Chapter 6. Noninvasive Ventilation

Acute respiratory distress is a frequent problem encountered by emergency physicians and intensivists. Often the clinician must act to ensure adequate oxygenation and ventilation before a definitive diagnosis is achieved. The treatment of acute respiratory distress requires an aggressive approach that entails use of medications, oxygen, and often positive pressure ventilation. Historically, patients who required positive pressure ventilation underwent endotracheal intubation (ETI) and were placed on a mechanical ventilator. However, over the past decade there has been an increased use of noninvasive positive pressure ventilation (NIV).1,2 As opposed to ETI, NIV uses an external mask interface to deliver positive pressure to the patient.

NIV and ETI with conventional ventilation are not synonymous alternatives; NIV is not warranted when the airway needs to be secured. Rather, NIV should be considered an additional tool that can augment medical care to possibly prevent ETI. This chapter will discuss the use of NIV in the emergency medicine patient with acute respiratory distress.

Nomenclature

There are two major types of NIV (or noninvasive positive pressure ventilation) that are used in the prehospital and emergency medicine setting: continuous positive airway pressure (CPAP) and bilevel positive airway pressure (BiPAP).

CPAP provides continuous positive airway pressure throughout the entire respiratory cycle (see Figure 6-1). There are small variations in pressure that are dependent on patient respiratory effort: a drop in pressure that occurs with each spontaneous inspiration and a rise in pressure that occurs with each exhalation. The set pressure will very closely approximate mean airway pressure (Pma). The amount of flow or tidal volume (Tv) will depend on patient effort, lung compliance, and the fit of the mask.

BiPAP consists of two applied pressures: inspiratory positive airway pressure (IPAP) and expiratory positive airway pressure (EPAP) (see Figure 6-2). EPAP is similar to applied positive end-expiratory pressure (PEEP) on the mechanical ventilator, maintaining positive pressure throughout the expiratory cycle. IPAP provides a higher positive pressure during inspiration to support the work of breathing and augment ventilation.

CPAP and EPAP: Opening the Lung and Keeping It Open

CPAP and EPAP (in BiPAP) are analogous to applied PEEP in the mechanically ventilated patient, applying pressure above that of atmospheric pressure during the expiratory cycle. The addition of positive pressure during the expiratory cycle can have multiple physiologic effects, and depending on particular patient pathophysiology, these effects can be beneficial or harmful.

The addition of CPAP and EPAP (applied PEEP) is helpful in the treatment of hypoxic respiratory failure that is refractory to supplemental oxygen. The beneficial effects of applied PEEP in hypoxic respiratory failure occur primarily through the opening of collapsed, atelectatic, or fluid-filled alveoli that have a low ventilation to perfusion (V/Q) ratio. In these alveoli, there is absent or inadequate ventilation and blood shunts from the right side of the circulation to left side without unloading carbon dioxide or oxygenating hemoglobin. Some of these fluid-filled or collapsed alveoli are “recruitable” and have the potential to be “opened” and take part in gas exchange, depending on the underlying disease process and severity of illness. “Recruitable” alveoli can open and close during the respiratory cycle or remain closed throughout the entire cycle. Applied PEEP can help limit expiratory collapse by providing positive pressure during exhalation that splints the alveoli open, or it can serve as a pressure head that opens collapsed alveoli. Opening collapsed alveoli decreases the shunting of venous blood and leads to an improvement of hypoxemia.3–5

At higher pressures, applied PEEP has negative effects that may outweigh the benefits of recruitment. First, some diseased alveoli are not “recruitable” and increasing levels of applied PEEP will not improve shunt effect.4 Second, high alveolar pressures can cause overdistention of healthy alveoli leading to barotrauma and the release of inflammatory cytokines, propagating pulmonary and nonpulmonary organ injury.4 Third, high PEEP can have negative effects on venous return and cardiac output in preload-dependent states (e.g., sepsis, hypovolemic shock), leading to decreased oxygen delivery and tissue perfusion.4 Finally, high PEEP can paradoxically worsen V/Q mismatch by overdistending alveoli and decreasing blood flow to previously perfused healthy lung segments.3–5

The level of applied PEEP provided in NIV is usually beneficial to patient oxygenation. Provided that the patient is not in a preload-dependent state, the negative effects of applied PEEP usually occur at higher pressures that are not well tolerated by patients because of discomfort and air leaks from the mask interface.

If a patient requires higher levels of CPAP or EPAP (greater than 12 or 15 cm H2O) to maintain oxygenation, this should signal that the clinical status is worsening and that the disease state is not amenable to NIV. Conventional ETI with mechanical ventilation should be strongly considered in these situations.

Flow and Tidal Volume: Getting Air in and out Of the Lungs

Airflow and ventilation are directly dependent on the pressure gradient and inversely related to airway resistance from the atmosphere to the alveolus. This can be conceptually explained as Ohm's law:

Rearranging Ohm's law for flow gives you the following:

Airway resistance is important in many different disease states (e.g., chronic obstructive pulmonary disease [COPD] and asthma), and it is essential that the clinician optimize treatment (i.e., steroids and β-agonists) to decrease resistance and maximize potential flow when indicated. From a schematic standpoint, we will assume airway resistance remains relatively constant from one breath to another, and that the pressure gradient between the atmosphere and the alveolus becomes the major determinant of air flow.

In the spontaneously breathing individual, the pressure gradient between the alveolus and the atmosphere is accomplished by creating negative pressure in the thorax. At the onset of inspiration, the diaphragm and intercostal muscles contract, increasing intrathoracic volume and decreasing intrathoracic pressure. Relative to the atmosphere, the alveolus is at a negative or lower pressure and air flows through the airways down a pressure gradient into the alveolus. At the end of inspiration, the elastic recoil of the chest wall increases alveolar pressure, creating a positive pressure relative to the atmosphere; air flows out of the lungs down a pressure gradient.3,5

The same principle of air flowing down a pressure gradient applies to NIV and other forms of positive pressure ventilation. In both cases, during inspiration the ventilator provides support in the form of positive pressure at the airway to create a pressure gradient for air to flow from the atmosphere (or ventilator) into the alveolus. Atmospheric pressure is made more positive (as opposed to alveolar pressure becoming more negative) to create the pressure gradient necessary for inspiratory flow to occur. Exhalation is similar to the nonventilated patient in that it is a passive phenomenon where the elastic recoil of the chest wall is used to exhale air down a pressure gradient.

