. In a ferret model, exposure of airway epithelium to hypertonic saline resulted in a dose-dependent increase in mucus production.⁹⁷ The authors cautioned that hyperosmolar solutions should be used with caution in patients with mucus hypersecretion and impaired mucus clearance. The use of inhaled hypertonic saline in adults with COPD before exercise 3 times/week resulted in less improvement in a 6-min walk test at the end of 8 weeks compared with a control group who received normal saline before exercising.⁹⁸ In addition, 12% of the treatment group experienced cough or bronchospasm as adverse events, whereas no adverse events were reported in the control group. A recent prospective study in intubated and mechanically ventilated children demonstrated no reduction in the duration of mechanical ventilation in a group who received 3% inhaled hypertonic saline compared with those who received normal saline aerosols.⁹⁹ Additionally, there was no improvement in lung mechanics or chest radiographic appearance following hypertonic saline administration, but 2 subjects had to be removed from the study because of hypoxemia that developed after hypertonic saline administration. Radioisotope studies of mucociliary clearance in healthy subjects treated with 3% inhaled hypertonic saline showed an initial acceleration of mucociliary clearance over the first 30 min after hypertonic saline inhalation but a slowing in mucociliary clearance at 3–6 h.¹⁰⁰ The authors speculated that the reduction in clearance was the result of depletion of mucus in the central airways. Once again, these data collectively suggest that caution should be used when considering hypertonic saline therapy for patients with NMD. Studies are required to determine under what conditions inhaled hypertonic saline therapy is appropriate to use and will be most effective for children with NMD, as well as to identify the optimal dosing regimen.

Inhaled anticholinergic drugs like ipratropium bromide, tiotropium, atropine, and glycopyrrolate block mucus hypersecretion that is triggered by inflammatory or pharmacologic stimulation of M3 muscarinic receptors, but they do not reduce the normal production of mucus or increase its viscosity.^101,102 In hypersecretory conditions like COPD, their use has been shown to decrease sputum volume.¹⁰³ No such effect has been studied in patients with NMD. One of the adverse effects of these drugs is oral dryness, and they are frequently used enterally to treat sialorrhea.^104,105 They are less effective anti-sialogogues when used by spraying the inhalational form of the drug into the mouth.¹⁰⁶ Although excessive thickening of secretions is a common concern when using these medications enterally, it is not a frequently reported adverse effect.^105,107

Epidemiology of Sleep-Disordered Breathing

The constellation of problems that can occur during sleep in children with NMD include hypoventilation, hypoxemia, central and obstructive hypopneas and apneas, frequent arousals, sleep fragmentation, decreased sleep efficiency, and even seizures.^{1,110,118,119} Collectively, these symptoms comprise the term sleep-disordered breathing (SDB). The prevalence of SDB among patients with neuromuscular weakness has not been well studied, but one report found a frequency of SDB of 42% among 60 unselected adults and children with a variety of NMDs.¹¹⁷ That center, however, was at a high altitude (1,500 m), and only 10 of the subjects were children, so the generalizability of information to a pediatric population at sea level must be questioned. Nevertheless, studies involving smaller numbers of children with different NMDs estimate the prevalence of SDB to be as high as 70%.^109,120,121 In addition, one study in boys with DMD showed a prevalence of obstructive sleep apnea of 30%, a rate 10 times higher than the prevalence reported in the general pediatric population.¹²² In the absence of large, longitudinal population-based studies, the prevalence of SDB in children with NMD at different centers will vary, depending on the age at which the children are studied and the type of NMD being studied.

