Noninvasive Positive Pressure Ventilation To Treat Respiratory Failure

Methods

The primary approach for the review was search of the MEDLINE database for the past 20 years using the key words respiratory failure and mechanical ventilation. Studies using noninvasive ventilation techniques to manage acute or chronic respiratory failure were selected for analysis. This database was supplemented by searching the tables of contents of recently published issues of respiratory journals and by scanning the reference sections of papers identified in the MEDLINE search. Case reports were excluded, but all other original studies describing the use of noninvasive positive pressure ventilation in acute or chronic respiratory failure were reviewed. Because of the paucity of randomized, controlled studies, uncontrolled cohort studies were not excluded. Selection criteria, numbers of patients, diagnoses, types of ventilatory techniques used, duration of use, and outcomes, including changes in gas exchange and pulmonary functions, symptom relief, and success rates were extracted and tabulated, and complication rates were noted. Although variability and inadequacy of study design precluded meta-analysis of many of the studies, they were all reviewed critically and weaknesses in the existing database were identified. Conclusions relating to the use of noninvasive positive pressure ventilation were drawn whenever possible, and areas needing further investigation were highlighted.

Historical Perspective

Early mechanical ventilators were first described in the late 1700s. They consisted of various noninvasive devices that applied positive pressure to the upper airway through a bellows type of apparatus or through negative or positive pressure applied externally to the chest, back, or abdomen [4-8]. These early ventilators typically were powered manually and had varied success as acute resuscitators of infants and drowning victims [5]. Techniques for sustained ventilatory support awaited the wide availability of electricity, the development of electric motors, and the need for more pa tients with chronic respiratory failure. This need arose during the polio epidemics of the 1920s through the 1950s, when various noninvasive ventilators, such as the rocking bed [9, 10], and negative pressure devices, such as the iron lung [11], poncho-wrap, and tortoise shell ventilators [12-14], were developed. These ventilators successfully supported many survivors of the polio epidemics who had chronic respiratory insufficiency, sometimes for decades [10, 12-16]. During the early 1960s, however, control of the polio epidemics with the Salk and Sabin vaccines and the concomitant proliferation of positive pressure ventilation through endotracheal intubation caused a marked decrease in noninvasive ventilator use [17-19].

Interest in noninvasive ventilators resurged during the early 1980s when intermittent use (mostly nocturnal) of negative pressure ventilators was found to reverse daytime gas exchange abnormalities and symptoms of chronic hypoventilation in patients with severe kyphoscoliosis and other neuromuscular diseases, including muscular dystrophy, multiple sclerosis, and the postpolio syndrome [20-24]. Unfortunately, these “body” ventilators have several limitations, including bulkiness, lack of portability, difficult application in severely disabled patients, and the tendency to cause musculoskeletal discomfort [16]. Frequent and severe oxygen desaturation caused by transient upper airway obstruction during sleep in certain patients using negative pressure ventilators is the greatest concern [25-27]. Switching to noninvasive positive pressure ventilators from negative pressure ventilators ameliorates oxygen desaturation [26] and improves portability and ease of application. These advantages compared with other noninvasive forms of ventilation have increased the use of noninvasive positive pressure ventilation.

Invasive positive pressure ventilation was first introduced during the late 1800s and early 1900s [28] but was not extensively used for ventilatory support until the late 1950s and 1960s, when experience using anesthesia for surgery led to the development of positive pressure ventilators that reliably delivered preset pressures and volumes. These early positive pressure ventilators were used almost exclusively with artificial airways [18]. Noninvasive administration of intermittent positive pressure became popular during the 1950s through the 1970s for several respiratory disorders with the use of intermittent positive pressure breathing through a mouthpiece [29, 30]. Although some early studies showed that intermittent positive pressure breathing could transiently reverse carbon dioxide narcosis in patients with acute respiratory failure [31, 32], successes were inconsistent [33], and intermittent positive pressure breathing became used primarily to deliver aerosolized bronchodilator medication. It fell into disfavor after several studies, including a National Institutes of Health–sponsored multicenter trial [30], found no benefit of intermittent positive pressure breathing compared with standard nebulizer therapy for patients with chronic obstructive pulmonary disease exacerbations.