Ensuring adequate tidal volume and minute ventilation is necessary for carbon dioxide elimination. Assuming a constant airway resistance, tidal volume is dependent on the pressure gradient between the alveolus and the atmosphere. Understanding this concept is helpful when manipulating the noninvasive ventilator. Rearranging Eq. (2) to noninvasive parameters yields the following:

Tv and flow are dependent on the level of pressure support provided by the BiPAP. Pressure support equals the difference between IPAP and EPAP. Increasing pressure support, provided that the patient has an adequate respiratory rate (RR), will increase Tv and minute ventilation.

Patient Selection

The appropriate selection of patients with respiratory distress for application of either NIV versus ETI and conventional ventilation is critical to minimize additional morbidity and mortality.6 First, the patient should have a derangement in pulmonary physiology that requires positive pressure respiratory support. Clinically, the patient should have moderate to severe respiratory distress with evidence of tachypnea, accessory muscle use, or paradoxical abdominal muscle use. Supplemental or laboratory evidence of moderate to severe respiratory distress includes a respiratory acidosis (pH <7.35 with a Paco2 >45 mm Hg) or severe hypoxemia (oxygen saturation <92% despite supplemental oxygen or a Pao2/FiO2 ratio [P/F] <300), and can also be used to guide the physician in selecting patients for NIV. However, the physician should interpret laboratory data appropriately given the clinical scenario: a Paco2 of 40 mm Hg and oxygen saturation of 92% in an asthmatic can have a significantly different meaning than the same blood gas value in a patient with a COPD or CHF exacerbation. Second, the patient should have a disease process that is amenable to treatment with NIV and has a high likelihood of reversibility, for example, exacerbations of COPD or CHF. In these patients, NIV should be started as soon as possible to prevent fatigue, further organ dysfunction, and worsening respiratory distress. Finally, contraindications and predictors of failure to NIV, such as apnea or respiratory arrest, medical instability, inability to protect airway or manage secretions, excessive agitation, poor mask fit, or recent upper airway or gastrointestinal surgery, should be absent.7 An exception to this rule is patients that have a Do-Not-Intubate (DNI) order. Acute respiratory distress and failure are often multifactorial, and the emergency medicine clinician is frequently required to treat and stabilize with incomplete patient data. Provided that contraindications are not present, a 1- to 2-hour trial of NIV may be warranted if the clinician deems NIV appropriate.6,7 If there is improvement in the patient's clinical status and blood gas parameters, NIV should be continued or weaned if applicable. However, if the patient's status does not improve, deteriorates, or it is determined that the patient has a disease process that is not amenable to NIV, the plan of action should be reassessed and therapy should be adjusted. In select patients who are DNI or DNR, palliative care measures should be initiated. A major risk of NIV is delaying ETI in patients who require mechanical ventilation, and there is a potential for an increased morbidity and mortality in those patients who “linger” on NIV.8 If there is no improvement after a trial period, or any deterioration, ETI should be conducted to prevent intubation under “crash” conditions.

Patient and Noninvasive Ventilation Interface

There are three basic interfaces or masks that the clinician can use to provide NIV: helmet, nasal mask, and full face mask. Each has its own strengths and weaknesses.

The helmet interface encases the patient's entire head. There is some concern that with the helmet interface patients can rebreathe carbon dioxide, particularly if the ventilator becomes disconnected from the patient.6,9 The helmet interface can also be louder than the oral or nasal interface.6,9 However, there is less risk of skin breakdown and this interface may be more comfortable for prolonged NIV use.6,9 The majority of clinical studies did not use a helmet interface, and the experience with this type of interface in US centers is limited.6

The nasal mask is a partial mask that goes around the nose of the patient. The patient's mouth is not covered. This type of interface is commonly used in chronic conditions such as obstructive sleep apnea. The nasal mask may be more comfortable than the standard full face mask and provokes less claustrophobia. However, it is not particularly suited for the acute setting because of the potential for large air leaks and loss of pressure when the patient opens or breaths through his or her mouth.6,7,9

The most commonly used interface in the emergency department (ED) and critical care setting is the full face mask.6,7,9 The full face mask covers the nasal bridge, goes around the nose, and then forms a seal around the chin and mouth. When using the full face mask, it is important that there is no leak of air around the face mask. An air leak will limit the amount of pressure and volume provided to the patient. Extended use or an overly tight fit can lead to pressure sores on the face. Some patients experience discomfort and a sensation of claustrophobia with the full face mask and positive pressure ventilation.6,7,9 It is often necessary for the clinician to be at the bedside to provide reassurance and make adjustments to the ventilator and face mask. The clinician may decide to provide the patient with analgesia or anxiolysis; however, the negative effects and risks of respiratory and mental status depression must be taken into consideration.

Initial Settings and Patient Monitoring

When choosing initial NIV settings, it is important to consider the patient's underlying disease process, the need for positive pressure support, and patient comfort and compliance. Increased patient compliance at lower pressures must be weighed against improved respiratory mechanics, ventilation, and oxygenation at higher pressures. A common BiPAP starting point for many emergency medicine physicians is to begin with an IPAP of 10 cm H2O and an EPAP of 5 cm H2O—“10 and 5.” This is an acceptable starting point, and lower initial pressures may help facilitate patient compliance. However, “10 and 5” provides a pressure support of about 5 cm H2O, slightly less than or equal to the amount of pressure support utilized in weaning trials. There is a subtle yet important difference in nomenclature when describing pressure support in BiPAP and conventional mechanical ventilators. “10 and 5” on the BiPAP translates to an IPAP of 10 cm H2O and an EPAP of 5 cm H2O. The “10 and 5” on the conventional mechanical ventilator utilized during a pressure support wean is actually “10 over 5”—an inspiratory pressure of 15 cm H2O and expiratory pressure of 5 cm H2O. It is therefore critical that the clinician monitors response and titrates pressures appropriately to ensure adequate gas exchange and decrease work of breathing. Inadequate pressure support or PEEP (EPAP) may increase work of breathing.6,7

When monitoring patients on NIV, it is important to look for certain subjective and objective criteria (see Table 6-1). First, examine the patient–ventilator interface for air leaks. Air leaks may be audible. You can also place your hands around the mask to feel for air leaking. Air leaks decrease the amount of support provided to the patient and can lead to NIV failure. One method to compensate for air leaks is to increase the applied pressures to increase the actual support that the patient receives. However, increasing the applied pressures can lead to increased air leak. The best way to overcome air leaks is to readjust the mask or change to a different interface. Second, the clinician should look at clinical parameters, such as the patient's mental status, use of accessory muscles, comfort level, and presence of subjective dyspnea or chest pain. Worsening mental status can imply a worsening respiratory status, worsening Paco2, and an increased risk of aspiration, and should signal the end of the NIV trial. The patient should report an improvement in dyspnea, and the clinician should look for decreased accessory muscle use and good patient and NIV synchrony. Third, objective clinical data such as Tv on the NIV, RR, heart rate (HR), oxygen saturation (SaO2), and blood pressure (BP) should be monitored continuously, and a baseline arterial blood gas (ABG) and 1- to 2-hour ABG should be obtained to measure pH, Paco2, and Pao2.