These 2 factors, patient age and natural history of the underlying disease, can also affect the type of respiratory disturbance identified. Suresh et al¹²² studied 34 subjects with DMD age 1–15 y referred to a sleep program for respiratory assessment. Ten had polysomnographic evidence of obstructive sleep apnea at 1–14 y of age (median 8 y). Over the 5-y observation period, 11 of the 34 subjects were found to have hypoventilation during sleep, and 9 began NIV at 11–15 y (median 13 y). The authors speculated that the observed bimodal distribution of SDB reflected age- and disease-related differences in muscle dysfunction: younger subjects with DMD who had better preservation of diaphragm function presented with obstructive sleep apnea. As respiratory muscle weakness ensued with increasing age, however, the boys might not have been able to generate enough negative intrathoracic pressure to develop upper-airway obstruction and thus presented more with sleep hypoventilation.¹²³ This interpretation is consistent with the finding that the type of SDB encountered reflects the distribution of respiratory muscle involvement.¹¹⁰

As patients develop respiratory muscle weakness significant enough to affect gas exchange, they will compensate for hypoventilation with an arousal response. The arousal limits the extent of hypoxemia and hypercapnia by changing the sleep state, recruiting respiratory and upper-airway muscular activity, and depth of ventilation.¹ Although arousals constitute a protective mechanism against hypoventilation, they cause sleep disruption, poor sleep quality, sleep fragmentation, and reduced sleep time. Sequelae of frequent arousals include daytime somnolence and fatigue as well as poor school performance.¹²⁴ With persistent hypoventilation and its accompanying sleep fragmentation and deprivation, ventilatory chemoreceptor responses become blunted, resulting in fewer arousals, longer periods of REM sleep, and longer periods of hypoventilation with hypoxemia.^1,110 Renal compensation for persistent hypercapnia further dampens chemoreceptor sensitivity, causing even more severe hypoventilation and a blunted respiratory drive. Hypercapnia, initially confined to REM sleep, extends into non-REM sleep and eventually results in diurnal hypercapnia.^125–127 Such a progression is life-threatening: mean survival in patients with DMD once diurnal hypercapnia developed was 9.7 months in the absence of ventilatory assistance.¹²⁸

This sequence of events leading to progressive hypoventilation is common among patients with NMD. There are some conditions, however, where alterations in control of breathing are a part of the primary condition and not the result of progressive muscle weakness. Children with myelomeningocele and Chiari malformation can demonstrate periodic breathing, hypoxemia, and abnormal responsiveness to CO₂ challenge immediately after birth.¹²⁹ A primary disorder in ventilatory control has also been reported in children and adults with myotonic dystrophy, perhaps related to neuronal loss and gliosis in the reticular activating system and brainstem.^130,131 Similarly, patients who experienced bulbar poliomyelitis can demonstrate SDB and abnormal responses to hypoxic and hypercapnic challenges years after the primary infection.¹³²

Evaluation of Sleep-Disordered Breathing

Children with NMD who have SDB can have subtle symptoms that may be difficult to distinguish as arising from something other than the underlying neuromuscular disease. For instance, children may complain of fatigue or excessive daytime sleepiness. Fatigue can result from progression of skeletal muscle weakness and not be at all associated with disorders of breathing. Excessive daytime somnolence is a hallmark of myotonic dystrophy and is not directly related to the presence of SDB.¹³⁰ Symptoms and signs frequently associated with SDB are listed in Table 3.^{1,108,110,114,118,121,125} These often appear insidiously and should be actively sought by the clinician. Although they offer clues to the presence of SDB, they are not specific. Furthermore, there is poor correlation between results of symptom questionnaires in patients with NMD and polysomnographic findings of SDB.^{108,117,119,122,133–135}

Because symptoms and signs alone are unreliable in detecting SDB and polysomnography is both cumbersome and often limited in availability, investigators have sought various daytime tests that might predict the presence of SDB. There are, however, no simple relationships between awake measurements of pulmonary function or respiratory muscle strength and SDB.¹¹⁰ For instance, a fall in FVC ≥25% associated with a change from seated (upright) to supine position is a good indicator of diaphragmatic weakness.¹³⁶ Although diaphragm weakness is a risk factor for hypoventilation during REM sleep, no correlation between a fall in the supine FVC and the presence of SDB could be found in 21 subjects with DMD.¹²⁰ Similarly, several investigators have shown little or no correlation of pulmonary function studies like the FVC or FEV₁ and SDB events.^{119,120,122,137}