During the 1960s and 1970s, a few centers used noninvasive positive pressure ventilation to provide ventilatory support to patients with chronic respiratory failure using lip seals or facial masks as interfaces [2, 3], but the technique required cooperative, motivated patients and considerable patience and coaching from medical staff. In 1981, Sullivan and colleagues [1] described nasal continuous positive airway pressure used as a pneumatic splint to maintain upper airway patency in patients with obstructive sleep apnea and, several years later, relatively comfortable nasal masks became commercially available. During the mid-1980s, investigators began applying intermittent positive pressure ventilation through nasal interfaces and rapidly discovered that it augmented ventilation in patients with chronic respiratory failure, particularly during sleep [26, 34-43]. Patients also seemed to adapt to these more readily than to previously available interfaces. Most recently, small, relatively inexpensive, easily portable ventilators were developed to be used specifically with noninvasive positive pressure ventilation using a nasal mask [44-47] (Figure 1). Application of noninvasive positive pressure ventilation has been tested in patients with many varieties of acute [48-57] and chronic [34-43] respiratory failure, as an aid to weaning after bouts of acute respiratory failure [58], in obstructive sleep apnea unresponsive to high nasal continuous positive airway pressure [47], and in patients having difficulty after extubation [53].

View larger version:

• In this window
• In a new window

Figure 1. Moustaches (or beards in patients using facial masks) may interfere with effective mask sealing, and a nocturnal study to monitor oximetry and mask pressure should be done if the patient insists on growing one. A typical noninvasive positive pressure ventilation system using a nasal mask and a portable pressure-limited ventilator.

Noninvasive Positive Pressure Ventilation in Acute Respiratory Failure

For the past several decades, acute respiratory failure has been treated primarily with positive pressure ventilation through endotracheal intubation. To avoid endotracheal intubation and its attendant complications [59], investigators recently began administering noninvasive positive pressure ventilation to selected patients with acute respiratory failure. The theoretical advantages of this approach are to improve patient comfort; to reduce the need for sedation; to avoid the complications of endotracheal intubation, including upper airway trauma, sinusitis, otitis, and nosocomial pneumonia [59]; and to maintain airway defense, speech, and swallowing. Of course, limitations exist, including the need for patient cooperation; the lack of direct access to the airway, which could promote mucus plugging or atelectasis in patients with copious secretions; facial skin ulcers caused by mask pressure; and aerophagia. Noninvasive positive pressure ventilation has been tested in several uncontrolled series [50-56], in a study with historically matched controls [49], and in one randomized, controlled trial [57], as listed in Table 1.

Several interfaces and ventilator modes have been used to deliver noninvasive positive pressure ventilation to patients with acute respiratory failure (Table 1), including facial masks covering both the nose and mouth [49-51] and nasal masks [52-57]. Ventilator modes were pressure-limited [49-51, 53] and volume-controlled ventilation [52, 54-57]. Success rates, defined as the percentage of patients who tolerated the device, had improved gas exchange, and avoided endotracheal intubation, were similar regardless of the type of interface or ventilator mode used. Theoretically, facial masks allow less air leakage through the mouth, and nasal masks preserve speech and swallowing, but no study has directly compared their efficacy. Nevertheless, success appears to be related more to optimization of mask fit, comfort, and patient cooperation than to the particular mask or ventilator mode used.

Most studies have used broad selection criteria for patient entry, and patients with various diagnoses have been included (Table 1). Exacerbation of obstructive airways disease was the most common cause of acute respiratory failure among the studies. Fewer patients had respiratory failure caused by restrictive diseases, congestive heart failure, or pneumonia. In one study [53], patients with respiratory failure after surgery composed the largest diagnostic category. Success rates ranged from 67% to 88% and did not differ substantially among the diagnostic categories. Mean arterial pH and carbon dioxide tension values obtained within several hours after initiation of noninvasive positive pressure ventilation showed consistent improvement (Table 1). In one study [51], failure of gas exchange to improve or of the respiratory or heart rate to decrease during the first 2 hours of noninvasive positive pressure ventilation was predictive of ultimate failure.

Schedules for continuation of noninvasive positive pressure ventilation after the initial trial also differed widely among the studies. Patients in most studies used noninvasive positive pressure ventilation for 8 to 20 hours per 24-hour period, but in one study, patients were given 1-hour rest periods after 2 or 3 hours of use, amounting to approximately 18 hours of use per 24-hour period [54]. Patients continued using daily noninvasive positive pressure ventilation for several days to weeks, and in one study, some patients continued to use the ventilator on a long-term basis [54]. None of the studies discussed criteria for discontinuation of noninvasive positive pressure ventilation. Throughout the studies, patients tolerated the devices well and encountered few complications, consisting primarily of nasal skin ulcerations.