Clinical Example: Ventilation and Work of Breathing

To give a clinical example, a patient is placed on BiPAP for hypercarbic respiratory failure secondary to a COPD exacerbation. This particular patient has inadequate ventilation, respiratory acidosis, and elevated work of breathing. The clinician selects a “standard” setting of IPAP 10 cm H2O and EPAP 5 cm H2O, adjusting the FiO2 to maintain SaO2 between 88% and 92%. Additional SaO2 is not required, and can run the risk of inhibiting respiratory drive. An EPAP of 5 cm H2O is a good starting point to decrease the work of breathing needed to overcome the inspiratory threshold created by intrinsic PEEP. An IPAP of 10 cm H2O provides a pressure support of 5 cm H2O. This is minimal, but patient compliance is a concern so the clinician begins with lower pressures. After a trial period, the clinician finds that the patient is not pulling enough Tv on the BiPAP and the Paco2 level is not dropping appropriately. A decision is made to continue the BiPAP and increase pressure support. Rather than increase both the IPAP and EPAP together, the clinician should only increase the IPAP (see Figure 6-3). Increasing the IPAP and EPAP equally together will maintain the same pressure gradient between the alveolus and the ventilator and the same tidal volume and ventilation. By selectively increasing the IPAP, the clinician increases the pressure gradient between the atmosphere and ventilator and therefore will increase the flow of air and tidal volume.

Clinical Example: Hypoxia

In the second clinical example, a patient with a CHF exacerbation and hypoxic respiratory failure is placed on BiPAP. Again the clinician selects a “standard” setting of IPAP 10 cm H2O over EPAP of 5 cm H2O, and applies supplemental FiO2. Initially the clinician will increase the level of FiO2. Often, 100% FiO2 will not completely correct hypoxia. If the patient does not adequately improve, the clinician must improve Pma (see the following equation) and alveolar recruitment to improve V/Q mismatch, reduce shunt, and effectively treat hypoxia.10

This is best accomplished by increasing the EPAP (or PEEP) level of support as this will have a greater impact on the mean airway pressure than increasing pressure support. Both IPAP and EPAP are increased equally together (see Figure 6-4). This will provide the same level of inspiratory support while increasing Pma and oxygenation.

Before examining the different articles studying NIV use, it is important to discuss their validity and application to bedside care. The respective studies, in addition to being conducted in centers with extensive experience with NIV, had strict inclusion and exclusion criteria. The majority of studies excluded patients with hemodynamic instability, multiorgan dysfunction, altered mental status, difficulty in maintaining their airway, or excessive secretions. In addition, the sickest patients were intubated before randomization, and were not included. NIV is appropriate for only a select subgroup of patients with acute respiratory distress.

COPD and Hypercarbic Respiratory Failure

The use of NIV in acute COPD exacerbations and hypercarbic respiratory failure is well supported by multiple clinical trials.11–14 NIV is believed to help improve respiratory mechanics and symptoms in patients with COPD exacerbations through a number of mechanisms. First by providing pressure support, BiPAP can partially offload the work of breathing of the diaphragm and other respiratory muscles. Second, providing extrinsic PEEP in certain scenarios may reduce air trapping and dynamic hyperinflation and overcome intrinsic PEEP, leading to improved pulmonary function. Finally, NIV can decrease the cost of breathing. Normal breathing utilizes approximately 2% of cardiac output; this can increase to 20% in patients with acute respiratory distress. By improving Paco2 and pH, NIV can improve mental status and respiratory muscle function, improving the efficiency of breathing.4

The use of NIV in COPD, as in other disease states, is intended to bridge the patient through the exacerbation until medical therapy can ameliorate and reverse the disease process. For the appropriately selected patient without exclusion criteria, NIV should be considered first-line therapy for the patient with respiratory distress from COPD exacerbations. NIV success rates have been cited near 80–85%.11,12 The beneficial effects of NIV can be attributed to decreasing the complications associated with ETI and conventional mechanical ventilation—for example, oversedation, ICU-associated weakness, ventilator-associated pneumonia (VAP), and pneumothorax.6,7,11–14 In addition, patients on NIV “wean” faster than patients on standard mechanical ventilation.6,7,11–14

In 1995, Brochard et al12 published the results of a randomized multicenter trial on the use of NIV in patients with acute exacerbations of COPD. In this study, the authors compared standard medical therapy alone with a combination of standard medical therapy and NIV in 85 patients with an acute exacerbation of COPD. The NIV group had a lower rate of ETI (26% vs. 74%, P < .001), frequency of complications (16% vs. 48%, P < .001), hospital length of stay (LOS) (23 ± 17 days vs. 35 ± 33 days, P = .005), and in-hospital mortality (9% vs. 29%, P = .02). Importantly, the sickest patients (approximately 30%), who required emergent intubation or were hemodynamically unstable, were excluded from the trial. This study was conducted in an ICU setting.

In 2000, Plant et al13 conducted a similar but larger multicenter trial on the use of NIV in acute exacerbations of COPD. The authors included patients with COPD exacerbations that had tachypnea, hypercarbia, and mild to moderate acidosis (defined as pH 7.25–7.35). This study differed from the earlier study by Brochard in that it was conducted in a general respiratory ward and not in an ICU—still not a busy ED but in theory closer to the functioning and staffing of the ED than an ICU. Plant found similar results to the prior work of Brochard. In 236 randomized patients, NIV was associated with a reduced need for intubation (15% vs. 27%, P = .02), lower in-hospital mortality (10% vs. 20%, P = .05), and a more rapid improvement in breathlessness and RR.13