Others, however, have shown statistically significant relationships between pulmonary function tests and nocturnal breathing events.^{121,125,135,138} Katz et al¹³⁵ evaluated 46 children with progressive NMD and found a prevalence of nocturnal hypoventilation of 15%, although the group was not selected for concerns about SDB. The investigators found a significant relationship between nocturnal hypoventilation and an FVC <70% of predicted (sensitivity 71.4, specificity 64.1) or FEV₁ <65% of predicted (sensitivity 71.4, specificity 79.5). The authors noted that their cutoff values in otherwise asymptomatic children were higher than those reported in adults who were symptomatic at the time of testing, many of whom had diurnal hypercapnia.¹³⁸ Mellies et al¹²¹ evaluated 49 children 5–18 y old with a variety of NMDs for SDB. Lung functions, including an inspiratory vital capacity (IVC), FVC, FEV₁, and maximum inspiratory pressure were measured and compared with polysomnogram results. The investigators found significant correlations between both the IVC and maximum inspiratory pressure and SDB events. Furthermore, those children who had an IVC <60% of predicted were noted to have the onset of SDB without hypercapnia (sensitivity 97%, specificity 87%); those with an IVC <40% of predicted were likely to have hypercapnic hypoventilation (sensitivity 96%, specificity 88%). The same investigators studied a group of 42 children and adults 14–63 y old with primary myopathies and found similar correlations between values of percent-of-predicted IVC and SDB.¹²⁵ It may be that because the IVC primarily reflects inspiratory muscle function, whereas the FVC is affected by both inspiratory and expiratory muscle function, the IVC will prove to be a better predictor of SDB resulting from diaphragm dysfunction. Similar comparisons using the IVC will need to be made in other centers to confirm its utility in predicting nocturnal breathing events.

With these relationships in mind, a panel of 10 experts developed consensus recommendations for the respiratory assessment of patients with DMD.¹³⁹ The experts strongly recommended assessment of gas exchange during sleep for any patients with signs or symptoms of hypoventilation. They noted that the assessment of gas exchange during sleep should be strongly considered in any patients with a baseline FVC <40% of predicted, an awake baseline end-tidal partial pressure of carbon dioxide (P_ETCO2) >45 mm Hg, and/or an awake baseline S_pO2 <95%. In teenagers and older patients, similar assessments should be made if the FVC falls below 1.25 L.¹³⁹

As noted, history and physical examination findings are neither sensitive nor specific in identifying patients with SDB. Furthermore, daytime tests of lung function do not reliably identify patients with SDB in the absence of diurnal hypercapnia. Thus, screening tests are required to detect SDB in children with neuromuscular weakness before the advanced finding of daytime hypercapnia develops. Guidelines for the care of children with NMDs recommend at least annual testing for patients who become non-ambulant because of weakness or for those who never walked.^13,14,140 Other statements acknowledge the need for testing but do not specify intervals.^18,141 There are, in fact, no studies that evaluate the required frequency of testing for SDB in patients with NMD. There may be situations in which more frequent monitoring is required (eg, if a child's course of weakness becomes rapidly progressive).