Although most studies reported success rates of 60% or more for noninvasive positive pressure ventilation in patients with acute respiratory failure caused by chronic obstructive pulmonary disease exacerbation, one small study reported no success in patients with this disease and found that application of noninvasive positive pressure ventilation was very time-consuming for nurses [52]. A second study [55] compared noninvasive positive pressure ventilation plus standard therapy with standard therapy alone in patients with acute chronic obstructive pulmonary disease exacerbations. Unfortunately, this study was not optimally controlled because patients who did not tolerate a brief noninvasive positive pressure ventilation trial composed the standard therapy group. The authors found no substantial advantage of adding noninvasive positive pressure ventilation to standard therapy. This contrasts with the findings of Brochard and coworkers [49], who administered noninvasive positive pressure ventilation to 13 patients with respiratory failure due to chronic obstructive pulmonary disease, only one of whom required subsequent endotracheal intubation. Among 13 historically matched controls, 11 required endotracheal intubation. Substantial reductions in the length of intensive care unit and total hospital stays were also observed among study patients compared with the control group. A favorable effect of noninvasive positive pressure ventilation was also shown in the one randomized, controlled trial [57]. In this study, 60 patients with acute respiratory failure due to chronic obstructive pulmonary disease were randomly assigned to receive nasal noninvasive positive pressure ventilation plus conventional treatment or conventional treatment alone. Patients treated with noninvasive positive pressure ventilation had an increase in pH compared with a decrease in controls, a larger decrease in PaCO₂ value, and less breathlessness. The mortality rate at 30 days was also substantially lower than in controls but only after exclusion of four patients randomly assigned to receive noninvasive positive pressure ventilation who did not actually use it.

The studies listed in Table 1 show that use of noninvasive positive pressure ventilation in selected patients with acute respiratory failure is well tolerated and associated with improved alveolar ventilation, but several caveats must be noted. First, only one adequately controlled published trial showed an advantage of noninvasive positive pressure ventilation compared with standard therapy [57]. Historically matched controls cannot be considered adequate because treatment strategies may change with time, and such studies usually favor the treatment group [60]. The findings relating to time consumption by nursing staff are particularly important because use of noninvasive positive pressure ventilation in the acute setting could pose a greater strain on intensive care unit resources than would standard practice alone. Until more controlled studies are published, noninvasive positive pressure ventilation to treat acute respiratory failure should be considered investigational and reserved for carefully selected patients.

Based on criteria used in the studies listed in Table 1, suggested guidelines for the selection of patients with acute respiratory failure to receive noninvasive positive pressure ventilation are listed in Table 2. Such patients should have physiologic evidence of acute respiratory failure or insufficiency, including acute respiratory acidosis and evidence of respiratory distress manifested by tachypnea, use of accessory muscles of inspiration, or abdominal paradox. Patients must also be sufficiently cooperative to follow instructions related to use of the mask, which must fit comfortably and without excessive air leakage. Noninvasive positive pressure ventilation should be avoided in patients with frank respiratory arrest or in those who are unstable because of hypotension, uncontrolled arrhythmias, or gastrointestinal bleeding. Patients at high risk for aspiration due to obtundation or swallowing difficulties and those with excessive secretions [48] are poor candidates and should be treated with endotracheal intubation if ventilatory support is necessary. Patients who are very obese may be difficult to ventilate because high inspiratory mask pressures are needed [51], but in our experience, such patients sometimes are supported successfully with noninvasive positive pressure ventilation. These guidelines should minimize risks associated with noninvasive positive pressure ventilation and allow for subsequent endotracheal intubation if the patient deteriorates despite the use of noninvasive positive pressure ventilation. Until more information is available, the best candidates for noninvasive positive pressure ventilation may be those with end-stage respiratory disease who have refused endotracheal intubation but who want temporary ventilatory assistance during an exacerbation [56] and those with acute decompensation of cystic fibrosis awaiting lung transplant [61, 62] who would be eliminated from transplant lists if they were endotracheally intubated.

