The Growing Role of Noninvasive Ventilation in Patients Requiring Prolonged Mechanical Ventilation

Chronic Obstructive Pulmonary Disease

Given the success of NIV during COPD exacerbation,⁹ one might expect that this therapy might also be helpful in patients with stable COPD. NIV at night might improve gas exchange, it could improve sleep quality, and it could rest chronically fatigued respiratory muscles. Moreover, symptom management and prevention of exacerbations are important in reducing morbidity and mortality associated with COPD. There have been numerous randomized controlled trials^10–18 and observational studies^19–31 that have evaluated the use of NIV in patients with stable COPD. It is fair to say that the use of NIV for this indication is controversial.^32,33

A meta-analysis by Wijkstra of 4 studies^10,11,19,23 of nocturnal NIV in patients with COPD was published in 2003.³⁴ NIV was applied via a nasal or oronasal mask for at least 5 h/d for at least 3 weeks. The authors of this meta-analysis concluded that NIV support did not improve lung function, gas exchange, or sleep efficiency.

A more recent systematic review was published by Kolodziej et al.³⁵ They included 6 randomized controlled trials^10–15 and 9 observational studies^19–27 in their review. The authors concluded that NIV use in patients with severe stable COPD may improve gas exchange, exercise tolerance, dyspnea, work of breathing, frequency of hospitalization, health-related quality of life (HRQL), and functional status. However, these findings were based primarily on the results of the observational studies. For example, improved gas exchange was observed in the observational studies but not in the randomized controlled trials.

Two more randomized controlled trials^17,18 were reported that are not included in the meta-analyses of Wijkstra and Kolodziej.^34,35 Duiverman et al¹⁷ randomized 72 patients with COPD to nocturnal NIV in addition to rehabilitation (n = 37) or rehabilitation alone (n = 35). Inclusion criteria were stable clinical condition (no exacerbation in the 4 weeks prior to study participation, together with pH > 7.35), severe COPD (FEV₁/FVC < 70% and FEV₁ < 50% predicted), hypercapnia at rest (P_aCO2 > 45 mm Hg while breathing room air), and age 40–76 years. Outcome measures of functional status and gas exchange were assessed before and after the 3-month intervention period. For those who completed the protocol, the mean inspiratory positive airway pressure (IPAP) was 20 ± 4 cm H₂O, the mean expiratory positive airway pressure (EPAP) was 6 ± 2 cm H₂O, the mean respiratory rate on NIV was 18 ± 3 breaths/min, the inspiration time was 0.9 ± 0.2 s, and the rise time was 1.2 ± 0.6 s. Most patients used an oronasal mask (70%); the remaining used a nasal mask. Five patients (16%) dropped out due to intolerance of NIV. Those who completed the protocol used NIV an average of 96% of the days, with a median daily NIV use of 7.7 h (interquartile range 5.8–8.5 h/d). The authors found that NIV augmented the benefits of pulmonary rehabilitation in patients with COPD and chronic hypercapnic respiratory failure, as it improved several measures of HRQL, functional status, and gas exchange.

McEvoy et al¹⁸ conducted a multicenter randomized controlled trial of NIV plus long-term oxygen therapy (LTOT) versus LTOT alone in 4 Australian university hospitals. Enrolled patients had severe stable smoking-related COPD (FEV₁ < 1.5 L or < 50% predicted and FEV₁/FVC < 60% with awake P_aCO2 > 46 mm Hg and on LTOT for at least 3 months) and age < 80 years. Patients with sleep apnea or morbid obesity were excluded. Outcome measures were survival, spirometry, arterial blood gases, polysomnography, general and disease-specific quality of life, and mood. One hundred forty-four patients were randomized (72 to NIV + LTOT, and 72 to LTOT alone). Adherence to NIV was 4.5 ± 3.2 h/night; only 41/72 patients (60%) used NIV > 4 h/night and were included in the per protocol analysis. IPAP was 12.9 (12.5–13.4) cm H₂O and EPAP was 5.1 (4.8–5.3) cm H₂O. NIV was associated with an improvement in survival but no change in daytime arterial blood gases, pulmonary function measurements, or hospitalization rates. Disease-specific and general quality of life did not improve with NIV. In fact, it appears that the addition of NIV to usual care in these severely disabled COPD patients resulted in deterioration in general and mental health and some aspects of mood. The authors concluded that nocturnal NIV in stable oxygen-dependent patients with hypercapnic COPD may improve survival, but this appears to be at the cost of worsening quality of life.

In a retrospective observational study, Oscroft et al²⁹ evaluated the outcomes of 35 patients with a P_aCO2 > 56 mm Hg and normal pH following hospital admission with COPD exacerbation. Patients were initiated on long-term NIV; 4 patients were intolerant. NIV was titrated to a mean inspiratory pressure of 26 ± 3 cm H₂O and expiratory pressure of 4 ± 1 cm H₂O; 77% used oronasal masks. Patients used NIV 8.4 ± 3.5 h/d. Daytime arterial blood gases and nocturnal ventilatory parameters improved significantly on NIV. Median survival was 28.6 months.

High intensity NIV is an approach that has been promoted by Windisch and colleagues. The first report of this approach was in 2002.³⁶ After a period of acclimatization, patients were instructed to use NIV during the night and for up to 6 h during the day (total 14 h/d) until normocapnia was achieved. The time to reach normocapnia was 8.8 ± 3.8 days, and the mean hospital stay was 14.1 ± 5.1 days. From then on the ventilator was used at night and for up to 2 h during midday (total 10 h/d). Patients used a nasal mask. Ventilator settings were adjusted individually to normalize arterial blood gases, while maintaining the best comfort for the patient. The inspiratory pressure was increased stepwise until a further increase was not tolerated, after which the respiratory rate was increased beyond the spontaneous rate to establish passive ventilation. Further increases in respiratory rate were aimed at a progressive decrease of P_aCO2 toward normocapnia. Patients were discharged when normocapnia (P_aCO2 37–43 mm Hg) during NIV had been achieved or P_aCO2 did not show a further decrease during NIV for a minimum of 3 days. Of the 14 patients enrolled, 11 continued with the therapy for at least 6 months post discharge. Inspiratory pressure and rate required to reach normocapnia were 30 ± 4 cm H₂O (range 24–34 cm H₂O) and respiratory rate 23 ± 2 breaths/min (range 20–26 breaths/min).

