Special Feature

Noninvasive Positive-Pressure Ventilation: Current Evidence Supporting Practice

By Dean R. Hess PhD RRT

There currently exists much evidence to direct the application of Noninvasive Positive Pressure Ventilation (NPPV). NPPV for the treatment of patients with acute respiratory failure has generated several meta-analyses and systematic reviews,1-4 including a recent one by me.5 A prospective survey6 for 3 weeks in 42 intensive care units found that NPPV was used as first-line therapy in 16% of mechanically ventilated patients. The first meta-analysis of NPPV was published by Keenan et al and concluded that the use of NPPV in acute respiratory failure affords a beneficial effect on survival and decreases the need for endotracheal intubation. Moreover, the survival advantage was greatest in patients presenting with an exacerbation of chronic obstructive pulmonary disease (COPD). Peter et al,2 in their meta-analysis, concluded that there was a substantial reduction in mortality and the need for intubation with NPPV in acute respiratory failure—especially in the COPD subgroup. Lightowler et al3 conducted a systematic review and meta-analysis restricted to the use of NPPV for COPD exacerbation and reported that NPPV significantly reduced the risk of mortality, the risk of endotracheal intubation, complications of treatment, and length of stay in hospital (Figure 1, below). In another meta-analysis by Keenan et al,4 they concluded that patients with severe COPD exacerbations benefit from the addition of NPPV to standard therapy.

NPPV for COPD Exacerbation

The strongest evidence for use of NPPV is for patients with COPD exacerbation.5 A number of studies reported the use of NPPV only in patients with COPD and report benefit in this patient population with the exception of those with mild exacerbation. The use of NPPV for patients with COPD exacerbation is now considered a standard of care, the evidence for which is established in 2 meta-analyses.3,4

NPPV for Acute Asthma

Compared to COPD, considerably less evidence exists for the use of NPPV in patients with asthma. Soroksky et al7 reported the results of a randomized, controlled trial of NPPV for asthma exacerbation. Hospitalization was required for 3 of 17 patients (17.6%) randomized to NPPV compared to 10 of 16 patients (62.5%) in the control group (P = 0.01). Soroksky et al concluded that, in selected patients with a severe asthma, the addition of NPPV to conventional treatment can improve lung function, alleviate the exacerbation faster, and reduce the need for hospitalization. Before recommendations can be made regarding the use of NPPV in the treatment of asthma exacerbation, additional studies will be needed with larger sample sizes.

NPPV for Hypoxemic Respiratory Failure

An area of considerable controversy is the role of NPPV in patients with hypoxemia who are not hypercapnic. Unlike COPD, hypoxemic respiratory failure is a heterogeneous group of diagnoses. Five randomized controlled trials have reported success with the use of NPPV in patients with acute hypoxemic respiratory failure.5 Evolving evidence supports the use of this therapy in such patients, albeit with evidence less compelling than that for COPD. As shown in Figure 2, below, intubation rate and mortality were lower in the patients receiving NPPV in each study evaluating its use in patients with acute hypoxemic respiratory failure.

Cardiogenic Pulmonary Edema

Another area of controversy is the use of NPPV in the treatment of patients with acute cardiogenic pulmonary edema (CPE). Pang et al8 conducted a systematic review of continuous positive airway pressure (CPAP) on mortality and the need for intubation in cardiogenic pulmonary edema. CPAP was associated with a decreased need for intubation (risk difference 26%; 95% confidence interval, 13 to 38%) and a trend towards a decreased hospital mortality (risk difference, 6.6%; 95% confidence interval, 3 to 16%) compared with standard therapy. Evidence was also lacking on the potential for harm associated to the use of CPAP in this patient population.

In contrast to the situation with CPAP, insufficient high-level evidence exists to allow recommendation of NPPV in the treatment of patients with acute CPE. Of concern is the risk for harm, with 2 studies reporting significantly greater rates of myocardial infarction in patients treated with NPPV.9,10 Moreover, there is no clear evidence of benefit in terms of intubation rate or mortality with the use of NPPV. Subgroup analysis, however, suggests a benefit for the use of NPPV in hypercapnic patients with acute CPE.11 Given the high level of evidence supporting the use of CPAP in this patient population, it seems reasonable to recommend the use of CPAP in hypoxemic patients with acute CPE, and to reserve NPPV for those who are also hypercapnic.


