Matt Semler, MD and colleagues performed an RCT on Apneic Oxygenation in Medical ICU patients. I got a chance to sit down and interview him on the trial.
The Trial
The Fellow Trial (Published ahead of Print)
Bottom Line Review
Trial Summary
Additional Written Comments from Matt Semler
as provided by Matt Anderson (@ccinquisivist)
From: Anderson, Matthew J
Dr Semler
I'm a CCM fellow at the University of Wisconsin, interested in airway mgmt in the ICU.Just saw your article published in ATS. I had a question about the airway mgmt protocol (or if there was one?). Did the airway team leader/intubator maintain a patent airway (ie were they instructed to do this just prior to the intubation attempt when getting sedation/analgesia and/or NMB)? I am unable to find/get to the supplement which this information may be listed but I didn't see any mention in the main manuscript, which I think is a extremely important discussion point. Previous, studies in the OR w/ or w/o maintenance of airway patency resulted in ‘no difference in the non-airway patency (ie jaw thrust/head tilt chin lift) group vs ‘stat significant difference in the airway patency group' during apneic oxygenation. If airway patency was maintained in your study this would be one of the first ‘negative' results I have seen with apneic oxygenation. If airway patency was not required, this may explain the ‘no difference' that was found in your study, which in my opinion, makes the use of apOX still an important part of endotracheal intubation. Until a randomized control trial to evaluate apOx with airway patency versus no apOx with airway patency confirms that previous. Further trials may need 30 degree ramp/optimal positioning, as well?
Thanks for taking the time to answer my questions and publish/perform important ICU airway research.
Matt Anderson
Critical Care Medicine Fellow, PGY5
________________________________________ From: Semler, Matthew
Matt,
Thanks for your interest in the trial. You ask two really important questions — actually two of the same points Rich Levitan emphasized when he visited during the conduct of the trial.
(1) When discussing the effect of airway patency on outcomes of apneic oxygenation, the time-period in question is between administration of RSI medications (with anything prior to induction technically a part of pre-oxygenation) and the onset of laryngoscopy (when patency of the airway is directly established by the laryngoscope better by external maneuvers). Objectively assessing whether the airway is patent during this period is challenging. For the 30% or so in the trial who were on BIPAP between induction and laryngoscopy, the airway was known to be patent through monitoring of the returned tidal volumes. In cases where NIV was not present, the operator was charged with maintaining patency of the airway between induction and laryngoscopy. In 60% of cases this required an oral airway and a head-tilt-chin-lift maneuver. In around 40 patients, the operator felt the airway was patent without such a maneuver. Whether these maneuvers were effective in maintaining patency or whether patency was truly present in those patients who were not felt to require a maneuver is difficult to know. We did analyze the subgroup of those who were on BIPAP and we were certain the airway was patent and there was not a significant effect of apneic oxygenation on lowest oxygen saturation in this group — though obviously this is a not a large population.
An important thing to consider when thinking about the period between induction and laryngoscopy is that high flows of oxygen are being delivered in BOTH arms. The AO group was getting 15LNC on top of a non-rebreather or BIPAP or bag-valve-mask but its not like in the usual care group their preoxygenation device (BIPAP, NRB, etc) was REMOVED at induction. So, even if airway patency were maintained PERFECTLY in both groups, usual care and AO both had high flows of oxygen available during this period and it was not until laryngoscopy started that the difference in available oxygen became more extreme (usual care group had NO oxygen and AO had 15L/min NC).
All that said, a big potential difference between our study and prior studies was patient population. I don't know of any randomized trial of AO previously in any population except those with healthy lungs undergoing elective intubation. The one trial I do think is really relevant is the PREOXYFLOW trial (attached) which functionally compared apneic oxygenation with 60L/min NC to no oxygen during laryngoscopy. They included a similar population of sick ICU patients with a lot of pulmonary dysfunction and saw pretty similar results to ours. I think the idea that oxygenation generally doesn't work as well when you have profoundly damaged lungs make me more inclined to want to see a study of AO in acute intubations for trauma or stroke or some setting where lung function is preserved to see if the results would be different there.(2) I could not agree more that patient positioning is a potentially important subject for emergent airway management research. Historically, the recommendation has been for “sniffing” position based on intubations in the operating room but I agree with you that there are some interesting preliminary studies of ramping between 25-35 degrees (with regard to both grade of view and duration of apnea without desaturation). To help untangle this question, we are currently performing a randomized trial of ramped versus sniffing position for ICU intubations (clinicaltrials.gov # NCT02497729).