In a retrospective study of 34 patients, Windisch et al evaluated 2-year survival, changes in lung function, and blood gases in patients with stable hypercapnic COPD in whom high intensity NIV was used.³⁷ NIV was predominantly used during the night, but patients were also instructed to use NIV for up to 4 h during daytime if necessary to control symptoms. The mean ventilator settings were an inspiratory pressure of 27.7 ± 5.9 cm H₂O (range 17–40 cm H₂O), respiratory rate 20.8 ± 2.5 breaths/min (range 14–24 breaths/min), and inspiratory time 1.0 ± 0.2 s (range 0.7–1.5 s). FEV₁ increased by a mean of 0.14 L, P_aCO2 decreased by a mean of nearly 7 mm Hg, and P_aO2 could be increased by nearly 6 mm Hg after 2 months of therapy. The authors suggest that earlier negative studies of NIV for stable COPD might have been due to lower levels of support than what was used in this study.

In another study, Windisch et al reported 2-year and 5-year survival rates for 73 patients receiving high intensity NIV of 82% and 58%, respectively.³⁸ In a prospective crossover study performed in 17 patients with severe stable hypercapnic COPD, Dreher et al³⁹ reported that high-intensity NIV, compared to lower intensity NIV, resulted in significant improvements in exercise related dyspnea, daytime P_aCO2, FEV₁, vital capacity, and the Severe Respiratory Insufficiency Questionnaire Summary Score. In an assessment of sleep quality, Dreher et al⁴⁰ reported that high inspiratory pressures used with long-term high intensity NIV produce acceptable sleep quality that is no worse than that produced by lower inspiratory pressures.

The reports of high intensity NIV are intriguing. Because the reports to date have all come from the same group of investigators, the generalizability of the results are unknown. Moreover, all of the data come from observational studies. The cost effectiveness of a program that requires a 2-week admission to the hospital is also unknown, and it is not known whether a similar program could be implemented on an out-patient basis.

Physical activity may be the strongest predictor of all-cause mortality in patients with COPD.⁴¹ A number of studies have evaluated NIV to augment exercise training in patients with COPD.^12,42–52 Benefits of NIV during exercise in patients with COPD include greater walking distance, cycle endurance, and the reduction in minute ventilation during exercise⁴⁸; improved exercise tolerance and dyspnea⁴⁹; improvements in heart rate, systolic blood pressure, blood lactate, and oxygen consumption after training⁵⁰; lactate/speed, maximum inspiratory pressure (P_Imax), 6-min walk distance, leg fatigue, S_pO2/speed, oxygen consumption, and dyspnea better with NIV than with supplemental-oxygen.⁵² Interestingly, one study found that dyspnea and walking distance were not improved, due to the burden of carrying the ventilator and oxygen in a backpack (Fig. 1).⁵¹ At present, use of NIV during exercise training is limited to stationary exercise (eg, treadmill or cycle) until smaller portable equipment is available.

• Download figure
• Open in new tab
• Download powerpoint

Fig. 1.

A patient with COPD using noninvasive ventilation in addition to supplemental oxygen. (From Reference 51, with permission.)

The overlap syndrome refers to patients who have both sleep-disordered breathing and COPD.⁵³ Machado et al⁵⁴ reported their experience in patients with COPD referred for LTOT. Patients with OSA symptoms were referred for polysomnography, and about 15% were confirmed to have the overlap syndrome. Marin et al⁵⁵ reported that CPAP eliminated the additional mortality risk of OSA in overlap patients, compared to COPD-only patients. It should be noted that the appropriate therapy for the overlap syndrome is usually CPAP, not NIV.

In 1999, a consensus group published recommendations for the use of NIV in patients with stable COPD (Table 1).⁵⁶ Overall, these recommendations seem reasonable today. Perhaps more important from a practical standpoint, reimbursement criteria of the Centers for Medicare and Medicaid Services are shown in Table 2. It is still unclear how NIV should be titrated if it is used in this population (eg, to comfort while awake, by polysomnography while asleep, to normalize blood gases). Evidence is lacking for benefit in the population of patients with severe COPD, although individual patients might benefit. Perhaps benefit will occur with approaches like high intensity NIV, but this requires further study before it can be recommended. Use of NIV may be beneficial in some patients during exercise training, but the studies that have been reported to date have evaluated physiologic end points and not patient important outcomes. Moreover, current Medicare criteria will most likely not cover NIV for this application.

Cystic Fibrosis

Cystic fibrosis (CF) is characterized by a progressive decline in lung function, which often progresses to end-stage respiratory failure. A physiologic rationale can be made for the use of NIV in patients with acute and stable CF, where it might unload respiratory muscles, relieve dyspnea, and improve alveolar ventilation and gas exchange. There have been a number of reports of the use of NIV for CF, but most have been short-term physiologic studies and observational reports.^57–80 In a survey of the use of NIV for CF in France, Fauroux et al⁷⁶ found that respiratory exacerbation was the most common criterion to initiate NIV. The primary indication for NIV in stable patients was diurnal hypercapnia, though certain other factors, such as respiratory sleep disturbance, were also important. It was also found that CF centers have high expectations about the potential benefits of NIV, but this contrasts with the lack of evidence of those benefits in this population. In an observational study, Fauroux et al⁷⁷ evaluated the effect of 1 year of NIV on lung function in patients with advanced CF. Patients who used NIV (n = 41) were compared to matched controls (n = 41). They found that long-term NIV was associated with stabilization of the decline in lung function. However, patients were not randomized in a prospective manner, and selection bias may have favored the group receiving NIV. Validated criteria for selecting patients with CF for NIV are not available, and the long-term benefits of NIV in this patient population are not known.⁸⁰

Several studies have evaluated the use of NIV as an adjunct to airway clearance therapy in patients with CF.^64,72,75 Fauroux et al⁶⁴ found that the use of NIV during chest physiotherapy was associated with an improvement in respiratory muscle performance and with a reduction in oxygen desaturation. Holland et al⁷² evaluated the use of NIV during chest physiotherapy in 26 patients with CF and found that NIV improved inspiratory muscle function, oxygen saturation, small airway function, and dyspnea. Placidi et al⁷⁵ evaluated 17 patients with CF and found no significant difference in sputum dry weight between mask positive expiratory pressure (PEP), CPAP, NIV, and directed cough. However, patients felt less tired after CPAP and NIV sessions than after mask PEP.