There are several different tests that are used to detect SDB. They vary in complexity and availability. The accepted standard for determining the presence of SDB in patients with neuromuscular weakness is the polysomnogram.^13,14,140 The polysomnogram provides monitoring of cardiorespiratory function including heart rate, breathing frequency, abdominal and chest-wall movements, nasal air flow, oximetry, limb movement, sleep position, and video monitoring during sleep, as well as simultaneous monitoring of sleep with continuous electroencephalogram, electromyogram, and electrooculogram. For patients with neuromuscular weakness, it is also critical that exhaled and/or transcutaneous CO₂ is monitored as part of the study to detect nocturnal hypoventilation. This constellation of monitoring not only provides information about gas exchange during sleep, but also yields important information about sleep quality. Polysomnography, however, is cumbersome, intrusive, labor-intensive, and expensive. The act of monitoring itself can disturb a child's sleep and yield inaccurate results if the child does not achieve the usual amount of REM sleep. Its availability for children varies widely, and its scarcity has been associated with differences in practice patterns of how children with NMD are monitored for SDB.¹⁴²

With the recognition that polysomnogram monitoring in a pediatric sleep laboratory may not be widely available, other screening tests have been used to detect nocturnal hypoventilation or other forms of SDB. Home pulse oximetry studies with a simultaneous pulse waveform are considered a reasonable method to screen for SDB.^13,14,140 Oximetry studies, however, cannot differentiate between hypoxemia secondary to obstructive apneas and hypoventilation. Since appropriate therapeutic interventions require such a determination, this is a significant shortcoming of the test. One can infer that hypoxemia is the result of obstructive sleep apnea when desaturations occur in clusters of a saw-toothed pattern cyclically during periods of REM sleep, as opposed to the prolonged and sustained desaturations noted with nocturnal hypoventilation.^14,118 Oximetry studies can be hampered by movement artifact, long signal averaging, and poor signal detection.¹¹⁸ Arousals that help to avoid episodes of hypoxemia also will not be detected. Another limitation of pulse oximetry monitoring was highlighted by investigators who performed simultaneous overnight transcutaneous CO₂ and oximetry recordings in children supported by NIV.¹⁴³ They showed that significant episodes of hypoventilation could go undetected, when the S_pO2 remained normal despite an elevation of transcutaneous CO₂ to ≥50 mm Hg. Nevertheless, the British Thoracic Society recognizes overnight pulse oximetry as an acceptable method of screening for nocturnal hypoventilation in asymptomatic children with NMD if polysomnography or simultaneous capnography monitoring is not available.¹⁴ The guidelines caution, however, that oximetry monitoring alone is not an adequate screening tool for symptomatic children.

The addition of capnography (either transcutaneous or end-tidal) to oximetry monitoring improves detection of nocturnal hypoventilation, although this combination has not been methodically studied. Such episodes would appear as prolonged periods of hypoxemia (low S_pO2) with simultaneous hypercapnia (elevated transcutaneous CO₂ or P_ETCO2). Most of the current guidelines for evaluation of children with various NMDs recommend a study with the combination of oximetry and capnography or transcutaneous CO₂ monitoring as a screening test if polysomnography is not available and an overnight oximetry study if neither polysomnography nor oximetry with capnography studies are available.^13,14,18,140 Attempts to perform overnight studies at home in adults and children with NMD are in the early stage of investigation but offer some promise.^144,145

The Impact of Ventilatory Support in NMDs

In general, the goals of chronic mechanical ventilatory support in children are to prolong survival, relieve dyspnea, normalize gas exchange, correct SDB and abnormal sleep architecture, prevent or reverse cor pulmonale, improve daytime function, and promote and sustain growth and development. For patients with progressive NMDs, the addition of mechanical ventilation at night or continuously has changed the natural history of their diseases. One of the first reports of improved survival resulting from ventilatory support in young men with DMD compared 5 subjects who used NIV at least 7 h/night with 5 subjects who refused NIV.¹²⁸ Four of those who refused NIV died 9.7 ± 5.8 months from the onset of daytime hypercapnia, whereas all of those who used NIV survived. Among 23 young men with DMD who had diurnal hypercapnia and evidence of severe nocturnal hypoventilation, institution of NIV not only improved survival, but also resulted in normalization of daytime P_aO2 and P_aCO2.¹²⁶ In a retrospective review, Passamano et al¹⁴⁶ divided 516 subjects with DMD into groups based on epochs of birth: 1961–1970, 1971–1980, and 1981–1990. They found improved survival related to the institution of mechanical ventilation at 20 and 25 y with each increasing decade of birth (Fig. 2). The overall mean age of death for those subjects who did not receive mechanical ventilatory support was 17.7 y (range 11.6–27.5 y), but it was significantly higher (mean 27.9 y, range 23–38.6 y) in those who used mechanical ventilation.