Use of Noninvasive Positive Pressure Ventilation in Chronic Respiratory Failure

Efficacy of nocturnal negative pressure ventilation in many forms of chronic respiratory failure was reported during the early 1980s [21-23], and the nasal mask combined with positive pressure ventilation to deliver intermittent noninvasive positive pressure ventilation was a logical extension of the earlier experience. Reference to delivery of noninvasive positive pressure ventilation through the nose first appeared during the mid-1980s [34], and the first reports that provided gas exchange and pulmonary function data appeared in 1987 [26, 35-38] (Table 3). In that year, Kerby and associates [35] and Bach and colleagues [36] each described five patients with various neuromuscular diseases whose daytime arterial blood gases and symptoms improved after a few weeks of nocturnal nasal noninvasive positive pressure ventilation. Both studies noted improved tolerance of the nasal mask compared with a lip seal and elimination of intermittent upper airway obstruction compared with negative pressure ventilators. Ellis and coworkers [26] directly compared ventilatory assistance administered by negative pressure ventilators with that using nasal noninvasive positive pressure ventilation in five patients with neuromuscular disease. They observed that severe oxygen desaturation caused by intermittent upper airway obstruction during rapid eye movement sleep while patients used negative pressure ventilators were ameliorated during ventilation with nasal noninvasive positive pressure ventilation. They concluded that nasal noninvasive positive pressure ventilation was preferable to negative pressure ventilation because it stabilized the upper airway during rapid eye movement sleep.

Subsequently, Ellis and associates [39] reported reversal of hypoventilation in five patients with chronic respiratory failure caused by severe kyphoscoliosis who did not improve with nasal continuous positive airway pressure therapy alone. Others have since confirmed these earlier observations of the efficacy of noninvasive positive pressure ventilation in larger groups of patients with neuromuscular disease and thoracic deformities [40, 46, 63-65] (Table 3). These latter studies showed conversion of patients from negative pressure ventilators to noninvasive positive pressure ventilators [63] and successful support of patients who depend entirely on mechanical ventilation [40]. More recently, Hill and colleagues [45] found a deterioration in daytime symptoms, an increase in average nocturnal transcutaneous carbon dioxide tension from 49 to 59 mm Hg, a decrease in average nocturnal oxygen saturation from 93% to 87%, and no change in daytime arterial blood gases in patients with neuromuscular disease or chest wall deformities from whom nasal noninvasive positive pressure ventilation was removed for average 1-week periods. Deterioration in symptoms and nocturnal gas exchange were reversed after resumption of noninvasive positive pressure ventilation. These studies showed that noninvasive positive pressure ventilation support in patients with chronic hypoventilation due to neuromuscular disease, thoracic deformities, and idiopathic hypoventilation is associated with amelioration of nocturnal hypoventilation and oxygen desaturation, alleviation of daytime symptoms of hypoventilation, and improvement in daytime gas exchange. However, most of the trials were uncontrolled cohort studies, in which success rates depend on the patients' severity of illness. Randomized studies would help confirm efficacy but have been constrained by the ethics of withholding effective therapy from patients with life-threatening illnesses.

Although successful in treating patients with restrictive thoracic disease, the efficacy of noninvasive positive pressure ventilation in treating patients with chronic respiratory failure due to severe chronic obstructive pulmonary disease has not been definitively proved. Whether intermittent ventilatory support provided by other types of noninvasive ventilators is beneficial for patients with severe chronic obstructive pulmonary disease is controversial. Several groups of investigators have shown improvements in gas exchange and pulmonary function after intermittent periods of ventilatory assistance using negative pressure ventilators in uncontrolled studies [66, 67] or in short-term controlled studies [68-70]. However, other groups have not found similar benefits using the poncho-wrap ventilator in longer-term controlled studies [71-73]. In the largest study [73], 184 patients with severe stable chronic obstructive pulmonary disease were randomly assigned to receive intermittent negative pressure or sham ventilation 5 hours daily for 12 weeks. No improvements in exercise tolerance, pulmonary functions, muscle strength, arterial blood gases, or dyspnea were detected in the ventilated group despite documentation of respiratory muscle rest during ventilator use.