Three studies evaluated nocturnal NIV in patients with CF.^62,68,78 Gozal et al⁶² conducted a single-night study in which patients received room air on the first trial night. If the patient exhibited substantial hypoxemia or hypercapnia, or both, the results were compared for a single overnight session of NIV and oxygen and to a single overnight session of oxygen. NIV improved sleep-related hypoxemia and hypercapnia, but did not affect sleep architecture and arousals. Milross et al⁶⁸ also conducted a single-night trial of NIV (with or without oxygen), compared to an overnight session of low level CPAP and oxygen. Both NIV and oxygen therapy improved nocturnal S_pO2, especially during rapid-eye-movement (REM) sleep. The rise in transcutaneous P_CO2 with REM sleep was attenuated with NIV. Young et al⁷⁸ conducted a randomized crossover study of NIV in 8 patients with CF and hypercapnia. Six weeks of NIV were compared with 6 weeks of placebo (no NIV). NIV improved the chest symptom score in a CF quality of life questionnaire, the transitional dyspnea index score, the nocturnal P_tcCO2, and exercise performance. However, NIV did not improve sleep architecture, lung function, or awake P_aCO2.

A Cochrane review has evaluated NIV for CF.⁷⁹ The authors found that NIV may be a useful adjunct to airway clearance techniques, particularly in patients with CF who have difficulty expectorating sputum. They also found that NIV, when used in addition to oxygen, might improve gas exchange during sleep to a greater extent than oxygen therapy alone in patients with moderate to severe disease. These benefits of NIV have been demonstrated primarily in single treatment sessions with small numbers of subjects. The authors of this review conclude that the impact of NIV on exacerbations and disease progression remain unclear. They recommend adequately powered long-term randomized controlled trials to determine the clinical effects of NIV in CF.

High level evidence is lacking to support the use of NIV in patients with stable CF. Its use seems reasonable in patients with severe disease and hypercapnia, in patients with sleep-disordered breathing, and in patients who become fatigued during chest physiotherapy. However, given the lack of robust evidence, the use of NIV for CF should be approached on an n-of-1 basis, and reimbursement might be problematic.

Restrictive Thoracic Disorders

NIV has been used in the treatment of patients with restrictive thoracic disorders for many years.^81–99 Potential mechanisms for benefit in this patient population include unloading of respiratory muscles and improvement in gas exchange. When NIV is used at night, it is hoped that these mechanisms will lead to better daytime gas exchange and less daytime fatigue.

In 49 patients with kyphoscoliosis, Marti et al⁹⁸ reported a 5-year survival of 75%. P_aCO2 ≥ 50 mm Hg after 1 month of NIV and comorbidity (Charlson Index ≥ 3) were independent predictors of mortality. In that same study, 61 patients with chest wall deformity due to tuberculosis sequelae experienced a 5-year survival of 69%. Jager et al⁹⁶ retrospectively evaluated 188 patients; 85 received mechanical ventilation (all but 2 with NIV), and 103 received oxygen therapy alone. Mechanical ventilation was associated with a significantly better survival than oxygen therapy alone. After adjustments for age, sex, concomitant respiratory disease, blood gases, and vital capacity, the adjusted hazard risk of death was 0.35 (95% CI 0.17–0.70). Gustafson et al⁹² retrospectively evaluated 100 patients who received mechanical ventilation (all but 3 with NIV), and 144 patients received oxygen therapy alone. Patients treated with mechanical ventilation had better survival, even when adjusting for age, sex, concomitant respiratory diseases, and blood gas levels, with a hazard ratio of 0.30 (95% CI 0.18–0.51). These studies, although retrospective and non-randomized, make a strong case for the use of NIV in patients with chest wall deformity. Ward et al⁹¹ studied a heterogeneous group of 47 patients with nocturnal hypoventilation secondary to neuromuscular disease, including some with chest wall deformity. They found that patients with neuromuscular disease and nocturnal hypoventilation are likely to deteriorate, with the development of daytime hypercapnia and/or progressive symptoms within 2 years, and thus might benefit from the use of nocturnal NIV before daytime hypercapnia ensues.

Menadue et al⁹⁹ evaluated 13 patients with severe kyphoscoliosis who performed endurance treadmill tests in random order: sham pressure support, pressure support of 10 cm H₂O, and high-level pressure support of 20 cm H₂O. Endurance time was significantly greater with a pressure support of 20 cm H₂O. In a study by Borel et al,⁹⁷ 9 patients exercised with NIV and 7 without NIV. Exercise training effects did not differ between patients training with or without NIV. This is in contrast to a previous study by Borel et al,⁹⁵ in which they reported that NIV during exercise significantly improved exercise duration and tolerance and increased alveolar ventilation in patients with severe restrictive disease. Vila et al⁹⁴ evaluated NIV and 15 cm H₂O pressure support with 4 cm H₂O PEEP in 8 patients with hypercapnia and kyphoscoliosis. They found that gas exchange, dyspnea, and perceived effort were improved with NIV.

A Cochrane review evaluated nocturnal mechanical ventilation for chronic hypoventilation in patients with neuromuscular and chest wall disorders.⁹³ Eight eligible trials included 144 patients, most with neuromuscular disease and only a few with restrictive thoracic disorders. The authors concluded that current evidence about the therapeutic benefit of mechanical ventilation is weak, but consistent, and suggests alleviation of the symptoms of chronic hypoventilation. Further larger randomized trials are needed to confirm long-term beneficial effects of nocturnal mechanical ventilation on quality of life, morbidity, and mortality, to assess its cost-benefit ratio in chest wall diseases, and to compare the different types and modes of ventilation.