The role of NPPV in the peri-extubation period remains to the determined. Only 4 randomized, controlled trials evaluated the role of NPPV to allow earlier extubation and only 2 of those were positive.5 The positive trials might relate to their enrollment of patients with COPD. Thus, one might consider the use of NPPV to facilitate earlier extubation of patients with COPD. Both randomized trials of NPPV for patients who failed planned extubation were negative,12,13 suggesting a limited role of NPPV in this setting. In the most recent study13 examining the role of NPPV in patients with failed extubation, the mortality rate was higher among patients assigned to NPPV than among those assigned to standard medical therapy, and the interval from the development of respiratory failure to reintubation was significantly longer with NPPV than with standard therapy.

Noninvasive Ventilation and Nosocomial Pneumonia

It is now well accepted that nosocomial pneumonia in mechanically ventilated patients is due to aspiration of pharyngeal secretions rather than to what is breathed from the ventilator. It follows that the risk of nosocomial pneumonia should be decreased if mechanical ventilation is provided with NPPV rather than through an endotracheal tube. The incidence of nosocomial pneumonia with NPPV has been compared with invasive mechanical ventilation in 7 studies and every one of them reported that the rate of nosocomial pneumonia was lower with NPPV. The combined risk of pneumonia is significantly reduced with the use of NPPV (see Figure 3).

Predictors of Success

NPPV is not universally successful in the avoidance of intubation. Although reported success rates vary, 25% or more of patients with acute respiratory failure who receive NPPV require intubation. It is useful to identify patients who have a higher likelihood for NPPV failure so that this may be anticipated and endotracheal intubation performed promptly if necessary. Predictors of NPPV failure include higher APACHE score, lower level of consciousness, lower pH, more air leak around the interface, greater quantity of secretions, poor initial response of NPPV, and the presence of pneumonia.5 NPPV may be least likely to be successful in patients who are most sick. This should not dictate that some patients should not receive a trial of NPPV, but should provide a lower threshold for intubation knowing that these patients have a high likelihood to failure of NPPV.

The Interface for NPPV

Both nasal and oronasal interfaces have been applied successfully in randomized controlled trials. An oronasal interface may be more effective and better tolerated than the nasal interface for patients with acute respiratory failure.14 An issue related to the interface and headgear is facial skin breakdown. The use of a mask of proper size, avoiding placing the headgear too tightly, and the use of wound-care tape on the bridge of the nose are important considerations to avoid facial skin breakdown.

A relatively new interface for application of NPPV is the helmet, which fits over the entire head of the patient and fits snugly around the neck.15 Potential advantages of this design include ability of the patient to interact with the environment, a fixation system that should have a lower risk of skin breakdown, and it can be applied to any patient regardless of facial contour. Concerns with this interface include the risk for rebreathing, and effective triggering and cycling of the ventilator. Until these issues are resolved, the helmet cannot be recommended for the treatment of hypercapnic respiratory failure with NPPV.

The Ventilator for NPPV

Although any ventilator can be used to provide NPPV, the most commonly used are the bilevel or BiPAP® ventilators. These ventilators are designed to operate in the presence of a leak. They typically apply an inspiratory positive airway pressure (IPAP) and an expiratory positive airway pressure (EPAP). Breaths are triggered by the patient and the difference between the IPAP and EPAP is the level of pressure support. They are blower devices that use a single limb circuit. There is no exhalation valve, with the fixed leak in the circuit serving as the exhalation port. The available evidence suggests that these ventilators trigger and cycle as well as, and sometimes better than, critical care ventilators.5

An issue that has received considerable attention with the portable pressure ventilators is the potential for rebreathing.16 Although a potential for rebreathing is present with bilevel ventilators, this can be minimized if the leak port is in the mask rather than the hose, if oxygen is titrated into the mask rather than the hose, if a higher level of EPAP is used, and with the use of a plateau exhalation valve. Anything that increases the leak increases the flow through the hose and more effectively flushes the hose and decreases the amount of rebreathing.

Most bilevel ventilators, however, do not have an oxygen control and thus supplemental oxygen is usually administered by adding it into the mask or the circuit. When administering oxygen with a portable pressure ventilator that does not have an oxygen blender, the delivered oxygen concentration is affected by oxygen flow, the site where oxygen in added into the circuit, the position of the leak port, the type of leak port, the amount of leak (intentional and non-intentional), and the IPAP and EPAP settings.17 Due to the complex interaction between these variables, pulse oximetry should be used to monitor oxygenation when using this therapy in patients with acute respiratory failure.

