Preoxygenation (PreOx) of patients is performed to extend the period of safe apnoea before the patient desaturates. This is a critical safety step before any procedure where apnoea is expected (eg intubation) or a possibility (eg procedural sedation).
PreOx is achieved through 3 main mechanisms:
– Preoxygenation of Blood
This is the maximisation of dissolved oxygen in blood. This is by far the least important PreOx mechanism as oxygen is poorly soluble in blood.
– Preoxygenation of Haemoglobin
Bringing the oxygen saturation as close as possible to 100%. In healthy lungs this can be achieved very quickly.
– Denitrogenation of the lungs
Washing out the nitrogen content of the lungs by replacing it with oxygen. This provides a large oxygen reservoir that will diffuse into the bloodstream in the event of apnoea. This is the most important aspect of PreOx, as the denitrogenated lungs represent 95% of the oxygen reservoir in apnoea compared to only 5% in blood. Denitrogenation can be achieved in 2 ways using a high FiO2 (Fraction of Inspired Oxygen) source (i.e. >90% FiO2).
1. Having the patient take 8 vital capacity (full) breaths in/out. This is obviously only achievable with an awake and compliant patient eg prior to procedural sedation
2. The patient takes normal tidal volume breaths for at least 3 minutes. This is usually the only option in Emergency Department (ED) intubations.
So it is not sufficient to just achieve good saturations. You must wait the required time for denitrogenation to occur. In the operating room, this time can be calculated using gas analysers of the expired oxygen content, however these are not generally available in Emergency Departments (ED). Therefore the Emergency practitioner should rely on the time estimate of at least 3 minutes as their marker of denitrogenation.
Devices that can provide a high FiO2 gas source for Preox in ED:
1. A Non-Rebreather face mask with reservoir bag (NRB)
2. A Bag-Valve-Mask (BVM)
3. Non-Invasive Ventilation (NIV)
Device settings required to create high Fi02
Positioning for Preoxygenation
Ideally to facilitiate preoxygenation the patient should be positioned in the semi-recumbent position with head elevation of approximately 30 degrees. They can then be repositioned to facility intubation after preoxygenation.
If a patient has C-spine precautions in place, the Reverse Trendelenberg position can be used for PreOx.
Problems achieving Pre-Ox? Here are the solutions!
Weingart article on Preoxygenation, Deoxygenation, Reoxygenation and Delayed Sequence Intubation
Weingart and Levitan article on Preoxygenation & Prevention of Desaturation in Emergency Airway Management
Non-rebreather masks will not reliably provide a sufficiently high FiO2 for adequate pre oxygenation. The FiO2 supplied by these devices is dependent on the design of the mask. Masks with only a non-rebreathing valve between the mask and the reservoir bag provide an FiO2 of only 70-80% (cf >95% with a BVM with an expiratory port valve). Those with additional valves on the mask may provide a higher FiO2 but can cause airway obstruction under certain conditions in very dyspnoeic patients. http://monashanaesthesia.org/fio2/
In general using a non-rebreather is not a good choice for PreO2. Apart from the limited FiO2 provided they remove the opportunity to confirm, prior to induction, that you are able to get a seal and an ETCO2 trace with the mask that will be used for ventilation if intubation is unsuccessful.
Nick, turning flow to >30 lpm will get you to ETO2 > 90% in the standard 2/3 valve (Model No 2012) covered reservoir face masks. All US flowmeters I have seen in the US deliver at least this much. This was tested but left out in your excellent post due to the variability of flow, correct.
We have always agreed that this NRB at 15 lpm is crap.
Lack of seal in a non-paralyzed patient in no way guarantees the absence of a seal in a paralyzed one. The converse is untrue as well.
I can’t make any comment on what could be achieved at 30L/min as the flowmeter we were using for the tests was subsequently found to be delivering 60L/min when the flows were “maxxed out”. This was not something we were able to reproduce on any other flowmeter in the hospital & the maximum O2 flow that could be reliably achieved across all flowmeters in our institution when “overcranked” was 20L/min. In not being calibrated beyond the markings, a flowmeter by definition cannot be relied upon to provide a particular flow rate when pushed beyond these. Given the high variability in what flowmeters do deliver when pushed beyond the calibrated markings, even within a single institution (let alone around the world), I think assuming that a particular O2 flow will be reliably achieved when the flowmeter is “overcranked” should be avoided – especially given there are other readily available devices/techniques which will dependably deliver an FiO2 >90% using flow rates for which the flowmeters have been calibrated.