Obesity Hypoventilation Syndrome

OHS is defined as the triad of obesity, daytime hypoventilation, and sleep-disordered breathing in the absence of an alternative neuromuscular, mechanical, or metabolic explanation for hypoventilation.^100,101 From 1986 to 2005 the prevalence of body mass index (BMI) ≥ 40 kg/m² has increased by 5-fold, affecting 1 in every 33 adults. The prevalence of BMI ≥ 50 kg/m² has also increased by 10-fold, affecting 1 in every 230 adults. The prevalence of OHS is between 10% and 20% in obese patients with OSA, and is higher in the subgroup of patients with obstructive sleep apnea and extreme obesity. The most widely used therapy for managing OHS is CPAP or NIV.

Priou et al¹⁰² retrospectively evaluated 130 patients with OHS, either under stable conditions or during ICU management of acute hypercapnic exacerbation. Arterial blood gases and the Epworth sleepiness scale were significantly improved after 6 months of NIV. After a mean follow-up of 4.1 ± 2.9 years, 24 patients had died and 24 had discontinued NIV. The 1-, 2-, 3-, and 5-year survival probabilities were 97.5%, 93%, 88.3%, and 77.3%, respectively. Supplemental oxygen therapy was the only independent predictor of mortality. The probability of continuing NIV was 80% at 3 years with a high rate of daily use (> 7 h), and female sex was predictive of lower long-term adherence. The authors concluded that NIV is an effective and well tolerated treatment of OHS, whether initiated in the acute or chronic setting.

Piper et al¹⁰³ conducted a randomized trial of CPAP versus NIV for the treatment of OHS in 36 patients without severe nocturnal desaturation. There was no difference in adherence between the 2 groups, and both groups had similar improvements in symptoms. The authors concluded that both CPAP and NIV are equally effective in improving daytime hypercapnia. But these results may be relevant only in a subgroup of patients with OHS without severe nocturnal hypoxemia.

Redolfi et al¹⁰⁴ evaluated CO₂ sensitivity before and after nocturnal NIV in 6 patients with OHS. NIV was associated with improvements in gas exchange and CO₂ chemosensitivity, and plasma leptin increased from after initiation in NIV. Heinemann et al¹⁰⁵ evaluated 35 stable patients with OHS. They found that long-term domiciliary NIV normalized hypercapnia, markedly improved hypoxemia, and increased expiratory reserve volume. In a study of 15 patients with OHS, Chouri-Pontarollo¹⁰⁶ found that the lower the daytime CO₂ response, and the higher the proportion of REM sleep hypoventilation and daytime sleepiness, the better short-term therapy with NIV improved all of these parameters. In a study of 126 patients with OHS, Budweiser et al¹⁰⁷ reported that gas exchange and lung function were improved after initiation of NIV. Hypoxemia, high pH, and elevated inflammation markers predicted poor survival. Perez de Llano et al¹⁰⁸ assessed the outcome of 54 patients with OHS who were treated with NIV, and found that it provided a significant improvement in clinical status and gas exchange. In a 12-month study of the use of NIV in 13 patients with OHS, de Lucas-Ramos¹⁰⁹ et al reported reductions in P_aCO2 and increases in P_aO2, FVC, and ventilatory response to hypercapnia. Masa et al¹¹⁰ assessed the effectiveness of NIV in 22 patients with OHS, and in another group of 14 patients with kyphoscoliosis, and reported that NIV improved the clinical symptoms and the respiratory failure of patients with OHS to a degree similar to that reported with kyphoscoliosis.

Although the studies to date are observational and many are retrospective, the available evidence supports that NIV provides the benefits of CPAP (maintaining upper airway patency), while increasing tidal volume and alveolar ventilation. Long-term outcomes of NIV include improvement in blood gases, quality of life, and ventilatory response to CO₂.¹¹¹ NIV should be considered the treatment of choice in patients with OHS if nocturnal hypoxemia is persistent despite an adequate CPAP titration. To be successful, the inspiratory pressure should be at least 8–10 cm H₂O above the expiratory pressure needed to eliminate obstructive respiratory events.¹¹²

Neuromuscular Disease

Neuromuscular respiratory failure can result from disease affecting the cerebral cortex, brainstem, basal ganglia, spinal cord, motor nerves, neuromuscular junction, and respiratory muscles.¹¹³ The respiratory effects of a wide range of neuromuscular diseases of various etiologies are similar. Consideration should be given to ability to ventilate, ability to cough, and aspiration risk.

Although all of the evidence is observational, there is little debate that NIV is life-prolonging in patients with Duchenne muscular dystrophy (DMD). Ishikawa et¹¹⁴ described survival in 3 groups of patients with DMD: group 1 patients were untreated (prior to 1984), group 2 patients underwent tracheotomy (1984–1991), group 3 patients were managed by noninvasive mechanical ventilation and cardioprotective medications (1992–present). The 56 patients of group 1 died at 18.6 ± 2.9 years of age; the 21 patients of group 2 died at 28.1 ± 8.3 years of age, with 3 still alive; and the 88 patients using NIV had 50% survival to 39.6 years (P < .001). Bach and Martinez¹¹⁵ described survival outcomes with NIV for full ventilatory support, and a mechanically assisted cough and oximetry protocol, in a series of patients with DMD. With advancing disease, 101 nocturnal-only NIV users extended their NIV use throughout the daytime hours and required it continuously for 7.4 ± 6.1 years to 30.1 ± 6.1 years of age, with 56 patients still alive. There were 8 tracheostomized patients decannulated to NIV, and 31 intubated patients were extubated to NIV plus mechanically assisted cough.