Similarly, conflicting results were reported when nasal noninvasive positive pressure ventilation was used to provide intermittent ventilatory assistance to patients with severe chronic obstructive pulmonary disease [74-76]. In a preliminary investigation, Elliott and associates [74] reported reversal of sleep hypoventilation and oxygen desaturation in patients using nasal noninvasive positive pressure ventilation. In a subsequent report, they [75] reported an improvement in daytime gas exchange in six patients with chronic obstructive pulmonary disease using nocturnal nasal noninvasive positive pressure ventilation for 6 months. However, Strumpf and coworkers [76] observed improvement only in tests of neuropsychologic function in a controlled study of patients with severe chronic obstructive pulmonary disease who used nasal noninvasive positive pressure ventilation in a 3-month cross-over trial. These investigators could not show any benefit in daytime gas exchange, pulmonary function, sleep quality, nocturnal oxygen saturation, or symptoms of dyspnea. Differing results between the studies may be explained in part by differences in the patient populations studied; on average, studies showing benefit of noninvasive ventilation [66-70] enrolled patients with more severe daytime and nocturnal gas exchange abnormalities than did studies showing no benefit [71-73, 76]. Thus, patients with chronic obstructive pulmonary disease and severe gas exchange derangement might benefit from intermittent noninvasive positive pressure ventilation. Nonetheless, current evidence suggests that intermittent noninvasive positive pressure ventilation offers little benefit in most patients with severe stable chronic obstructive pulmonary disease and minimal or no gas exchange derangement, and long-term efficacy is unproved in the subset of patients with more severe gas exchange derangement.

Despite convincing evidence that intermittent noninvasive positive pressure ventilation is effective in treating certain forms of chronic respiratory failure, patients must be carefully selected to optimize chances for success (Table 4). The best candidates are those with very slowly progressive disorders such as limb girdle muscular dystrophy, the postpolio syndrome, or severe kyphoscoliosis. Patients with obstructive sleep apnea and chronic hypoventilation that persists despite nasal continuous positive airway pressure therapy and those with obesity-hypoventilation and central hypoventilatory disorders are also candidates, although the very obese may be difficult to ventilate [51]. In more rapidly progressive neuromuscular disorders such as Duchenne muscular dystrophy, noninvasive ventilation appears to be temporizing, as weakness progresses during a period of months to years and patients are forced to use the ventilators for more hours each day [77]. Although noninvasive positive pressure ventilation can be used to provide uninterrupted ventilatory support in ventilator-dependent patients [3, 36], permanent tracheostomy has been recommended by some authors [78] for patients requiring ventilatory support most of the time (> 16 hours each day) as a more dependable means to assist ventilation. Optimal candidates for noninvasive positive pressure ventilation also should have minimal secretions and intact airway defenses. Therefore, patients with amyotrophic lateral sclerosis and the Guillain-Barre syndrome are usually poor candidates because of frequent bulbar involvement and swallowing dysfunction. Finally, although they are difficult to measure, cooperation and motivation are important qualities for optimal likelihood of success because the time necessary for patients to adapt to nocturnal use of noninvasive positive pressure ventilation may be prolonged.

Mechanisms of Action

On a cursory level, the mechanisms of noninvasive positive pressure ventilation seem self-evident: Intermittent positive pressure is transmitted through the upper airway to the alveoli, increasing transpulmonary pressure, inflating the lungs, and assisting alveolar ventilation, as is the case with positive pressure ventilation through an artificial airway. However, noninvasive positive pressure ventilation can be effective only if the patient can cooperate. For example, if the patient closes his or her glottis or allows air to leak out through the mouth during nasal ventilation, ventilatory support cannot be achieved [79]. Furthermore, patients receiving noninvasive positive pressure ventilation must learn to coordinate their breathing efforts with the ventilator so they allow it to assist their spontaneous breathing. In cooperative, awake patients, coordination with the ventilator usually can be achieved fairly quickly. However, because noninvasive positive pressure ventilation is often used nocturnally, patients must adapt their breathing patterns to the ventilator during sleep [80]. Even with spontaneous breathing during sleep, upper airway and respiratory muscle contractions must be synchronized to maintain upper airway patency [80-83]. Using noninvasive positive pressure ventilation through the nasal route, mouth or palatial closure must be synchronized with the ventilator to optimize ventilatory assistance. How much synchronization is necessary to achieve effective ventilatory assistance and the process by which this nocturnal adaptation takes place are poorly understood [80, 84]. In addition, it is not known whether structural or functional properties of the upper airway influence the efficacy of noninvasive positive pressure ventilation. For instance, nasal patency influences the severity of obstructive sleep apnea [85-88], and patients with high nasal resistances due to small nares may be less responsive to nasal noninvasive positive pressure ventilation than are those with low nasal resistance [40]. This could also be important if nasal resistance is increased transiently during upper respiratory tract infections.