In general, the beneficial results demonstrated by Brochard and Plant have been reproduced in multiple studies examining the use of NIV in acute COPD exacerbations. Ram et al,14 in 2001 and then again in 2004, conducted a systematic review of the literature for the Cochrane Database on the use of NIV in acute COPD exacerbations. In a pooled analysis, the use of NIV was associated with decreased mortality (RR 0.52; 95% CI 0.35–0.76), decreased need for intubation (RR 0.41; 95% CI 0.33–0.53), reduction in treatment failure (RR 0.48; 95% CI 0.37–0.63), less complications with associated treatment (RR 0.38; 95% CI 0.24–0.60), and shorter LOS (weighted mean difference [WMD] −3.24 days; 95% CI −4.42 to −2.06). In addition, NIV showed beneficial effects on physiologic respiratory parameters such as pH, Paco2, and RR. The authors concluded that data from good-quality randomized controlled trials (RCT) support the benefit of NIV as first-line therapy in conjunction with medical care in suitable patients with respiratory failure secondary to an acute exacerbation of COPD. In addition, they further recommend that NIV should be considered early in the course of respiratory failure, before severe acidosis ensues, as a means of reducing the likelihood of ETI, treatment failure, and mortality.14

The use of NIV in hypercarbic narcosis is controversial. Altered mental status has classically been an exclusion criteria or contraindication to the use of NIV. However, two studies suggest that NIV may be effective in patients with encephalopathy secondary to hypercarbic respiratory failure. Díaz et al15 conducted a prospective observational study comparing the use of NIV in patients with acute hypercapnic respiratory failure and a Glasgow Coma Scale (GCS) ≤8 (n = 76) versus those with a GCS >8 (n = 605). Their group found similar results between the two groups with regards to in-hospital mortality (33.2% in “No Coma” group vs. 26.3% in “Coma” group, P = .17) and success with avoidance of ETI (70.1% in “No Coma” group vs. 80% in “Coma” group, P = .04). In the subgroup with COPD, the results were even more encouraging with 89% of patients without coma avoiding intubation and 86.3% of patients with coma avoiding intubation. Improvement of GCS within 1 hour of therapy predicted NIV success (OR 2.32, 95% CI 1.53–3.53), again highlighting the need for the clinician to reassess the patient's response to NIV, search for objective criteria of success or failure, and appropriately escalate therapy if necessary. The major weakness of the Díaz study is the lack of a control group, but the results suggest that a trial of NIV may be warranted in patients with hypercarbic narcosis and especially in COPD patients with hypercarbic narcosis. There was only one episode of aspiration pneumonia in 76 patients. Scala et al16 conducted a case-controlled study of 80 individuals, matching comatose and noncomatose COPD patients. They found similar results to Díaz's group. However, patients with worsening mental status depression had higher failure rates and higher mortality than matched controls with normal mental status. There were no cases of aspiration in the selected patients. Patients who improved usually did so within the first hour, and the majority of cases of NIV failure were secondary to hemodynamic instability and need for vasopressors. Again the authors suggest that a trial of NIV may be warranted in patients with hypercarbic narcosis.

In 2009, the Global Initiative for Chronic Obstructive Lung Disease (GOLD) recommended NIV for use in COPD exacerbations in patients with moderate to severe disease, defined by dyspnea with use of accessory muscles, tachypnea (RR >25 breaths/min), acidosis (pH ≤7.3), or hypercapnia (Paco2 >45).11 NIV in multiple clinical trials has consistently improved respiratory acidosis and decreased RR, sensation of breathlessness, hospital LOS, complication rate, need for intubation, and mortality. It should be considered in all patients with moderate to severe dyspnea from a COPD exacerbation who do not require immediate intubation. It is probably prudent to institute NIV as early as possible in the disease course. An altered level of consciousness is usually a contraindication to NIV; however, if the depressed mental status is secondary to CO2 retention, the clinician can consider a trial of NIV. The patient should be closely monitored for evidence of hemodynamic instability, worsening mental status, respiratory failure, apnea, and aspiration. If the patient does not improve clinically within 1–2 hours, therapy should be escalated. It is probably prudent to place a nasogastric tube in patients with depressed level of consciousness to decrease gastric distention and the incidence of aspiration, although this statement has been both supported15 and disavowed16 by different authors. Furthermore, a nasogastric tube can worsen an air leak.

Acute Cardiogenic Pulmonary Edema

The use of NIV and CPAP in patients with respiratory distress secondary to acute cardiogenic pulmonary edema (ACPE) is generally well supported in the literature. BiPAP and CPAP are thought to benefit in ACPE by decreasing both preload and afterload, reducing the sensation of breathlessness, decreasing CO2 retention when present, and, when BiPAP is used, decreasing the work of breathing.4 The application of external PEEP increases intrathoracic pressure. This increase in intrathoracic pressure is believed to decrease venous return to right side of the heart, dropping preload and placing the heart on a more favorable part of the Starling curve. Some studies have called this theory into question however. Venous return is based largely on the gradient between right atrial pressure (RAP) and the mean systemic pressure (MSP) of the peripheral circulation. Jellinek et al17 showed that RAP and MSP increased equally with the application of PEEP. The gradient for venous return stayed the same while venous return decreased. Therefore, the drop in venous return that accompanies the application of PEEP is likely related to both an increase in intrathoracic pressure and more complicated interactions between the right side of the heart and the peripheral circulation.4,17

Venous return and cardiac output over any given time period are equivalent. In preload-responsive states, this drop in venous return is more pronounced, leading to a potentially dangerous drop in cardiac output, BP, and tissue perfusion. In conditions where the heart is adequately fluid resuscitated or volume overloaded (e.g., ACPE), this drop in cardiac output is negligible.4

The application of PEEP, via BiPAP or CPAP, is also believed to be beneficial in patients with ACPE through a reduction in afterload. Afterload is the force opposing ventricular contraction. This force is determined by two major variables: systemic arterial resistance and the left ventricular transmural pressure. The left ventricular transmural pressure is equal to the difference between the systolic BP and the intrathoracic pressure.4 In patients with ACPE, the clinician administers nitrates and vasodilators, decreasing afterload by reducing systemic arterial resistance and also reducing transmural pressure by reducing systolic BP. The application of PEEP through BiPAP or CPAP raises intrathoracic pressure that leads to a decrease in left ventricular transmural pressure and left ventricular afterload, leading to a decrease in edema formation in the lungs.4

Numerous studies have compared the use of BiPAP and CPAP in ACPE, with mixed results.

In 2008, Vital et al18 published a systematic review in the Cochrane Database examining the use of BiPAP or CPAP in ACPE. They included a total of 21 trials with 1,071 patients. Their results suggest that BiPAP or CPAP significantly reduced hospital mortality (RR 0.6, 95% CI 0.45–0.84) and ETI (RR 0.53, 95% CI 0.34–0.83). The number needed to treat (NNT) to prevent one death was 13 and to prevent one intubation was 8. ICU LOS was reduced by 1 day (WMD −1.07 days, 95% CI −1.60 to −0.53). No significant increases in acute myocardial infarction with BiPAP during (RR 1.24, 95% CI 0.79–1.95) or after (RR 0.82, 95% CI 0.09–7.54) its application were observed.