Guidelines for the respiratory management of DMD have been published. Recommendations from an expert panel were published in 2010, related to necessary equipment, procedures, tests, diagnostic evaluations, and a structured approach to the assessment and management of the respiratory complications of DMD.¹¹⁶ This document recommends use of nocturnal NIV when there are symptoms of hypoventilation, S_pO2 < 95% or end-tidal carbon dioxide pressure (P_ETCO2) > 45 mm Hg, or when the apnea-hypopnea index is > 10 events/h or 4 or more episodes of S_pO2 < 92%, or drops in S_pO2 of at least 4% per hour of sleep. Daytime NIV is recommended for self extension of nocturnal NIV into the waking hours, abnormal deglutition due to dyspnea that is relieved by respiratory assistance, inability to speak in full sentences due to dyspnea, and symptoms of hypoventilation, with S_pO2 < 95% and P_ETCO2 > 45 mm Hg while awake. In 2007, a consensus statement was published on the respiratory management of patients with DMD undergoing anesthesia or sedation.¹¹⁷ For patients at increased risk of respiratory complications (FVC < 50% predicted, and especially for patients with FVC < 30% predicted), it is recommended to consider preoperative training in the use of NIV. For patients at high risk of ineffective cough (peak cough flow < 270 L/min or peak expiratory pressure > 60 cm H₂O), consider preoperative training in manual and mechanically assisted cough. It is also recommended to consider extubating patients with FVC < 50% predicted, and especially those with FVC < 30% predicted, directly to NIV. In a consensus statement by the American Thoracic Society,¹¹⁸ nocturnal NIV is recommended for sleep-related upper-airway obstruction and chronic respiratory insufficiency, and daytime ventilation is recommended when awake P_aCO2 is > 50 mm Hg or when daytime S_pO2 < 92%.

NIV is also commonly used in patients with amyotrophic lateral sclerosis (ALS). Bourke et al conducted a randomized controlled trial to assess the effect of NIV on quality of life and survival in patients with ALS. Patients were assigned to NIV (n = 22) or standard care (n = 19) when they developed either orthopnea with P_Imax < 60% of that predicted or symptomatic hypercapnia. NIV improved survival in the subgroup with better bulbar function, but not those with severe bulbar involvement. This subgroup showed improvement in several measures of quality of life, and a median survival benefit of 205 days (P = .006). An evidence-based practice parameter by the American Academy of Neurology¹¹⁹ recommends that NIV should be considered to treat respiratory insufficiency in order to lengthen survival, that it may be considered to slow the decline of FVC and improve quality of life, and that early initiation may increase adherence. The authors of a Cochrane review concluded that evidence from a single randomized trial of NIV involving 41 participants (discussed above) suggests that it significantly improves and maintains quality of life and prolongs survival in patients with ALS.¹²⁰

Several measures have been used to identify the appropriate time to initiate NIV in patients with respiratory muscle weakness associated with neuromuscular disease. An FVC < 50% predicted is the most common respiratory assessment used as an indication to initiate NIV. However, it may be a less sensitive indicator of respiratory muscle weakness than P_Imax. A P_Imax > −60 cm H₂O is also used as an indication to initiate NIV. Mendoza et al¹²¹ reported that there were no cases among 161 patients where FVC < 50% antedated P_Imax > −60 cm H₂O. Patients reached the P_Imax criterion 4–6.5 months earlier than the FVC criterion. In patients with diaphragm weakness, supine FVC (orthopnea) may be a better indicator of the time to initiate NIV than an erect FVC. A sniff nasal pressure can also be used, which is measured through a plug occluding one nostril during sniffs through the contralateral nostril.¹²² Overnight oximetry may be useful, as desaturation suggests nocturnal hypoventilation. The use of polysomnography to identify the need for NIV and titration of setting is controversial, although it is recommended in one consensus statement.¹¹⁸ Recommendations from a consensus committee suggest that NIV should be initiated in patients with neuromuscular disease with symptoms (such as fatigue, dyspnea, morning headache) and one of the following physiologic criteria: P_aCO2 ≥ 45 mm Hg, nocturnal oximetry demonstrating oxygen saturation ≤ 88% for 5 consecutive minutes; for progressive neuromuscular disease, P_Imax > −60 cm H₂O or FVC < 50% predicted. Criteria for reimbursement are shown in Table 3.

NIV can be used to provide respiratory support during gastrostomy tube placement in patients with neuromuscular disease. Several case series have reported that this is safe in patients with ALS and DMD.^123–130 Alternatively, the patient can be intubated for the procedure and extubated to NIV.

Successful extubation typically follows successful completion of a spontaneous breathing trial. Bach et al¹³¹ reported successful extubation of patients with neuromuscular disease and weakness who could not successfully complete a spontaneous breathing trial. Data were collected on 157 consecutive patients who could not pass a spontaneous breathing trial before or after extubation. The first-attempt protocol extubation success rate was 95%; 6 of 8 patients who initially failed extubation succeeded on subsequent attempts, so only 2 underwent tracheotomy. These results, although observational, suggest the NIV can be used not only to prevent intubation and tracheostomy, but can also be used to allow successful extubation.

Some patients with a tracheostomy and chronic respiratory failure may be able to breathe without mechanical ventilation during the day, but still require nocturnal ventilation to manage hypoventilation and improve respiratory muscle function. These patients might also be successfully transitioned to NIV at night, with a capped tracheostomy tube, and ultimately decannulated (Fig. 2).¹³²

• Download figure
• Open in new tab
• Download powerpoint

Fig. 2.

Noninvasive ventilation via nasal pillows, with a capped tracheostomy tube. (From Reference 132.)

For patients with chronic respiratory failure, appropriate airway clearance is important in addition to support of ventilation.^113,133 A peak cough flow (PCF) > 160 L/min has been reported to predict success in extubation and decannulation.¹³⁴ A PCF of < 160 L/min has been used as a threshold for initiating cough-augmentation therapy for patients with neuromuscular disease and cough insufficiency. A PCF < 270 L/min is considered a risk factor for complications associated with respiratory infection, because the PCF can fall below 160 L/min during respiratory infection.¹³⁵ A peak expiratory pressure < 60 cm H₂O or a history of repeated hospitalization for respiratory infection and an inability to clear secretions are also indications for initiating cough-augmentation therapy.¹¹⁸ Cough augmentation techniques include manual cough augmentation, hyperinflation maneuvers, and mechanical in-exsufflation therapy.^133,136

Despite the fact that many patients with chronic neuromuscular respiratory failure can be managed with NIV, tracheostomy becomes necessary for some. Indications for tracheostomy include intolerance of NIV, complications associated with NIV, aspiration of upper airway secretions, repeated lower-respiratory-tract infections despite aggressive cough augmentation, and patient wishes.