In regards to mask seal I’m not sure I agree with you but I think we need to distinguish between two different situations which I’ll call primary & secondary leaks.
1. Primary Leak: this refers to the situation in which the mask is simply not firmly opposed to the patient’s face around its entire perimeter, allowing gas to be entrained during spontaneous ventilation and gas to escape during positive pressure ventilation. Barring a non-compliant patient I would say that inability to get a good seal with a mask in a spontaneously breathing patient would almost always guarantee the absence of a good seal once the patient is anaesthetised/paralysed and thus the persistence of a “primary leak”. It would depend on the cause of the poor seal but the common problems: poor technique, inappropriate mask size, beards, edentulous, etc are not going to improve once the patient is paralysed. Off the top of my head I can’t think of a circumstance in which the cause of a primary leak would be expected to resolve by anaesthetising/paralysing the patient. I don’t think it’s reasonable to assume that a significant leak that is present with negative pressure ventilation would necessarily disappear solely by virtue of converting to positive pressure ventilation. If there’s a gap somewhere between the perimeter of the mask and the patient’s face you will have a primary leak during preoxygenation and I would say that does absolutely guarantee you a primary leak during positive pressure mask ventilation once the patient is paralysed.
2. Secondary Leak: this refers to the situation in which the mask is firmly applied to the patient’s face around its entire perimeter (=> no primary leak) during spontaneous ventilation but when the pressure inside the mask is elevated for the purposes of positive pressure ventilation, the seal pressure around the mask perimeter is exceeded and gas escapes around the mask (=> secondary leak). If airway obstruction is present which cannot be overcome with pressures below the seal pressure of the mask then a secondary leak is inevitable. Thus in complete airway obstruction during mask ventilation a secondary leak must always occur regardless of the adequacy of the mask seal, as air is being forced under pressure into a blind passage so “something’s gotta give” – the gas can’t get into the lungs so it escapes around the mask. The primary problem in this situation though is not the mask seal, it’s the patency of the airway. In this regard if you had a secondary leak in a patient you were trying to deliver PPV to pre-induction/paralysis, it would be reasonable to expect that (since mask ventilation generally becomes easier and able to be performed with lower inspiratory pressures as pharyngeal tone decreases with onset of paralysis) the secondary leak might resolve once the patient is paralysed. Obviously though, this would not be a scenario in which you would be preoxygenating with a non-rebreather anyway. This is more an example of your final point, with which I absolutely agree – the absence of a (primary) leak during pre oxygenation of an awake patient most certainly does not guarantee the absence of a (secondary) leak developing in a paralysed one.
One of the difficulties with using a self inflating BVM for positive pressure ventilation during airway management is that, unlike the soft collapsible bag on an anaesthetic circle or Mapleson circuit, it does not allow you to distinguish between a primary and secondary leak and thus you can’t determine the best intervention to address the problem. When using a soft collapsible bag, if there is a leak around the mask and the bag collapses I know I have a primary leak and that I need to improve my seal around the mask. If however there is a leak around the mask with positive pressure ventilation but the bag remains full I know that there is a reasonable seal around the mask (no primary leak) but that this is being overcome by the excessive positive pressure needed to overcome impaired airway patency (secondary leak). Thus the appropriate intervention is to address the primary problem and try to implement manoeuvres to improve airway patency. When using a BVM the bag self inflates regardless of whether or not a primary leak is present. The poor “feel” for the amount of pressure required to produce the leak when using a BVM further impairs the ability to identify secondary leaks and target strategies to improve ventilation appropriately.
Thus I’d suggest that not only is a non-rebreather mask not a good device for pre-oxgyenation during airway management, but that neither is a BVM. Airway experts should be proficient with use of an O2 delivery device with a soft collapsible bag. Devices such as at the Mapleson circuit have numerous advantages including the ability to more precisely troubleshoot difficulties with mask ventilation.