The validity of the systematic review in critical care medicine has recently been called into question.19 Lumping together populations from multiple studies will increase the sample size, tighten the confidence interval, and decrease the chance of random error, but unfortunately these populations are often heterogeneous. Analyzing heterogeneous study populations together can serve to amplify systematic errors, threatening accuracy or usefulness of the evidence for the clinician at the patient's bedside.19 Furthermore, meta-analyses fail to predict the outcomes of future large multicenter randomized trials 35% of the time.20

In 2008, Gray et al21 published the Three Interventions in Cardiogenic Pulmonary Oedema (3CPO) trial, examining NIV in cardiogenic pulmonary edema. This trial was a large, randomized multicenter trial involving 1,069 ED patients with ACPE. Patients were randomized to standard care (supplemental oxygen) versus CPAP or BiPAP. The 3CPO trial found that CPAP and BiPAP were better than standard care in relieving dyspnea, improving HR, RR, hypercapnia, and acidosis at 1 hour. Unfortunately, these benefits did not translate into a lower hospital mortality rate (9.8% vs. 9.5%, P = .87) or intubation rate (2.8% vs. 2.9%, P = .90). CPAP and BiPAP were equally effective, and there was no associated increase in the incidence of acute myocardial infarction with BiPAP (27.2% vs. 26.8%, P = .90). The authors recommend that NIV be used as an adjunctive therapy in patients with ACPE with severe respiratory distress or in patients who do not improve with standard pharmacologic therapy (nitrates, diuretics, and possibly afterload-reducing agents).

There are some important critical elements and controversies to point out about the 3CPO trial. First, the sickest patients, those who required immediate life-saving interventions (i.e., intubation), were excluded. This exclusion possibly selected a “healthier” group of patients. The intubation rate in the 3CPO trial was approximately 3%.21 In all other trials besides 3CPO, the intubation rate was around 27% in the control groups.17,22 It is harder to find a difference in outcomes (i.e., intubation and mortality) when these outcomes are rare to begin with.

Second, this trial was conducted as an open trial with an “intention-to-treat” analysis. Briefly, intention to treat means that regardless of the actual treatment given, a particular patient is analyzed in the group to which he or she is randomized. With regards to this trial, if a patient was randomized to receive standard care, but shortly after randomization was deemed to need additional support and crossed over to an NIV arm, he or she was still analyzed in the standard care arm (and vice versa). From a research standpoint, the intention-to-treat study design is critical to protect the randomization process but from a practical standpoint it can make the results difficult to interpret and apply at the patient's bedside. Researchers attempt to minimize confounding results from crossover among treatment groups. This is not always practically or ethically possible. In the 3CPO trial, between 15% and 24% of patients did not complete the trial in the group they were randomly assigned.21 This high rate of crossover serves as a confounding variable because some patients treated with standard medical treatment likely benefited from NIV application, yet were analyzed as successes in the standard treatment group. Approximately 15.5% of the patients treated with standard therapy were put on NIV because of worsening respiratory distress or deterioration in blood gas values.21 Similarly, 5.2–8.4% of patients treated with NIV did not complete the trial because of discomfort with the mask and ventilator.21 A major determinant of NIV success or failure is patient compliance and tolerance of treatment.

The aggregated data on the use of NIV in ACPE suggest that there is a definite benefit in terms of physiologic and clinical parameters. This seems to support its use in the care of patients with ACPE. However, this does not necessarily translate statistically into a reduction in mortality and the reduced need for intubation. Weng et al22 conducted another systematic review and included the data of the 3CPO trial investigators. They concluded that even with the equivocal results of the 3CPO trial, the previous assessments that the use of CPAP reduces mortality and intubation rates in patients with ACPE and that BiPAP reduces the need for intubation compared with standard therapy still appear to be true.

Perhaps the most useful information from these, at times, conflicting data is that the use of NIV in ACPE does not cause harm. One of the first studies published by Mehta et al23 in 1997 found an increased rate of myocardial infarction in the BiPAP group (71%) when compared with the CPAP group (31%), leading to an early termination of their study. However, the patients in the BiPAP group had a higher percentage of chest and jaw pain, and likely presented with myocardial infarction rather than developing one on BiPAP. In addition, this study had a small sample size. The 3CPO trial21 and multiple systematic reviews18,22 did not find an increased incidence of myocardial infarction with the use of NIV, particularly BiPAP.

The question of whether to use CPAP or BiPAP in ACPE is controversial. From a purely statistical standpoint, it would appear that CPAP has a slight advantage. From the studies of Vital et al18 and Weng et al,22 CPAP was associated with a statistically significant decrease in hospital mortality and reduced need for intubation. BiPAP was associated with a less robust decrease in mortality and intubation.18, 22 From a physiologic standpoint, BiPAP would seem to be more logical. BiPAP provides all the benefits of CPAP as the EPAP is essentially the same as CPAP. Plus BiPAP better offloads the respiratory muscles and work of breathing, better augments ventilation and improves hypercarbia and respiratory acidosis when present, and better improves dyspnea, HR, hypoxemia, and RR.18, 21 When compared directly with one another, CPAP and BiPAP perform identically in terms of mortality and need for intubation.21,24 The difference appears to be statistical rather than actual. For ACPE, the use of CPAP and BiPAP appears to be equally effective. However, patients with coexisting COPD or any degree of hypercapnia or respiratory acidosis may benefit more from BiPAP over CPAP.

NIV, in addition to aggressive medical management, should be considered first-line therapy for the appropriately selected patient with ACPE. The patient should not be hypotensive, in shock, or otherwise hemodynamically unstable. In addition, active ischemia, STEMI requiring emergent intervention, or an unstable arrhythmia should not be present. These patients are better managed by ETI and mechanical ventilation. The BiPAP should be set with an emphasis on a higher EPAP to increase mean airway pressure and lung recruitment that can be used to improve oxygenation, along with supplemental oxygen. Alternatively, CPAP can be used alone. Pressure support, if used, should be adjusted by raising the IPAP to decrease the work of breathing and improve ventilation, hypercapnia, and respiratory acidosis as needed. As with all patients on NIV, they should be monitored closely for signs of NIV failure, and the clinician should be prepared to escalate care and perform ETI if necessary. Objective data from ABGs, and other clinical parameters such as HR, RR, SaO2, and dyspnea, should be frequently reassessed. After 1–2 hours of NIV, clinical improvement should be evident. If there is no improvement, or deterioration occurs at any time, intubation should strongly be considered.