• Download figure
• Open in new tab
• Download powerpoint

Fig. 2.

Kaplan-Meier curves demonstrating the improved survival of men with DMD by decade of birth from the 1960s to the 1990s. Innovations in treatment that included use of NIV, ACE inhibitors for cardiomyopathy, and deflazacort to preserve muscle function were introduced in this center at the beginning of the 1990s. From Reference 147, with permission.

The effect of nocturnal NIV on diurnal gas exchange has been confirmed elsewhere. In a prospective trial of NIV in 30 children from 6 to 19 y of age with a variety of NMDs, Mellies et al¹⁴⁷ demonstrated normalization of nocturnal gas exchange in all subjects and normal diurnal gas exchange in those who had demonstrated diurnal hypercapnia. The effect was sustained over 25.3 ± 12.7 months of observation. Furthermore, in 10 of the subjects who agreed to forego NIV for 3 consecutive nights, measurements of nocturnal and diurnal gas exchange quickly deteriorated but were then promptly restored with the reinstitution of NIV.

There have been other salutary outcomes attributed to the introduction of NIV for patients with NMD. One retrospective review involving 15 children between 3.4 and 17.8 y of age with a variety of NMDs also compared outcomes in the year before and after NIV initiation.¹⁴⁸ There was a significant reduction in number of days spent in the hospital and in length of stay in the pediatric ICU in the year following commencement of NIV versus the prior year. Two of the subjects also reported improvement in school performance. Another retrospective review of 14 children with NMD between 1.5 and 16 y of age who were treated with NIV reported a 73% reduction in hospitalizations in the year after institution of NIV support versus the year before it was begun.¹⁴⁹ There was also a trend toward a reduction in pediatric ICU days from 10.2 to 2.3 d (P = .06) and a significant reduction in health-care costs per patient from $55,129 in the year before NIV use to $14,914 in the year after (P = .003). In addition, a survey of the subjects and families showed no adverse effects on quality of life related to the introduction of NIV in this small cohort. In addition to improving gas exchange and sleep quality, NIV can also alter chest-wall growth. In children with SMA, the use of NIV with inspiratory pressures >14 cm H₂O has been shown to prevent or reverse the development of pectus excavatum.^150,151 It is not known, however, whether such alterations in chest-wall shape also promote lung growth by enhancing alveolarization.¹⁵⁰

When to Institute Ventilatory Support

Once sleep-related hypoxemia or hypoventilation is recognized in patients with NMD, recommendations are to offer patients ventilatory support, typically in the form of NIV.^13,14,139 Aside from the presence of hypoxemia or hypercapnia while asleep or awake, however, there is not consensus on when exactly to begin NIV in different diseases. Some recommend that markers like acute respiratory failure, failure to thrive, recurrent pneumonia, or symptomatic diurnal hypercapnia be used as triggers to institute long-term NIV.¹⁵² An apnea-hypopnea index >10/h or ≥4 episodes of S_pO2 <92% or drops in S_pO2 ≥4%/h of sleep on polysomnography have also been recommended as indications to begin NIV.¹³⁹ One center in Sweden surveyed their patients to determine the primary motive for commencing NIV.¹⁵³ Of 352 adults and children with a variety of NMDs, mechanical ventilation was begun electively in 268 (76%) and during an acute illness in the remainder. This cohort represented mostly adults, and no patient <5 y old was included. In those who began NIV electively, daytime sleepiness was the most common symptom, but in the subgroup of children with SMA, an insufficient cough was the most common reason to begin NIV. Hypercapnia occurred more frequently than hypoxemia in those who began NIV electively; notably, however, hypercapnia did not occur more frequently in those who complained of daytime sleepiness or headache versus those who did not.