There is controversy related to the need for humidification during NPPV. Unlike invasive mechanical ventilation, the upper airway is not bypassed with NPPV. When a bilevel ventilator is used, much of the delivered gas is from the surrounding room (except for the supplemental oxygen) and should thus have the same humidity content that the patient would breathe if not receiving NPPV. Anecdotally, I have found greater patient comfort, greater compliance, and less upper airway drying when NPPV is used with a humidifier. Aerosolized bronchodilators can be effectively delivered during NPPV and thus there is no need to temporarily interrupt NPPV to administer these. The evidence for use of MDI during NPPV is not as strong as that for use of nebulizer, but the available evidence suggests that MDI can be used effectively during NPPV.5

Several studies have evaluated the combination of heliox with NPPV in patients with COPD.5 However, further work is needed before a recommendation can be made regarding heliox administration during NPPV. Several short-term studies suggest physiologic benefit when heliox is combined with NPPV in patients with COPD exacerbation.18 But there is only one randomized controlled study that assessed outcomes such an intubation rate and mortality and the results of that study were inconclusive. Of concern is the potential for ventilator malfunction when used with heliox.

Clinical Application

Incorporation of NPPV into usual clinical practice requires a concerted effort by physicians, respiratory therapists, and nurses. A suggested approach to initiation of NPPV is as follows:

  • Determine that patient is a good candidate for NPPV (eg, COPD exacerbation);
  • Choose a ventilator capable of meeting patient needs;
  • Choose the correct interface; avoid mask that is too large;
  • Explain therapy to the patient;
  • Silence alarms and choose low settings;
  • Initiate NPPV while holding mask in place;
  • Secure mask, avoiding a tight fit;
  • Titrate pressure support (IPAP) to patient comfort;
  • Titrate FIO2 to SpO2 > 90%
  • Avoid peak inspiratory pressure > 20 cm H2O
  • Titrate PEEP (CPAP/EPAP) per trigger effort and SpO2
  • Continue to coach and reassure patient; make adjustments to improve patient compliance


High-level evidence exists for the use of NPPV in patients with COPD exacerbation. This form of ventilatory support has also been used successfully in selected patients with acute hypoxemic respiratory failure and to allow earlier extubation of mechanically ventilated patients following COPD exacerbation. The role of NPPV in patients with acute cardiogenic pulmonary edema is inconclusive. Both nasal and oronasal interfaces have been used successfully with NPPV, although the oronasal interface is often preferred for acute respiratory failure. Bilevel ventilators with the pressure support mode are most commonly used for NPPV, although any ventilator and mode can be used successfully. NPPV can be combined with inhaled bronchodilators and heliox.


1. Keenan SP, et al. Crit Care Med. 1997; 25:1685.

2. Peter JV,et al. Crit Care Med. 2002; 30:555.

3. Lightowler et al. BMJ. 2003; 326:185.

4. Keenan SP, et al. Ann Intern Med. 2003; 138:861.

5. Hess DR. Respir Care. 2004;49(7):810-29.

6. Carlucci A, et al. Am J Respir Crit Care Med. 2001;163:874.

7. Soroksky A, et al. Chest. 2003;123:1018.

8. Pang D, et al. Chest. 1998;114:1185.

9. Mehta S, et al. Crit Care Med. 1997;25:620.

10. Sharon A, et al. J Am Coll Cardiol. 2000;36:832.

11. Nava S, et al. Am J Respir Crit Care Med. 2003;168:1432.

12. Keenan SP, et al. JAMA. 2002; 287:3238.

13. Esteban, et al. N Engl J Med. 2004;350;24.

14. Kwok H, et al. Crit Care Med. 2003;31:468.

15. Antonelli M, et al. Anesthesiology. 2004;100:16.

16. Ferguson GT, et al. Am J Respir Crit Care Med. 1995;151:1126.

17. Schwartz AR, et al. Respir Care. 2004;49:270.

18. Hotchkiss JR, et al. Am J Respir Crit Care Med. 2001;163:374.

Dean R. Hess, PhD, RRT, Respiratory Care, Massachusetts General Hospital, Department of Anesthesiology, Harvard Medical School, is Associate Editor for Critical Care Alert.