Properties of the ventilator also may be important in the efficacy of noninvasive positive pressure ventilation. Because matching inspiratory air flow with the patient's inspiratory effort helps determine how much assistance a ventilator actually provides [88-90], the ventilator triggering mechanism may be a critical factor. Several triggering mechanisms are available on portable ventilators, including time, pressure, and flow-cycled mechanisms and those that combine spontaneous and pressure-triggered modes such as synchronized intermittent mandatory ventilation. Modes that require minimal inspiratory muscle activity from the patient are most attractive on a theoretical basis because they afford the greatest amount of respiratory muscle rest [89-91]. Time-cycled modes that are adjusted to suppress spontaneous patient respiratory activity [88] and highly sensitive flow-triggered modes, such as those available with pressure support ventilation [89-91], may be preferable to standard assist-control and synchronized intermittent mandatory ventilation modes.

Although patients with acute respiratory failure clearly should benefit from noninvasive positive pressure ventilation if they can coordinate their breathing with the ventilator and reduce their efforts to breathe, the mechanism for sustained improvement of gas exchange in patients receiving only intermittent noninvasive positive pressure ventilation for chronic respiratory failure is not so clear. Such patients using noninvasive positive pressure ventilation and other forms of noninvasive ventilation may have sustained reversal of daytime hypoventilation and symptomatic improvement when the ventilator is used nocturnally for as little as 4 to 8 hours each night.

Some theories have been proposed to explain the sustained improvement. One postulates that chronic respiratory failure is associated with chronic respiratory muscle fatigue [92, 93]. Intermittent respiratory muscle rest afforded by noninvasive positive pressure ventilation alleviates this fatigue and improves daytime ventilatory muscle function. Supporting this theory are studies showing that respiratory muscles do rest during noninvasive ventilation [79, 94-96] and that indices of respiratory muscle strength and endurance may improve in patients with chronic respiratory failure after varying periods of noninvasive ventilatory assistance [66-70]. Among studies showing resting of respiratory muscles, Carrey and associates [79] found that noninvasive positive pressure ventilation reduced phasic diaphragmatic electromyographic activity and esophageal pressure swings in three normal, four restrictive, and five chronic obstructive pulmonary disease patients. On the other hand, chronic respiratory muscle fatigue has never been adequately defined or convincingly shown [97-99], and several studies have not found improvement in respiratory muscle function despite amelioration of gas exchange abnormalities and symptoms after periods of intermittent ventilatory assistance [63, 71-73].

A second theory postulates that intermittent ventilatory assistance improves respiratory system compliance by reversing microatelectasis of the lung, thereby diminishing daytime work of breathing [100, 101]. This theory is supported by studies showing improvements in forced vital capacity without changes in indices of respiratory muscle strength after periods of positive pressure ventilation [100, 101]. However, data are again conflicting, with some studies showing no changes in vital capacity after periods of noninvasive ventilation [39, 46, 64, 65, 71-73], and no convincing evidence shows that a change in lung compliance occurs consistently during the course of noninvasive positive pressure ventilation.

A third theory proposes that chronic hypoventilation develops as the respiratory center adapts to limited capabilities of the respiratory system [38, 39, 43, 92, 93]. In this context, the respiratory center is thought to adjust its output so the work of respiratory muscles will not exceed the level that would precipitate muscle fatigue, a process sometimes called central fatigue [73, 98]. Particularly at night, when upper respiratory muscle tone and activity of nondiaphragmatic inspiratory muscles diminish, progressive nocturnal hypoventilation may occur, permitting an upward resetting of the respiratory center and worsening daytime hypoventilation [102]. Nocturnal noninvasive positive pressure ventilation prevents the worsening of hypoventilation at night and allows the respiratory center to be reset, thus reducing daytime hypercarbia. Evidence for this theory is found in a study showing amelioration of hypoventilation during nocturnal ventilatory assistance in patients with chronic respiratory failure [24]. When ventilatory assistance was discontinued for 1 night, nocturnal hypoventilation was less severe than before initiation of nocturnal noninvasive ventilation, suggesting that respiratory center sensitivity for carbon dioxide was reset. In a more recent study, worsening of nocturnal hypoventilation, oxygen desaturation, and daytime symptoms occurred without loss of respiratory muscle strength or vital capacity when nocturnal noninvasive positive pressure ventilation was discontinued for 1 week in patients with restrictive thoracic diseases whose chronic respiratory failure was reversed by nocturnal noninvasive positive pressure ventilation [45]. The worsening nocturnal hypoventilation and symptoms were promptly alleviated by resumption of noninvasive positive pressure ventilation. These latter studies suggest that prevention of nocturnal hypoventilation with possible resetting of respiratory center sensitivity for carbon dioxide may be the most important mechanism of noninvasive positive pressure ventilation. However, the three theories are not mutually exclusive and all could contribute to different degrees depending on the patient.