In EDs and Crit Care units, the wall oxygen supply will always supply much greater than necessary pressure to get well beyond the 60 lpm mentioned as this same supply is used for ventilator hook-ups. So we can eliminate inadequate supply. Then we are left with the flowmeters. In the US, in the >100 programs I have visited, there is always the exact same model of flowmeter. Rich Levitan has told me the same for the programs he has visited. In the US at least, a maximally opened flowmeter with NRB is a perfectly adequate solution, especially when combined with a nasal cannula placed during the preoxygenation phase. It is not ideal, but it is by far the best option.
A BVM with a PEEP valve or exhalation valve will give higher fiO2, but if the operator lifts the mask for even one breath and the patient breathes room air, denitrogenation goes to crap. This is a waste of time for little yield.
As to demonstrating mask seal before intubation, this is a fundamental philosophical divide between crit care and anesthesia. We don’t care and it will not influence our airway management at all to know if we can get a mask seal on a conscious, unparalyzed patient.
While I like your division between primary and secondary leak, unfortunately it has not been supported by the anesthesia literature. Patients you define as primary leakers will often be baggable when paralyzed. These two articles have eliminated most of the argument that we must be able to bag a patient before administering a relaxant (at least for US anesthesiologists).
If I have the time in a stable patient, predicted difficult to bag may make me consider awake intubation, but I rarely have a pt stable enough for this.
SGA has changed this equation; a patient who may be difficult to bag won’t get bagged, they will get empiric LMA placement for any need to reoxygenate.
We don’t check if we can mask a pt, because we don’t care if we can mask a patient when awake. All we care about is can we mask them when relaxed, and if we can’t we will immediately move on to SGA.
Scott, I agree with you that this is not an issue of gas supply pressure. In Australia wall O2 supply is always 4kPa, clearly enough to deliver 60L/min since one of our flowmeters did exactly that. Whether or not a flowmeter is capable of delivering 60L/min depends on the minimum resistance of the flowmeter.
All the flowmeters I assessed were exactly the same model, yet the maximum flows varied between 20-60L/min. There must be manufacturing variations between the same model of flowmeters that lead to wide variations in the flow delivered when opened beyond the calibrated markings. If you and Richard have measured the maximum flow delivered by the flowmeters in your hospitals and found it to be consistent that’s a different matter but I would caution anyone else reading this against making that assumption without specifically checking the flowmeters being used in their clinical area. (Out of interest what is the model of flowmeter being used throughout the US? That model must have substantial world domination also. The vast majority of flowmeters I’ve encountered in Australia are the model shown in the oxygen delivery video. I’ll find out what it is – if its the same then there’s no reason the situation in the US should be any different from that in Australia). The simplest resolution would be to ask the manufacturers what is the maximum flow they would reliably say could be achieved with the flowmeters fully opened.
I agree also that using 15L/min supplementary nasal oxygen changes everything. In our demonstration both a BVM without an expiratory port valve and a ⅓ NRM (#2008) achieved ETO2 concentrations following preoxygenation which were consistent with an FiO2 >90%, with flow to the device at only 15L/min. This would seem to be the most reliable way to confirm an adequate FiO2 when using these devices. Whilst it does resolve the issue of % O2 delivery by a NRM, however, it doesn’t address the other issues.
The operator lifting the mask and ruining preoxygenation is just an example of poor technique. As I said above, in a particularly uncooperative or claustrophobic patient I can see there might be difficulties maintaining a seal with a face mask and in that situation the NRM is definitely a useful option. Beyond that the answer is DON’T lift the mask. I’ve intubated many sick, emergency patients in ED/ICU & theatre and I’ve rarely found avoiding this to be a significant challenge.
We may have to have a verbal discussion about mask seal as I suspect the issues are getting too complex to address efficiently in text – but I’ll give it a try!
I think you’ve introduced an additional issue in the articles you’ve cited (by the way I the Anesthesiology reference you cited linked to an article that seemed to be about difficult airway algorithms) that is unrelated to my point about primary & secondary leak. There two separate issues. The first is whether a “leak” in an “spontaneously ventilating” patient will correspond to a “leak” in an “apnoeic” patient. This is the issue of “primary leak” and the answer is “yes”. The second issue is whether “difficulty mask ventilating” an apnoeic “unparalysed” patient corresponds to “difficulty mask ventilating” an apnoeic “paralysed” patient. This is an issue relating to resistance to ventilation and may or may not involve occurrence of a secondary leak. I think it’s important we distinguish between “leak” & “difficult mask ventilation” and the transition from “spontaneous ventilation to apnoea” (& therefore positive pressure ventilation) and that from “unparalysed to paralysed” (which in the study cited both involve PPV). The issues with each are very different and I think have been bundled together until this point in the discussion.