Interface

The interface is the weak link in the application of NIV. Desirable characteristics of an interface are listed in Table 4. The most commonly used interfaces are nasal masks and oronasal masks, and there are advantages and disadvantages of each (Table 5). Other interfaces used for chronic respiratory failure include nasal pillows, total face masks, and mouthpieces. Many sizes and designs are commercially available.

The nasal mask fits just above the junction of the nasal bone and cartilage, directly at the sides of both nares, and just below the nose above the upper lip. Some are gel-filled and others use an open cushion with an inner lip, in which pressure inside the mask pushes the cushion against the face. Nasal pillows, sometimes called nasal prongs or cushions, are available from several manufacturers. This interface consists of soft plastic plugs inserted into the nares, shaped in a way that the pressure applied during inspiration helps to seal the wall of the pillows against the inner surface of the nasal vestibule.

In a crossover design, Navalesi et al¹⁴¹ assessed 26 stable hypercapnic patients with restrictive thoracic disease or COPD who were naïve to NIV. They reported that the nasal mask was better tolerated than the other interfaces, but P_aCO2 was lower with an oronasal mask than with nasal mask, and minute ventilation was higher with oronasal mask than with nasal mask. A criticism of this study¹⁴¹ is that the authors used a very comfortable nasal mask (gel cushion) but a less comfortable oronasal mask (air inflatable cushion). In a randomized crossover trial involving 24 healthy volunteers, Holanda et al¹⁴² evaluated adverse effects and comfort of total face masks, oronasal masks, and nasal masks during NIV. They found no differences among the masks in terms of comfort score, respiratory rate, heart rate, and S_pO2. No differences in adverse effects were observed. The full face mask avoided pain on the bridge of the nose and presented no air leaks around the eyes and mouth. A criticism of this study is that it was conducted in healthy volunteers and thus the implications for long-term ventilation are unclear.¹⁴³

Willson et al¹⁴⁴ studied the effectiveness of a nasal mask versus an oronasal mask in patients with nocturnal hypoventilation. They reported that the type of interface did not affect gas exchange, and arousal indices were comparable for both masks. Meyer et al¹⁴⁵ assessed the frequency of occurrence and effect on sleep quality of air leaking through the mouth during nocturnal nasal ventilation in 6 patients with chest wall and neuromuscular disease using nocturnal nasal NIV. All patients had air leaking through the mouth for the majority of sleep, and air leaking through the mouth was associated with frequent arousals during lighter stages of sleep that interfered with progression to deeper stages, thus compromising sleep quality. Teschler et al¹⁴⁶ found that mouth leak reduced effective respiratory support, increased transcutaneously measured partial pressure of carbon dioxide (P_tcCO2), and disrupted sleep architecture.

Leaks through the mouth are common with a nasal interface. This can affect comfort, it can result in dry mouth, it can result in less effective ventilation, it can affect patient-ventilator interaction (trigger and cycle), and it can disrupt sleep architecture.¹³⁹ The patient can be coached to keep the mouth closed, but this may be ineffective, particularly during sleep. A chin strap can be tried, and the current generation of chin straps can be quite effective. Willson et al¹⁴⁴ found the chin strap effective in 14 of 16 subjects. Gonzalez et al¹⁴⁷ reported that a chin strap can reduce air leak and P_aCO2. However, the chin strap was effective only in about a third of patients. If persistent mouth leak occurs, an oronasal mask is often required. A nasal interface can be used during the daytime, and an oronasal mask is used at night to minimize mouth leak and improve sleep quality.

A potential problem with nasal and oronasal masks is facial skin breakdown, which most commonly occurs on the bridge of the nose. Perhaps the most important approach to prevent skin breakdown is to avoid strapping the mask too tight. A mask that is too large or that is too small increases the likelihood of poor fit and facial soreness. A mask with a forehead spacer or an adjustable forehead arm can be used to reduce the pressure on the bridge of the nose. Hypoallergenic care tape can be applied to the bridge of the nose, but this is less effective after substantial skin breakdown has occurred. Commercially available material is available specifically for this purpose. One can also consider the use of a different interface (eg, nasal pillows, mouthpiece, total face mask, or different manufacturer).

A small, angled mouthpiece or straw-type mouthpiece has been used for many years for NIV in patients with chronic respiratory failure (Fig. 3). Usually, the patient uses a mouthpiece during the day, but changes to a nasal or oronasal mask at night. The typical setup involves the use of a portable ventilator.¹⁴⁸ Although the circuit is open, the pressure in the circuit prevents a low-pressure alarm.¹⁴⁹ The respiratory rate is set on a low setting so that the patient can take breaths as needed, without the ventilator triggering when a breath is not needed. The patient activates the breath by placing the mouth on the mouthpiece and creating a small negative pressure in the circuit by sipping or inhaling. Some of the breath is lost to leak around the lips, so the tidal volume is set higher than what would be set otherwise. Tidal volume is set per the patient's comfort, typically in the range is of 0.7–1.2 L. The mouthpiece can be mounted close to the head so that the patient can speak after each breath. Some mouthpieces are configured with a lip seal to minimize air leak. In a report by Bach et al¹⁵⁰ of 257 patients with acute or chronic respiratory failure, mouth NIV was the predominant method of daytime ventilator support in 228 of the patients. Mouthpiece NIV was also used by 163 patients, of whom 61 had little or no measurable vital capacity. A lip seal or custom orthodontic interface was used for nocturnal mouthpiece NIV. Toussaint et al¹⁵¹ evaluated daytime mouthpiece ventilation in 42 patients with DMD, and found it to be safe and to prolong survival.

• Download figure
• Open in new tab
• Download powerpoint

Fig. 3.

Mouthpiece ventilation. Left: Patient with mouthpiece and ventilator circuit in standard position. Right: Rear view of wheelchair with ventilator and mouthpiece circuit in place. (From Reference 148.)