Application of Noninvasive Positive Pressure Ventilation

No universally accepted method exists for initiating noninvasive positive pressure ventilation. The following guidelines are based on our clinical experience and methods described by others in the literature [35, 36, 42, 64]. Implementation of noninvasive positive pressure ventilation requires a cooperative patient with an appropriate diagnosis (as described previously), a comfortably fitting interface, and properly adjusted ventilator settings. Based on our experience, patients probably will not adapt to noninvasive positive pressure ventilation unless they have symptoms of hypoventilation and are motivated by the desire for symptom relief.

Selection of a well-fitting interface is important because inappropriate selections may cause excessive patient discomfort, facial sores, or air leakage that interferes with acceptance or efficacy of the system. Various interfaces are commercially available, including nasal masks, facial masks, and lip seals. Standard nasal continuous positive airway pressure masks (Respironics, Inc., Murrysville, Pennsylvania; Healthdyne Technologies, Marietta, Georgia) are commonly selected because of their ease of application and ready availability (see Figure 1). Selection of an appropriate size is important, but pressure sores over the bridge of the nose are common even with proper fitting. Also, air leakage through the mouth may impair efficacy with any of the nasal interfaces. If this problem does not abate with continued use, chin straps, facial masks, or lip seals can be tried.

Facial mask interfaces [103] (Vital Signs, Inc., Totowa, New Jersey) are usually less acceptable than nasal masks for long-term use of noninvasive positive pressure ventilation because they cover both the nose and mouth and interfere with speech and eating, both of which can be accomplished with a nasal mask. However, facial masks have been used successfully for patients with acute respiratory failure because they prevent air leakage through the mouth [49-51]. Lip seals (Puritan Bennett Corp., Lenexa, Kansas) have been used for decades to deliver noninvasive positive pressure ventilation [2, 3], but in our experience, patient tolerance is limited because of aerophagia and difficulty in swallowing saliva.

To deliver noninvasive positive pressure ventilation to patients with chronic respiratory failure, a portable positive pressure ventilator is usually selected. Standard portable volume ventilators offer various modes and alarms and are suitable for use with patients who depend entirely on ventilators and with those who require only nocturnal ventilation. Recently, portable pressure-limited systems such as the bi-level positive airway pressure device (BiPAP; Respironics, Inc.) have become popular with patients requiring only nocturnal ventilatory assistance because of simpler application, greater portability (approximate weight, 5 kg), and lower cost than standard portable volume ventilators. These ventilators cycle between two levels of positive airway pressure, with the higher level assisting ventilation during inspiration and the lower level maintaining airway patency during expiration. Tracings showing mask pressure, esophageal pressure, and tidal volume during spontaneous breathing, use of continuous positive airway pressure, and a bi-level positive airway pressure device are shown in Figure 2. Because they maintain high air flow to sustain mask pressure, these pressure-limited ventilators may compensate for mask leaks better than volume ventilators, but they lack alarms and should not be used in patients who depend entirely on mechanical ventilatory support unless appropriate alarms are added.

View larger version:

• In this window
• In a new window

Figure 2. Note that during use of bi-level positive airway pressure, tidal volume is augmented, whereas swings in esophageal pressure are reduced, indicating a reduction in the work of breathing. Adapted from actual tracings in references 79, 95, and 96. Typical mouth pressure (P_m), esophageal pressure (P_es), and tidal volume (V_T) tracings during (from left) spontaneous breathing, use of continuous positive airway pressure (CPAP) at 10 cm H₂O, bi-level positive airway pressure using an inspiratory positive airway pressure (IPAP) of 10 cm H_2O and expiratory positive pressure using an IPAP of 15 cm H_2O and expiratory positive airway pressure (EPAP) of 7 cm H₂O.