I agree that the process of “test ventilation” prior to administering muscle relaxant is a waste of time and have always opposed it. It had been popular amongst some anaesthetists at one time but I think a number of articles published over the last 10 years debunking its rationale and effectiveness have largely eliminated the practice now. The reason it doesn’t work though is that, as I said in my previous post, when the pharyngeal muscles in an unconscious patient relax the degree of airway obstruction typically decreases. This lessens the amount of positive pressure needed to ventilate the patient and makes mask ventilation easier. It may also cause a secondary leak to resolve by maintaining ventilation pressure below the seal pressure around the mask perimeter. Nothing has changed about the seal between the mask and the face though (and conceivably – what could have?). The phenomenon of increased ease of ventilation with onset of muscle paralysis is entirely about decreased airway resistance, not about changes in the effectiveness of the face mask seal. Where a leak is involved it is about resolving a secondary leak not a primary leak.
The only factors relevant to developing a primary leak are face mask technique, the contour of the perimeter of the face mask and the contour of the patient’s face.
The factors determining whether a secondary leak occurs are the seal pressure obtained around the perimeter of the mask (which will relate to the factors determining existence of a primary leak) and the pressure generated in the airway.
Apnoea & muscle relaxation will not make any difference to the factors affecting primary leak but will may make substantial differences to the factors affecting secondary leak (ie. resistance to ventilation)
Thus the issue of “test ventilating” an unconscious, apnoeic patient prior to giving muscle relaxant is separate and unrelated to that surrounding whether inability to get an adequate face mask seal in a conscious, spontaneously breathing patient translates to inability to achieve a seal after either sedating or paralysing. The former issue relates to a change in resistance to positive pressure ventilation and thus a decreased likelihood of a secondary leak, as the ventilation pressure is decreased relative to the seal pressure (which is unchanged) with the onset of paralysis. In contrast a primary leak present prior to paralysis would still be expected to be present after paralysis as the face mask technique, shape of the mask and shape of the patient’s face will not be altered by muscle relaxation. Similarly a primary leak present when awake (that is not due to poor patient cooperation) will remain present post induction/apnoea as none of the factors contributing to it will have changed.
I don’t understand how you can say you’re not be interested to know that you can’t get a mask seal before inducing the patient? Surely if the reason you can’t get the mask to seal is that it is the wrong size or their beard you would rectify these things? In the same way as you position the patient’s head, ensure the correct dose of relaxant and most appropriate laryngoscope blade, confiming you have the right sized mask and the ability to get a face mask seal are part of preparing for airway management and ensuring as many factors as possible are optimised prior to induction.
In summary I’m not talking about confirming you can positive pressure ventilate an apnoeic (or even an awake, spontaneously breathing) patient prior to paralysing them, I’m talking about confirming you can get a mask seal in an awake patient prior to inducing them. These are 2 completely different things because if you have a primary leak awake you won’t be able to positive pressure ventilate asleep, paralysed or not.
Testing for a mask seal in an awake patient doesn’t require positive pressure ventilation. It can be inferred to some extent by the shape of the ETCO2 trace and by the FiO2 achieved (typically not available in ED though in my experience) – but it would still be possible to overlook a leak that could significantly interfere with positive pressure ventilation if using these techniques alone. The most sensitive measure to detect whether you have a primary leak in an awake, spontaneously breathing patient is to have an O2 delivery device with a soft, collapsible bag – as this will deflate if a leak is present. This is not possible with a self-inflating bag and is again why I would suggest that a BVM is not a good device to use for pre oxygenation.
A couple of corrections having just re-read my last post:
I’ve written 4kPa as the wall O2 pressure – that should have said 400kPa (~ 4 atmospheres).