The pressure on facial structures from the mask has been associated with flattening of facial structures in children, because forces responsible for the broadening and outward growth of the face can be counteracted by the inward pressure exerted by the mask.¹⁵² This causes facial flattening, particularly of the mid-face,^153,154 and malocclusion with retrusion of the maxillary ridge (class III angle malocclusion).^154,155 Fauroux et al¹⁵⁵ quantified mask-related facial injuries and reported that global facial flattening was present in 68% of the children: 43% had flattening of the forehead, 38% had flattening of the malar area, and 28% had flattening of the maxilla. A concave face was reported in 12% of the children. Maxillary retrusion was found in 37% and was seen more commonly when the nasal mask was used. Facial flattening was not associated with patient age, cumulative use of nasal NIV, or underlying disease. Skin injury was much more common in children greater than 10 years of age. Young children treated with NIV should be followed by an orthodontist so that necessary mask adjustments can be made or corrective orthotic devices can be employed.¹⁵⁴

Ventilator

Table 6 lists considerations in the selection of a ventilator for NIV. There are 3 types of circuits commonly used with positive pressure ventilators for NIV. Dual limb circuits, such as those used with critical care ventilators, have inspiratory and expiratory valves, and separate hoses for the inspiratory and the expiratory gases. For portable ventilators, such as those typically used for ventilation in the home, a single limb circuit is used with an exhalation valve near the patient. The expiratory valve is actively closed during the inspiratory phase to prevent loss of delivered tidal volume. Because the expiratory valve is near the patient, rebreathing is minimized. For bi-level pressure ventilators, a single limb circuit is used. A leak port is present, which serves as a passive exhalation port for the patient. In some configurations, the leak port is incorporated into the circuit near the patient. In other configurations, the leak port is incorporated into the interface.

Bi-level pressure ventilators are blower devices that deliver inspiratory and expiratory pressures. They typically provide pressure support ventilation (PSV) or pressure controlled ventilation (PCV); the difference between the IPAP and EPAP is the level of pressure support (or pressure control). These ventilators use a single limb circuit with a passive exhalation port. For devices used in the home, supplemental oxygen, if required, is titrated into the circuit at the ventilator outlet or into the mask. F_IO2 is determined by the oxygen flow, respiratory pattern, and leak.

Intermediate ventilators are commonly used for home care ventilation. Most current generation devices use a single limb circuit with an active exhalation valve near the patient. Older generations provided only volume controlled ventilation (VCV), but newer generations provide VCV, PCV, and PSV. Some of these are designed for either invasive or noninvasive ventilation. They vary in their ability to compensate for leaks; some compensate well and some not at all. Most have internal batteries, which power the ventilator for many hours, and can be mounted on a wheelchair.

Important to the function of ventilators for NIV is their ability to tolerate leaks. The function of bi-level pressure ventilators depends on the presence of a leak. Leaks include intentional leak through the passive exhalation port as well as any unintentional leaks that may be present in the circuit or at the interface. Some ventilators are able to detect unintentional leak and adjust flow to accommodate the leak. Leak detection algorithms must adjust for changes in leak with inspiratory and expiratory pressure changes, as well as changes that may occur breath-to-breath due to fit of the interface. Some newer generations of bi-level pressure ventilators use redundant leak estimation algorithms. Storre et al¹⁵⁶ assessed leak compensation during PCV and VCV in 14 stable hypercapnic patients with COPD receiving long-term NIV. Leak compensation decreased leakage-induced dyspnea. However, much of the additional inspiratory volume during PCV was lost via leakage. The authors concluded that leak compensation largely feeds the leak and results in only a marginal increase of tidal volume. PCV, but not VCV, resulted in clinically important leak compensation.

For long-term care, the clinician must decide among VCV, PCV, and PSV. There are advantages and disadvantages of either approach for NIV (Table 7). PSV is commonly used for NIV. A theoretical advantage of PSV is that it varies the inspiratory flow to meet patient demand, which should improve patient comfort during NIV.¹³⁹ Girault et al¹⁵⁷ compared VCV and PSV in 15 patients with hypercapnic respiratory failure and reported that both modes similarly improved breathing pattern and gas exchange. The inspiratory work load was less with volume control, but patient comfort was greater with pressure support. In patients with stable CF, Fauroux et al⁶⁷ reported that both pressure and volume ventilation decreased respiratory muscle unloading. In 13 ventilator-naïve patients with chest wall deformity, Struik et al¹⁵⁸ compared volume-targeted and pressure-targeted ventilation in a randomized fashion. Two patients switched from volume to pressure ventilation due to lack of tolerance, but improvements in P_aCO2 and P_aO2 were not significantly different between the approaches. Available evidence has not shown any mode to be clearly superior to another for NIV.¹³⁷

An issue of some controversy is whether or not a back-up rate is needed for NIV in patients with stable chronic respiratory failure. This has financial implications because, in the United States, reimbursement for a device with a backup rate (E0471 respiratory assistance device) is more than twice that for a device without a backup rate (E0470 respiratory assistance device). In a retrospective study of 182 post-tuberculosis patients receiving long-term NIV, Tsuboi et al¹⁵⁹ found that patients treated with a pure controlled mode (rate determined by the ventilator) had longer survival rates than those treated with a mode that allowed patient triggering. In fact, the 10-year probability of continuing NIV in those receiving the controlled mode was approximately 3-fold higher. Dellweg et al¹⁶⁰ measured blood gas values, respiratory muscle strength, spontaneous breathing pattern, and lung function before and after a 3-month period of NIV in 305 patients; 91% of the patients were able to adapt to a controlled mode of NIV. They observed a significant reduction in daytime P_aCO2, increased daytime P_aO2, improved lung function, and improved maximum respiratory muscle strength after 3 months of NIV using the controlled mode. Fauroux et al⁷³ evaluated the effect of a back-up rate on respiratory effort during NIV. The back-up rate was progressively increased to the maximum that patients could tolerate, which resulted in decreased respiratory effort in 10 patients with CF. A back-up rate during NIV is particularly important in patients prone to hypoventilation with loss of wakefulness respiratory drive during sleep, such as those with neuromuscular disease^161,162 or stable chronic alveolar hypoventilation syndromes.¹⁶³