French investigators hypothesized that the introduction of NIV before boys with DMD became symptomatic would delay the loss of lung function. In a multi-center national trial, they randomized those with normal daytime gas exchange along with an FVC between 20 and 50% predicted to begin nocturnal NIV or to continue standard therapy.¹⁵⁴ Enrollment in the study was curtailed, however, because investigators noted a decreased survival rate among the ventilated patients compared with controls. They speculated that NIV caused subjects to have a false sense of security, and therefore they practiced less careful monitoring. There was little description of the types of interventions used for airway clearance, however, and 7 of 10 deaths resulted from retained secretions and respiratory infection. Further, within the NIV group, 15 (43%) were inadequately supported by nocturnal mechanical ventilation. There was also a greater proportion of subjects with cardiomyopathy in the NIV group. For all these reasons, the dire warning of the authors to avoid NIV use until mechanical ventilation becomes imperative must be interpreted with caution.

Recognizing that waiting until daytime hypercapnia develops before starting NIV might subject patients with NMD to uncontrolled respiratory decompensation, Ward et al¹³³ conducted a randomized trial of NIV use in subjects with NMD or restrictive chest-wall disease who had nocturnal hypoventilation but normal daytime gas exchange. Twenty-six subjects were randomized to receive nocturnal NIV (n = 14) or to be treated conservatively without NIV (n = 12), but those in the control group were started on NIV if they developed daytime hypercapnia, worsening symptoms of nocturnal ventilation, recurrent chest infections, failure to thrive, or acute ventilatory decompensation. Those in the NIV group demonstrated improvement in gas exchange at 24 months of follow-up and better quality of life scores at 18 months than the controls. Of the 10 control group patients who were followed for the entire 24-month period, 7 had begun NIV by 12 months, and all but one met criteria to use NIV by 24 months of observation. The authors noted, therefore, that once a patient with NMD demonstrates nocturnal hypoventilation, diurnal respiratory failure is likely within 12–24 months. These data provide some framework for caregivers to use in discussing the role of NIV once patients have been diagnosed with nocturnal hypoventilation.

Once ventilatory support is commenced, the goal of support is to restore normal blood gas values. Since patients with NMD do not have intrinsic lung disease, targets of support should include a normal awake P_ETCO2 (35–45 mm Hg) and an S_pO2 >95% while the patient breathes room air. In fact, a decrease in oxyhemoglobin saturation in patients already receiving NIV support is an early sign of lower-respiratory tract dysfunction like atelectasis or pneumonia and is a signal to increase airway clearance measures.¹⁵⁵

Although it is clear that NIV improves sleep quality and survival in patients with NMD, NIV use in the treatment of acute respiratory failure or identification of SDB does not always trigger a discussion by health-care providers with patients about starting long-term NIV. In one center where 50 of 73 patients (69%) commenced NIV non-electively because of respiratory failure caused by pneumonia, the authors noted 191 missed prior opportunities to discuss beginning NIV with their patients.¹⁵⁶ Having a well-defined educational plan for patients and families that takes into account the stage of the patient's respiratory dysfunction, as detailed in the American Thoracic Society guidelines,¹³ might help to facilitate such discussions.

Noninvasive Ventilation Versus Ventilation via Tracheostomy

Patients with progressive NMDs who initially present with diurnal hypercapnia often experience improvement in daytime gas exchange as a result of nocturnal ventilatory assistance but eventually again develop daytime hypercapnia as respiratory muscles continue to weaken.¹²⁷ Once that occurs, patients require extended periods of ventilatory support to normalize gas exchange. This can be accomplished with noninvasive methods or with invasive ventilation via tracheostomy. Although methods of NIV include negative-pressure body ventilators (tank ventilator or iron lung, cuirass, or chest shell) and in older children and adolescents positive-pressure body ventilators (pneumobelt), by far the most common method of providing NIV is via positive pressure through a nasal, oronasal, or oral interface.

