An ounce of PreVent-ion

No Comments

To bag or not to bag – does it affect RSI outcomes?

It is a busy day in the Emergency Department, and you’re faced with a critically ill patient.  He’s 64, short of breath, coughing, febrile and has become more hypoxic over the course of the last hour.  With a non-rebreather mask on, his oxygen saturations read 92% and his respiratory rate is 32 breaths per minute.  The chest x-ray shows multifocal pneumonia.  He’s full code, and you know it is time to intubate him.


There has been a barrage of useful information you’ve gleaned through talking with your colleagues, your once monthly attendance at the area residency program’s conferences and on social media.  You position the patient appropriately, add supplemental nasal cannula oxygen, dose the medications in accordance to his hemodynamics and tell the respiratory therapist what vent settings you would like based on his ideal body weight. Your team is ready to proceed with the intubation.

But out of the corner of your eye, you notice the blue saturation reading still lingering ominously at 94%.  You know the medications you’ve chosen will render him apneic for at least a minute, and likely longer before you get the tube in place and restart ventilations.  This patient is already at the edge of the cliff, and likely will not tolerate any further physiologic insults.

“Should I just bag this patient while the drugs kick in?” you think to yourself, but this only creates further turmoil in your decision-making as it goes against everything you were always taught.


Or does it?

The dogma

Rapid sequence intubation (RSI) has been taught as a means to achieve intubation in a critically ill patient safely.  The premise is the patient is not likely to be “prepared” for this intubation and may have a stomach full of that day’s meal.  To prevent aspiration, traditional teaching has been to avoid bag-valve-mask (BVM) ventilation during the apneic period.[1] The theory is that bagging will force open the lower esophageal sphincter, and facilitate aspiration of gastric contents.  There are certainly some clinicians who are strictly in the “never bag” camp.

On the flip side, some argue that occasionally, in the hypoxic patient, the risk of further hypoxia is too high during the apneic period, and you should provide supplemental ventilation.[2]  The risk of cardiac arrest and death can certainly increase – I can treat aspiration…treating death is a little harder!  Others argue that taking away the respiratory compensation of a severely acidotic patient, even for a brief period, may cause significant harm.

The problem is the available data to support one argument fully over another has been lacking.


A breath of fresh air?

A recent publication in the New England Journal of Medicine may provide some clarity to this question.[3]  Entitled “Bag-Mask Ventilation During Tracheal Intubation of Critically Ill Adults,” this paper from the PreVent (Preventing Hypoxemia with Manual Ventilation during Endotracheal Intubation) investigators ask the simple question: Will BVM ventilation during the apneic period significantly increase the lowest oxygen saturation from induction to two minutes after intubation when compared to no ventilations?  Their additional questions included identifying the incidence of aspiration.

What are the details?

This was a randomized controlled trial conducted in ICU patients at seven academic medical centers across the United States.  They included all adult patients (over age 18) requiring endotracheal intubation.  Their exclusions were listed as pregnant patients, prisoners and those patients whose clinical state demanded intubation before randomization could occur, or those in whom the treating physician determined they would definitely require BVM ventilation or in whom they felt it was contraindicated (e.g. hematemesis, hemoptysis, actively vomiting).

Block randomization occurred in a 1:1 ratio.  However, given the nature of the intervention, the proceduralists and research personnel could not themselves be blinded to the randomization result.

Patients in the intervention arm underwent BVM ventilation during the interval from induction to laryngoscopy.  They were very specific about how this should happen: two-handed mask seal; oxygen flow rate set to 15 liters per minute; PEEP valve attached set to 5 to 10 centimeters of water; oropharyngeal airway in place; appropriate head and airway positioning; and breaths applied at enough volume to allow visible chest rise.  Breaths were administered at 10 per minute.

The control arm was not allowed to receive BVM ventilation at all – except if there was a failed intubation attempt, the patient was severely hypoxemic (saturations <90%), or the clinician felt that this was needed for patient safety reasons.

Non-invasive ventilation was not allowed, but otherwise treating clinicians had free rein over all other aspects of the intubation procedure (including the use or not of apneic oxygenation).[4]

Show me the numbers

In total, 401 patients were randomized, 199 to the intervention arm and 202 to the control arm. The authors did a good job at matching the patient groups – they were primarily older (median age 60 years), with almost half having sepsis or septic shock and almost 60% with respiratory failure as the reason for intubation.  The first-pass success rate was 82%.

In terms of the primary outcome, the lowest median saturation in the control arm (no ventilations) was 93% versus the BVM arm which was 96%.  The control arm had more episodes of desaturation (defined as oxygen saturations <90%) than the BVM group – 40.1% versus 29.5%. Both these results reached statistical significance.

A bigger truth lies deeper in the numbers.  Extreme desaturations (<80%) occurred much less frequently in the BVM group – shown nicely in this bar graph.  This is a finding that carries significant importance and clinical relevance.

But what about aspiration? The trial showed no significant difference in terms of operator-reported aspiration – 4% in the control group and 2.5% in the BVM group. Aspiration was determined by the presence of a new chest x-ray opacity within 48 hours of intubation.

No trial is without limitations though. There were some variables left up to the clinician. Apneic oxygenation, for example, was utilized in all the BVM group patients, and only two-thirds of the control group ones. BVM ventilation was applied to some of the control group patients. The BVM ventilation technique itself was very scripted and would need to be replicated in its entirety to account for similar results in other patients. Finally, this was a study done on the ICU, where circumstances may be different from what we encounter in the Emergency Department.

The final word

Despite its limitations, this remains a well-conducted, multicenter randomized controlled trial. We see similarly critically ill patients in the ED, suffering from critical hypoxia who are at the edge of the cliff.  These patients are at high risk of aspirating, but also high risk of cardiac arrest from hypoxia. If nothing else, this trial provides us with a call to arms to improve our basic airway skills in using the BVM to safely maintain oxygenation during our intubation attempt, and provides a measure of reassurance that the aspiration risk with gentle, appropriately applied ventilation is not significantly different to providing no assisted breaths.


  1. Higg A, McGrath BA, Goddard C, et al. Guidelines for the management of tracheal intubation in critically ill adults. Br J Anaesth 2018;120:323-353
  2. Brown JP, Werrett G. Bag-mask ventilation in rapid sequence induction. Anaesthesia 2009;64:784-785
  3. Casey JD, Janz DR, Russell DW, et al. Bag-mask ventilation during tracheal intubation of critically ill adults. N Engl J Med 2019;380:811-821
  4. Binks MJ, Holyoak RS, Melhuish TM, et al. Apneic oxygenation during intubation in the emergency department and during retrieval: A systematic review and meta-analysis. Am J Emerg Med 2017;35:1542-1546


Zaffer Qasim, MD is an emergency and critical care physician in Philadelphia. His clinical interests include trauma resuscitation and airway management. Follow him on Twitter at @ResusOne.

Leave A Reply