Proportional-assist ventilation (PAV) has been used to improve patient-ventilator synchrony during acute respiratory failure.¹⁶⁴ In patients with chronic respiratory failure due to neuromuscular disease and chest-wall deformity, Hart et al¹⁶⁵ found that PSV and PAV produced similar improvements in physiologic variables. However, greater diaphragm unloading was observed with PSV than with PAV, which was associated with greater symptomatic benefit. Porta et al¹⁶⁶ compared the short-term physiologic effects of PSV and PAV in 11 patients with clinically stable COPD and 7 patients with restrictive chest-wall diseases. They reported that noninvasive PAV, set at the patient's comfort, was not superior to PSV in unloading the inspiratory muscles. Winck et al¹⁶⁷ compared the tolerance and physiologic effects of a 5-night treatment with either nasal PAV or PSV in patients with chronic respiratory failure (4 with COPD and 10 with restrictive thoracic diseases). PAV and PSV had similar patient tolerance and were equally effective in reducing daytime hypercapnia and improving nocturnal oxygen saturation and symptoms. PAV was associated with less nasal and oral dryness, but with more alarm noise. Poggi et al¹⁶⁸ compared PAV versus PSV in 8 male patients with chronic respiratory failure due to COPD. They found that PAV and PSV unload the diaphragm, but that PAV can be more efficient than PSV, by producing a greater minute ventilation for a similar rise in transdiaphragmatic pressure. Importantly, PAV is designed to only amplify patient effort. There is no minimal support provided in the setting of very weak or absent efforts.

Average volume-assured pressure support (AVAPS) maintains a tidal volume equal to or greater than the target tidal volume, by automatically controlling the minimum and maximum IPAP settings. AVAPS averages tidal volume over time and gradually changes the IPAP over several minutes to achieve the target tidal volume. If patient effort decreases, IPAP is increased to maintain the target tidal volume. On the other hand, if patient effort increases, IPAP is reduced. Because AVAPS is a form of adaptive pressure-controlled ventilation, there is a concern that the ventilator will inappropriately decrease support if respiratory drive increases.¹⁶⁹ In 10 patients with OHS who did not respond to therapy with CPAP, the effects of NIV with PSV and AVAPS on ventilation pattern, gas exchange, sleep quality, and HRQL were assessed in a randomized crossover trial.¹⁷⁰ NIV with PSV substantially improved oxygenation, sleep quality, and HRQL in patients with OHS. AVAPS provided additional benefits on ventilation quality, thus resulting in a more efficient decrease of P_tcCO2, but this did not provide further clinical benefits regarding sleep quality and HRQL. In a single-blind randomized crossover study of 28 patients with chronic respiratory failure, Ambrogio et al¹⁷¹ found that AVAPS was comparable to PSV with regard to sleep efficiency.

Another type of adaptive pressure-controlled ventilation is intelligent volume-assured pressure support. This mode is designed to guarantee minimum alveolar ventilation, defined as minute ventilation minus anatomical dead-space ventilation. The level of pressure support is continuously adjusted by comparing the continuously updated estimate of alveolar ventilation with the target alveolar ventilation. Battisti et al¹⁷² evaluated this approach in 19 patients receiving NIV for acute hypercapnic respiratory failure. They found that the alveolar ventilation based control of PSV during NIV was feasible and led to beneficial effects in patients, comparable to manually set PSV. In a randomized crossover trial of 25 patients with COPD and chronic respiratory failure, Oscroft et al¹⁷³ compared NIV with intelligent volume-assured pressure support to conventional PSV. Each treatment period consisted of 8 weeks. No significant differences were found in arterial blood gases, nocturnal oxygen saturation, or any of the secondary outcomes that included lung function, exercise capacity, nocturnal P_tcCO2, health status, or adherence. The authors concluded that the 2 modes were equally effective in the long-term management of ventilatory failure associated with stable COPD. Jaye et al¹⁷⁴ conducted a randomized crossover trial that compared NIV with intelligent volume-assured pressure support to PSV in 20 patients with neuromuscular and chest wall disease who required long-term ventilation for nocturnal hypoventilation. They found comparable control of nocturnal oxygenation without compromising sleep quality.

Rise time (pressurization rate) is the time required to reach the inspiratory pressure at the onset of the inspiratory phase with pressure support or pressure control ventilation.¹⁷⁵ Rise time should be set to maximize patient comfort. Ramp settings reduce the initial pressure and then gradually increase it to the pressure setting. A ramp has been used primarily in patients receiving CPAP for sleep apnea, the objective being to allow the patient to fall asleep more comfortably. The role of a ramp during NIV is unclear. Although it might be considered to improve patient tolerance of NIV, it also delays the onset of effective respiratory support. Another embellishment of NIV is Bi-Flex, which inserts a small amount of pressure relief during the latter stages of inspiration and the beginning part of exhalation. Evidence supporting the use of Bi-Flex with NIV is anecdotal.

Although the use of humidification during NIV is controversial,¹⁷⁶ it is commonly used to improve patient comfort and to prevent drying of the upper airway. Humidification can be provided by an active heated humidifier or a passive heat-and-moisture exchanger (HME). In a randomized crossover 12-month study,¹⁷⁷ 16 patients with stable chronic hypercapnic respiratory failure received either heated humidification or HME. At the end of the trial, 10 out of 14 surviving patients decided to continue with heated humidification. Lellouche et al¹⁷⁸ reported that minute ventilation was significantly higher with an HME than with a heated humidifier for a similar P_aCO2. Jaber et al¹⁷⁹ reported that P_aCO2 was significantly higher with an HME than with a heated humidifier, and concluded that the increased dead space of an HME may negatively affect respiratory function and gas exchange. Based on this evidence, the use of an HME during NIV is not recommended.

Ventilators for NIV can be battery-powered for safety and increased portability. Some ventilators that can be used for NIV have an internal battery. Others can be powered with a battery or uninterruptable power supply. The duration of a fully charged battery is determined by the size of the battery, ventilator settings, amount of leak, and whether or not a humidifier is used. When using a battery, it is generally best not to use a humidifier if possible, as this will extend the life of the battery.