Sage Advice: Indications for Non-invasive Ventilation

Written by Sage Whitmore, MD

whitemore blog picMy favorite thing to do on Thursday nights at 2:30am is pontificate on non-invasive ventilation, so…whatever that says about me…

It came up on a shift the other day that there are some patients with respiratory failure who are best served by a trial of BiPAP, and some patients who are best served by an early intubation. I wanted to review some of the literature and make some (generally agreed-upon in the ICU community at large) recommendations on this controversial topic. The major issue at play here is whether the patient in front of you is likely to be helped by BiPAP (avoid intubation, decrease LOS, decrease nosocomial infection, get better quicker, die less), or ultimately harmed (delayed intubation, more complications, worsened condition, die more).

There are generally 5 categories of patients for whom BiPAP is preferred over intubation:

COPD STRONG evidence.

–       Reduced need for intubation

–       Reduced mortality

–       Reduced nosocomial infections

–       Decreased LOS

CHF STRONG evidence.

–       Reduced need for intubation

–       Decreased LOS

Severely Immunocompromised MEDIUM evidence in patient with Neutropenic Fever, recent Solid Organ Transplant, Active Chemotherapy

–       Reduced intubation rate

–       Reduced mortality (if they avoid intubation)

–       Reduced nosocomial infections

Cystic Fibrosis/Advanced Interstitial Lung Disease (who are NOT transplant candidates) MEDIUM evidence

–       Less nosocomial infections

(ALSO, these are terminal condition at the time that the patient would need intubation, and they are VERY unlikely to come off the vent without a trach. Would rather have a palliative conversation than intubate these guys)

Asthma WEAK evidence

(But we know it works, and asthma is very difficult to manage with a high mortality on a ventilator )

Obesity-Hypoventilation Syndrome (often misdiagnosed as OSA +/- COPD) responds nicely to BiPAP as well, IF they are awake enough to participate and can be positioned sitting straight up to take the weight off their diaphragm.

There are generally 5 categories of CONTRAINDICATIONS to non-invasive ventilation

Decreased LOC Your patient should be a GCS of 14 or better, should be able to participate. Should be able to take a deep breath when you instruct them, should be able to give you a thumbs up, should be able to press their call light if they feel worse
Airway Issues Facial trauma, upper airway mass, obstruction, stridor, vomiting, hematemesis, poor swallow, poor cough, any recent Thoracic, Esophageal/Upper GI, ENT, or craniofacial surgery, or spine precautions = don’t do it!
Respiratory Conditions Respiratory arrest (duh), hypoventilation, initial presentation of myasthenia gravis, guillan barre, cervical cord injury, de novo pneumonia/ARDS in the absence of COPD/CHF*
Circulatory Issues Hemodynamic instability (diaphragm steals up to 50% of cardiac output during septic shock), active ischemia (new ECG changes, trop rising), malignant arrhythmias
Prognosis Cause of respiratory failure unlikely to resolve within 24-36 hours

*On the issue of BiPAP for Pneumonia/ARDS: There is no randomized trial comparing NIV to invasive mechanical ventilation in PNA/ARDS that I know of. There are old trials showing that compared to plain supplemental oxygen, NIV reduces the need for intubation (duh). The most important is the data is looking at how often patients with PNA/ARDS fail a trial of NIV, and what happens to them after they fail. Here is what I look at, based on studies that are referenced at the bottom:

  1. Predictors of NIV failure include: presence of pneumonia, hemodynamic instability
  2. NIV failure rates are >50% in presence of severe pneumonia and/or ARDS
  3. Overall mortality rates for patients who fail NIV are anywhere from 50-92% (highest being in neutropenic patients with HCAP/ARDS)
  4. Mortality rates for patients who fail NIV are 2x higher than otherwise predicted by severity of illness scores
  5. Patients with de novo hypoxic respiratory failure (i.e. PNA/ARDS) are >3x more likely to die if they fail NIV
  6. in a review from our QI database at U of A over the last two years, patients who were intubated after failing NIV are >2x more likely to significantly desaturate, become hypotensive, or aspirate during intubation, also associated with increased odds of death

The take home point is: patients with PNA/ARDS are more likely than not to fail NIV, and when they do, their intubation is more complicated and they are more likely to die. So, prevent an ugly crash intubation in the middle of the night; intubate your septic pneumonia patient early.