The Immunocompromised Patient

ETI and conventional mechanical ventilation are associated with a risk of VAP and other nosocomial infections. VAP can have a 20–50% mortality rate.25 NIV has been shown to benefit immunocompromised patients by reducing mortality, mainly secondary to prevention of VAP and other nosocomial infections and complications such as pneumothorax.

In a case-controlled study looking at NIV in 48 acquired immunodeficiency syndrome (AIDS) patients admitted to the ICU with acute respiratory failure secondary to pneumocystis carinii pneumonia (PCP), Confalonieri et al26 showed a decrease in ICU mortality in patients treated with NIV (75% vs. 38%), a reduction in the need for intubation in the NIV group (67% avoided intubation), and a decreased ICU LOS (7 ± 4 days vs. 10 ± 4 days) as compared with controls who were intubated on presentation of acute respiratory failure. The retrospective nature of the study makes it difficult to draw complete conclusions. It is very possible that those initially intubated were sicker and destined to do worse, but the fact that intubation was prevented in approximately two thirds of the patients treated with NIV suggests that a trial of NIV is warranted in AIDS patients with acute respiratory failure secondary to PCP pneumonia. Hilbert et al27 compared NIV with standard treatment (supplemental oxygen) in 52 patients with evidence of immunocompromise, acute hypoxic respiratory failure, pulmonary infiltrates, and fever. They found a decrease in the need for intubation (12 vs. 20, P = .03), fewer serious complications (13 vs. 21, P = .02), fewer ICU deaths (10 vs. 18, P 50.03), and fewer hospital deaths (13 vs. 21, P = .02). Again, it is important to emphasize that this study did not compare NIV with intubation, but rather NIV with standard oxygen therapy. The benefits of NIV may be most pronounced when used early in the disease process. All of the patients who were intubated died, regardless of initial treatment, highlighting the severity of underlying illness and poor prognosis associated with respiratory failure in the immunocompromised population. Similar results were found by Antonelli et al28 in immunocompromised patients with acute respiratory failure after solid organ transplantation. In 51 patients with acute respiratory failure after solid organ transplantation, the use of NIV was associated with a significant reduction in the rate of ETI (20% vs. 70%, P = .002), a lower rate of fatal complications (20% vs. 50%, P = .05), decreased LOS in the intensive care unit by survivors (mean [SD] days, 5.5 [3] vs. 9 [4], P = .03), and lower intensive care unit mortality (20% vs. 50%, P = .05). An initial trial of NIV is warranted in immunocompromised patients with acute respiratory failure as it may prevent ETI and the complications associated with standard mechanical ventilation.

Asthma

Clinicians can consider the use of NIV for the treatment of acute asthma exacerbations. However, the evidence behind the use of NIV for acute asthma exacerbations is not as strong as other disease states such as COPD or ACPE. Asthma-induced acute respiratory failure is secondary to airflow obstruction and dynamic hyperinflation, leading to increased intrinsic PEEP and an increased work of breathing. An early study by Meduri et al29 studied the use of NIV in 17 patients with status asthmaticus admitted to the intensive care unit, and showed that NIV improved clinical and physiologic parameters (pH, Paco2, RR). Only two of the patients required intubation. This was a case series and was therefore unable to show if NIV can prevent intubation. However, it did demonstrate that a trial of NIV may be safe in status asthmaticus. In a study conducted in an ED, Soroksky et al30 compared BiPAP with medical therapy alone in 30 patients with severe asthma exacerbations. They found a reduction in the need for hospital admission, more rapid improvement in FEV1, and a greater percentage of patients who improved their FEV1 to greater than 50%. This study was mainly limited by size, and only patients with moderate to severe disease were included. It is difficult to apply these data to status asthmaticus patients with acute respiratory failure because there were no intubations or morbidity in either group. Except for FEV1 and PEFR, other clinically important variables such as ICU admission and ICU LOS were not commented on. However, applying NIV early as opposed to waiting for ICU admission in status asthmaticus patients may explain the very positive results of this trial. A Cochrane meta-analysis by Ram et al31 found the use of NIV in asthma exacerbations to be “promising” but also still “controversial” because of lack of evidence. The meta-analysis included only the aforementioned Soroksky study, as no other trials were adequately designed for inclusion in this analysis. More evidence is needed to develop evidenced-based recommendations, but it is generally agreed that a trial of NIV is warranted in select patients with severe asthma exacerbation provided that there are no contraindications present.32 Again it must be emphasized that NIV is not a treatment for status asthmaticus, but rather a means of respiratory support. The clinician should optimize medical treatment and aggressively treat the patient with β-agonists, corticosteroids, and magnesium sulfate. Intubation and conventional mechanical ventilation in the status asthmaticus patient are sometimes necessary. If a trial of NIV is to be attempted, it should be done early and the patient should be monitored closely. If there are signs of worsening fatigue or respiratory failure, support should be escalated. Intubation should strongly be considered if the patient fails the NIV trial.32

Pneumonia

The use of NIV in pneumonia is controversial, but in a carefully selected patient, a trial of NIV may be attempted. Even with optimal antibiotic therapy and aggressive medical management, the course of pneumonia is usually not rapidly correctable. Therefore, standard mechanical ventilation may be a better option than NIV. Multiple studies have looked at acute respiratory failure and pneumonia either directly or via subgroup analysis, often with differing results.

One of the larger studies looking specifically at pneumonia patients, conducted by Confalonieri et al,33 compared 56 patients with acute respiratory failure secondary to severe community-acquired pneumonia (CAP) with NIV versus conventional therapy. They found a decreased need for ETI (21% vs. 50%, P = .03) and duration of ICU stay (1.8 ± 0.7 days vs. 6 ± 1.8 days, P = .04). However, there was no decrease in ICU mortality or 60-day mortality. Furthermore, the applicability of this benefit may be limited to patients with COPD. A post hoc analysis suggested that the subset of CAP patients with underlying COPD benefited the most from a trial of NIV. In this study, patients with COPD treated with NIV versus conventional therapy had a decreased need for intubation (0% vs. 55.5%), duration of ICU LOS (0.25 ± 2.1 days vs. 7.6 ± 2.2 days, P = .02), and 60-day mortality (11.1% vs. 62.5%, P = .05). A reduction in intubation, duration of ICU stay, and mortality was not seen in patients without COPD.

In an observational study looking at NIV in 24 patients with severe CAP, Jolliet et al34 found that NIV was associated with a moderate improvement of P/F and a decrease in RR. A large proportion (66%) of patients eventually required intubation and eight of those patients expired. Likewise, those patients who did not require intubation had a shorter ICU stay (6 days vs. 16 days) and hospital stay (9.5 days vs. 23 days).