If a patient requires continuous mechanical ventilatory support, a noninvasive approach is still feasible. Older patients can receive daytime ventilatory support through a mouthpiece placed close to the patient's mouth, which can be accessed whenever he or she wishes¹⁵⁷ (Fig. 3). Several home-care ventilators can be adapted to provide this method of NIV, and some ventilators include settings for mouthpiece ventilation.^158,159 If the child is too young to understand how to use this sip ventilation or the buccal musculature is too weak to prevent leakage of air from the mouth, a nasal interface can still be used for most or all of the day (Fig. 4). Although there have been no randomized controlled trials comparing continuous NIV with invasive ventilation via tracheostomy, observational studies in which other aspects of care are similar demonstrate prolonged survival with either technique.^160,161

• Download figure
• Open in new tab
• Download powerpoint

Fig. 3.

Use of mouthpiece (sip) ventilation in a young man with Duchenne muscular dystrophy. The angled mouthpiece is positioned within easy reach of the patient's mouth for access to assisted ventilation as needed. The ventilator is mounted on the back of the patient's wheelchair.

• Download figure
• Open in new tab
• Download powerpoint

Fig. 4.

Use of continuous nasal ventilation in a young man with Duchenne muscular dystrophy and excessively weak buccal musculature. A: Note the alteration of the nasal interface, transforming it from a vented to a non-vented interface for use with a dual-limb ventilator circuit. B: The circuit is attached to a portable ventilator mounted on the back of the wheelchair to facilitate mobility and extended duration of support away from an electrical power source.

Advantages of NIV over invasive ventilation via tracheostomy include less risk of laryngeal or tracheal injury, fewer respiratory illnesses, less problem with mucus hypersecretion, and less complexity of care.^157,160,162 NIV also preserves oropharyngeal function for speaking, swallowing, and coughing,¹⁶² although one group found that weight loss and feeding problems occurred more frequently in their NIV group than in those supported by invasive ventilation via tracheostomy.¹⁶⁰ Use of NIV is also associated with shorter length of hospitalization when it is initiated, and one study demonstrated that children and adults could begin NIV as out-patients with the same rate of success as those started as in-patients.¹⁶³ In adults, NIV was shown to be significantly less expensive than invasive ventilation via tracheostomy and more often associated with an ability of subjects to be cared for at home,¹⁶⁴ whereas those adults supported by invasive ventilation via tracheostomy are more likely to be institutionalized.^160,164 Complications of NIV include inadequate triggering and cycling of the device in small children; leaks through the mouth or around the mask that impair adequate ventilation or irritate the eyes; intestinal bloating and aerophagia; dryness of the nose, mouth, and airways; facial, sinus, ear, or nasal discomfort from the interface; skin breakdown or irritation; and orthodontic problems.^165,166 Young children can also develop flattening of the midface and maxillary retrusion following long-term use of nasal interfaces for NIV^167–169 (Fig. 5). The use of custom-made nasal interfaces, which are more readily available in Europe than in the United States, can circumvent many of the complications related to the interface.^167,170

• Download figure
• Open in new tab
• Download powerpoint

Fig. 5.

Alteration in the shape of the face of a boy with myotubular myopathy (A) and a girl with congenital myasthenia gravis (B). Note the flattening of the malar eminences and resulting concave profile of the face resulting from prolonged use of NIV via nasal mask.

Contraindictions for NIV include severe hypoxemia, moderate to severe bulbar weakness associated with aspiration, an inability to clear the airway of secretions despite augmented cough techniques, failure to thrive or recurrent infections despite the use of NIV, an inability to tolerate noninvasive interfaces, patient preference, and lack of health-care provider familiarity with continuous NIV.^157,162,171 In those situations, tracheostomy should be considered.