  1.       Kramer et al AJRCC 1995 – reduced intubation in COPD group
  2.       Keenan et al. Crit Care Med 1997 – reduced mortality and intubation in COPD
  3.       Antonelli et al. NEJM 1998 – non-COPD patients
  4.       Pan et al. Chest 1998 – metaanalysis, reduced intubation
  5.       Wood et al. Chest 1998 – small ED trial, included PNA patient
  6.       Confalonieiri et al. AJRCCM 1999 – benefits for severe pneumonia
  7.       Antonelli et al. J Crit Care 2000 – Review article, helps in CHF, COPD
  8.       Wysocki et al. Chest 1995 – non-COPD pts, no benefits of NIV
  9.       Antonelli et al. JAMA 2000 – NIV in fresh transplant is good
  10.    Cheung et al. Hong-Kong Medical Journal 2000 – failed NIV worsened mortality, esp PNA
  11.    Girou et al. JAMA 2000 – NIV decreased nosocomial infections in COPD and CHF
  12.    Martin et al. AJRCCM 2000 – need for intubation decreased in all comers
  13.    Moretti et al. Thorax 2000 – prediciting NIV failure in COPD
  14.    Sinuff et al. CMAJ 2000 – no clear predictors of NIV failure in mixed population
  15.    Ambrosino et al. Thorax 1995 – precense of PNA and low pH predict NIV failure
  16.    Azoulay, Alberti et al. Crit Care Med 2001 – NIV decreased mortality in Cancer Pts
  17.    Gilles, Didier et al. NEJM 2001 – NIV in immunosuppressed
  18.    Hilbert, Gruson, et al. Crit Care Med 2000 – NIV in neutropenic, all responders lived
  19.    Madden, Kariyawasam, et al. Europ Resp J 2002 – NIV in cystic fibrosis improves PaO2
  20.    Carrillo, Gonzales-Diaz. Inten Care Med 2012 – NIV for CAP, delayed ETI = mortality
  21.    Ferrer, Esquinas. AMJRCC 2003 – RCT of NIV v O2 for hypoxic resp failure
  22.    Gristina et al. Crit Care Med 2011 – NIV for heme malignancy, 5yr experience.
  23.    Jolliet, Abajo, et al. Intens Care Med 2001 – cohort of NIV for ARDS, 66% failed
  24.    Rana, Jenad, et al. Crit Care 2006 – risk factor for NIV failure in ALI
  25.    Carron, Freo, et al. J Crit Care 2010 – severe CAP on NIV, 56% failed
  26.    Demoule, Girou, et al. Intens Care Med 2006 – NIV for de novo vs chronic resp failure
  27.    Ferrer, Cosentini, et al. Europ J Intern Med 2012 – Review of NIV for PNA
  28.    Sorksky et al. Europ Resp Review 2010 – NIV for asthma, review
  29.    Zhan, Sun et al. J Crit Care Med 2012 – RCT of NIV vs simple O2 for Acute Lung Injury
  30.    Antonelli, Conti, Moro et al. Intens Care Med 2001 – predictors of NIV failure
  31.    Keenan et al. Crit Care Med 2004 – systematic review of NIV for hypoxic resp failure
  32.    Delclaux et al. JAMA 2000 – RCT of NIV vs simple O2 for Acute Lung Injury +/- CHF
  33.    Carron, Freo et al. British J Anaesth 2013 – Review of NIV complications
  34.    Emeriaud G, Crulli B, Ducharme-Crevier L, Nishisaki A. 429: Safety of pediatric tracheal intubation after non-invasive ventilation failure (POSTER). Critical Care Medicine 2012;40(12) Supplement 1