Studies looking at all-comers with hypoxic respiratory failure have found conflicting results with regards to the use of NIV in patients with pneumonia. In a study of 64 patients with acute respiratory failure from multiple different causes by Honrubia et al,35 a small subset of 8 patients with pneumonia all failed an NIV trial and required mechanical ventilation. In the Antonelli et al36 study of 354 patients with acute hypoxemic respiratory failure, the presence of CAP was an independent predictor of NIV failure, with 50% requiring intubation. Ferrer et al,37 looking at the use of NIV versus conventional therapy in 105 patients with acute hypoxemic respiratory failure, showed a decreased need for intubation (26% vs. 73%, P = .017) and ICU mortality (15.7% vs. 53%, P = .030) in a subset of 34 patients with acute hypoxemic respiratory failure from pneumonia.

The evidence surrounding the use of NIV in acute hypoxemic respiratory failure secondary to pneumonia does not provide a clear signal to the clinician. Although patients who avoid intubation do better, this effect may not be related to the efficacy of NIV but may rather suggest the obvious—sicker patients do worse. To complicate matters, with the exception of the sickest pneumonia patients, it may be difficult to initially differentiate the subset of pneumonia patients who may benefit from a trial of NIV from those who should be immediately intubated. The American Thoracic Society (ATS) and Infectious Disease Society of America (IDSA) in their 2007 guidelines for the management of CAP recommend a “cautious trial” of NIV in patients with CAP who have signs of respiratory distress and/or hypoxemia unless they are candidates for immediate intubation as evidenced by severe disease, bilateral infiltrates, or a P/F ≤150.38 The ATS/IDSA guidelines note that severe CAP and acute respiratory distress syndrome (ARDS) may be hard to distinguish clinically early in the disease process, and NIV has poor efficacy in ARDS (see below), further decreasing the benefit of NIV in these patients.38

A short trial (1 or 2 hours) looking objectively for signs of improvement or failure of NIV (fatigue, accessory muscle use, change in Pao2, and Paco2) may be warranted in the appropriately selected patient with pneumonia. Similar to COPD and CHF, NIV may be most useful in pneumonia patients earlier in the disease course. The clinician should anticipate that there is a possibility that the patient with CAP will fail his or her NIV trail and should take the appropriate steps to be prepared to endotracheally intubate the patient in the event of decompensation or failure to improve. However, if intubation can be avoided, there is a higher likelihood of favorable outcomes and decreased associated morbidity and health care costs. Again, patients with evidence of severe disease or any evidence of hemodynamic instability, septic shock, or additional nonpulmonary organ failure should not be considered for NIV.

ARDS/ALI

The use of NIV in ARDS and acute lung injury (ALI) draws many parallels to the previously discussed use of NIV in pneumonia. Similar to pneumonia, the evidence for the benefit of NIV in ARDS and ALI is not straightforward. Studies looking at the use of NIV in ARDS and ALI have shown rates of intubation or failure of NIV ranging from 46% to 85%.36,37,39,40 NIV may not be as efficacious at preventing intubation in ARDS/ALI when compared with COPD, CHF, and other disease processes. With regards to the following research studies, all included patients were relatively stable and not in shock. In addition, the sickest patients were endotracheally intubated immediately, before they had a chance of enrolling in the respective studies. Looking at patients with ARDS in multiple studies, NIV could be applied to only about 30% of patients with ARDS, and only succeeded in half, or about 16%, of total ARDS patients. Therefore, it should be clear that the use of NIV in ARDS applies to only a very specific, and small, subset of the ARDS population.

Looking at 354 patients with acute hypoxemic respiratory failure, and in particular 86 patients with ARDS, Antonelli et al36 found the presence of ARDS and P/F ≤146 to be predictors of NIV failure. Patients with pulmonary and nonpulmonary causes of ARDS required intubation about 46% and 54% of the time, respectively. Stated differently, NIV was successful in preventing intubation in ARDS patients approximately half of the time. Half of the ARDS patients who avoided intubation survived versus 9.5% of ARDS patients who were intubated. It is difficult to make conclusions about mortality benefit from this study because of its uncontrolled nature; rather the mortality difference may simply prove that sicker patients do worse.

In the study of Ferrer et al37 of 105 patients with acute hypoxemic respiratory failure, 15 patients met the criteria for ARDS. Nearly all of the patients who required intubation, in both the NIV and supplemental oxygen groups, expired. The results from both groups were equally dismal from a clinical standpoint and not statistically different. It is difficult to draw useful conclusions from such a small group of patients; however, these results suggest that NIV is less helpful in the treatment of ARDS than other disease states.

Rana et al in 200639 conducted an observational cohort study examining the use of NIV in patients with ALI. In analysis of 54 patients, 70.3% failed therapy with NIV and required intubation and standard mechanical ventilation. All 19 patients in shock failed the NIV trial. Additionally, patients with severe hypoxemia as evidenced by a median P/F <112 (70–157) and metabolic acidosis (base excess −4.0, range −7 to 0.2) had a higher incidence of NIV failure. While there was no control group in this study, it seems prudent to bypass an NIV trial and proceed directly to ETI and standard mechanical ventilation in patients with evidence of shock and metabolic acidosis.40

Antonelli et al41 in 2007 looked at the use of NIV as first-line therapy in 147 patients with ARDS who were not intubated. NIV was successful in avoiding intubation in 79 patients (54%). Multivariate analysis showed that a Simplified Acute Physiology Score (SAPS II) >34 (OR 3.6, 95% CI 1.66–7.7) and a P/F ≤175 (OR 2.34, 95% CI 1.1–5.15) after 1 hour of NIV were independently associated with NIV failure and the need for ETI. The ICU mortality rate was 28%, but ICU mortality was significantly higher in those who required ETI, 5% versus 36% (OR 21, 95% CI 6.4–76.5, P < .001). Patients requiring intubation developed severe sepsis or septic shock and VAP more often. Mortality was higher in those who failed the trial of NIV and required intubation (54% vs. 19%, P < .01). The authors recommend that patients with less severe disease (SAPS II <34) and with a P/F >175 after 1 hour of NIV will likely benefit from continuation of NIV, but those who fail to show substantial improvement in oxygenation after a 1-hour trial of NIV should be closely monitored with a low threshold for ETI.