A brief hospital stay is usually recommended for initiation of noninvasive positive pressure ventilation in patients with stable chronic respiratory failure [64]. The patient usually can be discharged after a few days to continue nocturnal ventilation for as long as tolerated after bedtime and to extend periods of use gradually until they can sleep through the night using the device. Patients require different periods of time before they fully adapt to nocturnal use, ranging from 2 nights to more than 3 months [45]. Approximately 20% to 25% of patients cannot adapt, mainly because of mask discomfort [64, 71]. As patients extend periods of nocturnal use, daytime arterial blood gases are obtained periodically, usually showing a gradual reduction in carbon dioxide tension, and nocturnal polysomnography is performed to assess adequacy of ventilation. Further adjustments in ventilator settings are made as necessary to control symptoms and to maintain a target arterial carbon dioxide tension in the range of 40 to 50+ mm Hg. Patients with chronic respiratory failure who cannot adapt to noninvasive positive pressure ventilation may try other noninvasive ventilators, including negative pressure ventilators [43].

During bouts of acute respiratory failure, more coaching and longer periods of initial use are needed because of the acute need for ventilatory support. Patients with acute respiratory failure using noninvasive positive pressure ventilation are at risk for respiratory arrest if the technique fails and they must receive intensive monitoring until their condition stabilizes. If air leaking through the mouth compromises efficacy of nasal noninvasive positive pressure ventilation during initial use, a facial mask can be tried. After initial adaptation, daily use of noninvasive positive pressure ventilation for acute respiratory failure ranges from 6 to 20 hours in most studies and is usually determined by patient tolerance (see Table 1). When failure occurs with acute respiratory failure, endotracheal intubation is usually needed unless the patient refuses intubation.

Conclusions

Compared with alternatives such as intubation with an artificial airway or use of negative pressure ventilators, noninvasive positive pressure ventilation offers several advantages. With acute respiratory failure, noninvasive positive pressure ventilation delivered by a facial or nasal mask may obviate the need for endotracheal intubation and thus preserve speech and the ability to eat, reduce trauma and infection, and possibly decrease the length of stay in the intensive care unit [49]. With chronic respiratory failure, noninvasive positive pressure ventilation delivered through the nasal route reverses nocturnal hypoventilation and improves daytime symptoms and gas exchange. It is more simple and less costly to administer at home than ventilation through permanent tracheostomy and obviates sleep-associated intermittent upper airway obstruction caused by negative pressure ventilators. It is also easier to apply and usually better tolerated by patients than are negative pressure ventilators. However, patients who receive noninvasive positive pressure ventilation must be selected carefully. With acute respiratory failure, patients must be sufficiently cooperative to coordinate their breathing with the ventilator and must not have excessive secretions. They should have sufficient upper airway function and cough to protect their airways and should not have hemodynamic instability. Because the use of noninvasive positive pressure ventilation for acute respiratory failure has not been fully evaluated, the most appropriate applications are in patients with acute or subacute respiratory failure who are reluctant to undergo endotracheal intubation and desire ventilatory assistance or in acutely decompensating patients with cystic fibrosis awaiting lung transplant.

Patients with chronic respiratory failure should be motivated and able to understand the use of the ventilator. Upper airway function should be intact, and ideally the patient should not be very obese. Patients with neuromuscular diseases, restrictive thoracic deformities, and idiopathic hypoventilation appear to respond best to noninvasive positive pressure ventilation. For patients with severe chronic obstructive pulmonary disease, indications must still be clarified. Those with substantial nocturnal oxygen desaturation or severe hypoventilation may respond favorably to noninvasive positive pressure ventilation, but controlled studies are lacking.

Several issues relating to the use of noninvasive positive pressure ventilation are unresolved. The optimal interface and ventilator design have not been determined, and these may differ among patients. Benefits of noninvasive positive pressure ventilation compared with intubation with an artificial airway in acute respiratory failure have not been firmly established. Complications of endotracheal intubation may be avoided, but costs and eventual outcomes have not been adequately compared. The mechanism by which intermittent use of noninvasive positive pressure ventilation improves chronic respiratory failure and indications for use in chronic obstructive pulmonary disease also need further clarification. Despite these unresolved issues, noninvasive positive pressure ventilation clearly represents an important addition to the techniques available to manage patients with respiratory failure, particularly as a home application for patients with chronic respiratory failure. As the technique is refined, it will probably be used more widely and may prove valuable for selected patients with acute respiratory failure.