Also whilst changes in upper airway resistance may occur, I imagine that the lower inspiratory pressure requirement (decreased “resistance to ventilation” as I’ve referred to it above) with the onset of paralysis, producing an increased ease of positive pressure ventilation, is predominantly due decreased respiratory/abdominal muscle tone and not decreased pharyngeal muscle tone as I seem to have suggested in my previous posts. It may be that decreased upper airway tone decreases results in some degree of pharyngeal dilation or more effective jaw thrust but I would think that the bulk of the effect relates to relaxation of muscles in the thorax & abdomen.
Ok, we have officially moved to the point where this would be more fun as a recorded skype/google hangout conversation that can get posted. I’ll email with some possible times in a couple of days.
Great idea and thanks for the excellent comments guys. This might be a good point for me to insert my 2 cents so you can hopefully cover this if appropriate in your recorded discussion +/- answer here.
1. It appears as discussed above that any discrepancies with NRM and BVM and their different designs seem to be removed by the addition of nasal prongs at 15L/min during the preox phase, at least from an FiO2 point of view. Given that apnoeic oxygenation should arguably be performed in all ED Urgent Endotracheal Intubations (UEI), then nasal prongs will be in place anyway, so commencing this during PreOx is no extra step.
Thanks heaps Nick for the link to your excellent demonstration
I had heard this FiO2 benefit of nasal prongs discussed but not seen it demonstrated till seeing this. Notwithstanding the limitations of the study, I do hope you plan to publish this as it would be a great contribution to the literature in this area.
2. One advantage of the NRM (cf BVM) in PreOx is that during an ED UEI it can be useful to have your 2 hands free to prepare everything else during your 3 minutes of PreOx (assuming your hands are not required to maintain a patent airway). This is not a small advantage.
3. I have concerns about the real FiO2 that is being delivered by BVM’s in patients who are hypoventilating with shallow breaths (eg resp fatigued, tox OD’s). It takes a fair bit of effort to generate the negative inspiratory pressure required to open the inspiratory valves of BVM’s, in particular the common duck-bill designs. Such patients will barely open the inspiratory valve so will get most of their air from leaks around the mask with minimal spurts of high FiO2 from a barely opening inspiratory valve. It may also explain their agitation and attempts to remove the mask from their face as they feel they are being suffocated. I’ve tried restful tidal volume breathing on such BVM’s and had substantial air hunger when I had a good seal.
Most assessments of BVM FIO2’s including the Monash one by Nick are on healthy patients so don’t answer this question.
It is worth noting that some BVM’s have a better designed inspiratory valve that mitigate this eg I recently tested out one of the latest Ambu bags which have a shutter valve instead of the duck-bill that requires much less inspiratory negative pressure to open.
4. I take Nick’s point re the leak detecting benefits of using an BVM instead of the NRM. However what is unclear is exactly how clinically useful this actually is? ie how often does one detect a leak and make modifications based on the CO2 trace, (without confusing this with an abnormal CO2 trace created by other pathology) AND by making these modifications how much benefit dId the patient experience once sedated in terms of prevention of desaturation. I presume this has not been studied so the answer would be opinion based.
5. I agree the Mapelson C would be superior to both BVM’s and NRM yet they are generally seem unavailable in ED’s in Australia though this could and probably should be changed. Nick, a colleague mentioned at his institution there were some infection control issues with Mapleson C a long time ago so they got rid of them. I’m unsure if he was talking about disposable or reusable ones but just thought I’d bring this up in case you had heard anything about this.
Scott – sounds good.
Anand – Scott and I will try to discuss some of these issues in more detail, but in short…
1. True – nasal O2 may compensate for some poorly performing devices (conversely a high FiO2 device does not compensate for absence of nasal O2).
2. I suspect more is lost than is gained.
3. Yes, our subject was surprised how much effort was required to breath through the BVM. Another advantage of the valveless Mapleson circuit.
4. Detecting a leak which could potentially impair or prevent effective mask ventilation before induction is common and VERY clinically useful – and the benefits extend far beyond just that individual patient (a point Scott & I should definitely expand on).
5. Never heard this. We have reusable Maplesons in the recovery room of our OR’s and have them re-sterilised without issues.
great look forward to the audio recording!
following the comments above I’ve edited and created a link above “Device settings required to create high FiO2” to clarify the post.
Hey Scott/Nick – have you guys had a chance to catch up for an audio debate regarding your points of difference above?
Pingback: FOAM Eye-Catchers 6 | Emergucate