In conclusion, there are two important factors that the clinician must consider before placing a patient with ARDS/ALI on NIV. First, patients with severe disease as evidenced by a severe hypoxemia (P/F ≤ 150), additional nonpulmonary organ failure, or hemodynamic instability with the need for vasopressors or significant fluid resuscitation should not be considered candidates for NIV as they have a very high likelihood of failure and may do worse with a trial of NIV.40 Second, if the patient does not show improvement within 1–2 hours of NIV, it is important that the clinician not permit the patient to linger on NIV, but rather escalate therapy.40 A cautious trial of NIV can be attempted in a certain subset of ARDS/ALI with less severe disease that does not have any of the aforementioned high-risk factors for failure.

Do-Not-Intubate

When considering the use of NIV in patients with a DNI, it is important for the clinician to have an understanding of the patient's goals of care as well as the underlying disease process responsible for respiratory distress. Some patients and their families may be amenable to an attempt of NIV; however, some may consider NIV to be unnecessary “life support” that merely prolongs suffering. The treating physician should explain the risks and benefits of NIV and determine the patient's and family members' wishes. NIV may alleviate dyspnea and air hunger or may worsen discomfort. As with any other patient placed on a trial of NIV, the clinician should frequently reassess the patient for evidence of NIV failure or success. In DNI patients, failure of NIV trial warrants escalation of opiate and other palliative measures in conjunction with the patient's and family members' wishes.

Similar to other patients placed on NIV, DNI patients with more reversible disease processes causing respiratory distress (i.e., COPD and CHF exacerbations) have more success with NIV than disease states such as pneumonia or ARDS.

Levy et al42 in 2004 looked at the use of NIV in 114 patients with DNI status. Forty-three percent of these patients survived to hospital discharge. The presence of a strong cough (OR 0.16, 95% CI 0.05–0.51), being awake (OR 0.18, 95% CI 0.05–0.62), having a high baseline Paco2 (OR 0.01, 95% CI 0.01–0.93), having COPD (OR 0.31, 95% CI 0.10–0.90), or having CHF (OR 0.14, 95% CI 0.02–0.75) as the underlying cause of respiratory failure had better outcomes in terms of hospital mortality. Patients with CHF and COPD survived to hospital discharge approximately 75% and 50% of the time, respectively. Patients with pneumonia, cancer, and other diagnosis did poorly, with less than 30% surviving to hospital discharge.

Schettino et al43 in 2005 found similar results in their observational trial of NIV in 131 patients with DNI status. Patients treated with NIV for COPD exacerbations had a hospital mortality rate of 37.5%, and those treated for an exacerbation of ACPE had a hospital mortality rate of 39%. However, patients had significantly higher mortality rates when treated with NIV for other conditions such as non-COPD hypercapnic respiratory failure (68%), postextubation respiratory failure (77%), advanced cancer (85%), and hypoxemic respiratory failure (86%). In addition, Schettino et al43 found that a baseline albumin ≤2.5 g/dL or a SAPS II score >35 also predicted mortality.

In 2007, the Society of Critical Care Medicine Palliative Noninvasive Positive Pressure Ventilation Task Force44 put forth a stratification system for the use of NIV in DNI and palliative care patients. They proposed three broad classifications of patients with acute or chronic respiratory failure that NIV can be used in: (1) patients without preset limits for life support, (2) patients with a preset limit for life support (i.e., a DNI order), or (3) patients who desire comfort care only. Each separate category should have different goals of care, definitions of success, type of escalation in case of failure, and the appropriate clinical setting where NIV may be used.

The first category, patients without preset limits for life support, is the standard group of critical care patients with respiratory failure who have no restrictions on care and therefore require all appropriate life-sustaining measures. In this group, the goal of care is to restore health. NIV can be used for these patients as a means to assist with ventilation and oxygenation as well as to prevent intubation. If NIV fails, patients in this category should be endotracheally intubated. According to the task force, these patients should be cared for in an ICU or step-down unit setting.44

The second category, patients who request life support but place a limit on care (in this particular situation ETI), comprises patients with respiratory failure where the use of NIV may prove to be beneficial. The goal in this group of patients is also to restore health if possible, with a secondary goal of minimizing discomfort. If NIV fails in this category of patients, palliative measures should be initiated, and NIV should be discontinued. These patients should also be cared for in an ICU or step-down unit setting; however, local institutional practice and resource availability should be taken into consideration.44

The final category involves patients who desire comfort care and palliation of symptoms. The use of NIV in this group is controversial, and there are little data to support the use of NIV in this situation.44 However, NIV may be helpful in improving dyspnea and cognition. In these situations, the use of NIV can be discussed with the patient and family members. If NIV fails to improve dyspnea, or the patient loses consciousness, NIV should be discontinued as other palliative measures are pursued. These patients can be cared for in an ICU or step-down setting, but a more appropriate locale is probably a hospice or palliative care unit with properly trained personnel.

The delineation of these categories is helpful when discussing treatment options and goals of care with patients and family members in order to ensure that the proposed treatment is congruent with their wishes.

In conclusion, the use of NIV can be considered in select patients who have a DNI status. An honest discussion should take place between the clinician and patient or family members to ensure that this level of treatment is acceptable. Patients who have respiratory failure secondary to a COPD or CHF exacerbation are most likely to benefit from NIV. As with all patients on NIV, they should be closely monitored for signs of NIV failure. If NIV failure occurs, care should focus on comfort and NIV should be discontinued.

The use of NIV to assist in the treatment of ED patients with acute respiratory failure has increased in the past decade. Patients should be appropriately selected for NIV—individuals who are hemodynamically unstable or are unable to protect their airway are not candidates for NIV. NIV works best in disease processes that are easily reversible. Good clinical evidence supports the use of NIV in exacerbations of COPD, ACPE, and acute respiratory failure in the immunocompromised, with weaker evidence for its use in asthma patients. A trial of NIV may be attempted in a very select subgroup of patients with hypoxemic respiratory failure secondary to pneumonia or ARDS, but there is a higher rate of failure. Regardless of the etiology of respiratory distress, aggressive medical management should be used in tandem with NIV. Patients on NIV should be monitored closely for signs of improvement or failure, and NIV therapy should be titrated appropriately to assist the patient with work of breathing, oxygenation, and ventilation. If improvement is not apparent within 1–2 hours, or any deterioration occurs, the patient should be endotracheally intubated, placed on conventional mechanical ventilation, and not allowed to linger on NIV. Patients placed on NIV should be admitted to an ICU or a respiratory care unit depending on disease etiology and severity, as well as local hospital resources and experience.

The authors would like to thank Lauren Houdek, B.S., David Ozimek, B.A., Elizabeth Barton, B.A., Richard Tan, B.A., and Sue Hahn, B.S., for their help with the creation of the manuscript.