Thanks for the great questions. There are not a lot of ICU folks interested in airway management research so if you all are working on projects that might be relevant to what we are trying to do or where there might be room for collaboration, please don't hesitate to contact me.
Thanks!
Matt
What do you think? Comment Below!
Now on to the Podcast…
Podcast: Play in new window | Download (11.3MB) | Embed
Subscribe: Apple Podcasts | Android | Google Podcasts | RSS | More
Scott Weingart
thanks mate for a great podcast so soon after article publication. Kudos to the investigators for questioning the role of ApoX in emergent RSI in their ICU setting.
I think you have hit the nail on the head in that BVM was used in about half of each trial arm, in other words totallly diluting the effect of Apox. We know ApOx works as the physiology has been well described for decades.
I think the FELLOW trial demonstrates that you can do other things that are just a good as ApOx adjunct in your RSI practice i.e good preox, adjunctive BVM during ETI attempt. I do note 2 episodes of aspiration in the control arm versus none in ApOx group but this was not statistically significant.
Great discussion, my hat is off to Matt and colleagues for completing such an interesting study. I agree with your comments Scott. If a patient is hypoxic at the beginning, then I think CPAP/BiPAP should be used for preox, and continued post-induction (or BVM used) – exactly as ‘Weingart and Levitan’ originally described in AnnEmergMed. I think ApOx with nasal prongs alone is only an option in patients who are not hypoxic from the start and not using CPAP/BIPAP/BVM. It wasn’t clear to me from reading the study that BVM was performed, or if the BVM was simply being used as an oxygen mask. This discussion has cleared this up nicely – they were being ventilated post-induction – not truly apnoeic. these patients, from first principles, are unlikely to benefit from ApOx with nasal prongs IMO. I will still use ApOx with nasal prongs in case I have a patient who has an unexpected difficult airway leading to prolonged laryngoscopy, and to facilitate apnoeic CPAP in patients I don’t want to ventilate post-induction (e.g. bowel obstruction). This study is a great reminder that ApOx with nasal prongs is just a useful adjunct – we must get the basics right to… Read more »
Great comments, thumper. WIll also have EMNerd on to discuss why the power stats don’t add up at all (or for the NIPPV trial that this was based on–just so happened that was positive, so it was not an issue)
Did I hear that right in the PC? Did he say it took roughly 2-3 minutes to pass the tube in the majority of the PTs from the time of pushing the induction meds?
No it was 1-2 minutes from pushing of meds to introduction of laryngoscope
Also at my FD ApOx is a must for all RSI PTs. The NC goes on immediately when we see we have a critical PT, a respiratory distress PT or a PT who may be a candidate for airway management and it doesn’t come off until after intubation. We don’t consider ApOx an extra step since it literally take 5 seconds to place the cannula on the PT and does not get in the way. We also have no compromise in mask seal with PTs on a BVM or CPAP ever since we use foam masks on all our PTs. Secondly in my simple thinking, if our PTs are truly hypoxemic how can we justify anymore withdrawing O2 for at least a minute or more and provide no O2 or PPV in PTs we know have the likely potential to desaturate quicker than we can intubate under even ideal conditions(appropriate PreOx and Postionsing of PT)? I don’t buy it… In our mind it is not an extra step, it does not get it the way, and if a patent airway is provided IT WILL deliver oxygen continuously to the high majority of our PTs during the apneic phase of RSI.… Read more »
[…] an interview conducted by Scott Weingart, from EMcrit, (the author of the original apnoeic oxygenation paper), with the lead author of this paper, some […]
I think this article will help clarify a few points on pre-oxygenation techniques. It may be, when airway management is performed using some of the techniques for pre-oxygenation that they used it may substantially negate the benefits of using a nasal cannula.
I don’t think most emergency medicine providers utilize B-PAP for pre-oxygenation and even though I believe that BVM ventilation is common prior to intubation it is my belief that this practice should be avoided in our setting.
I have unpublished data from prehospital RSI’s (a data set of about 260 patients) showing an 12.5% reduction in SpO2 dropping below 90% on first intubation attempts and a 25.8% reduction on the second attempt after the implementation of ApOx. I believe this is similar to what most EM physicians are doing and in my mind definitively proves ApOx works. I guess its time to write this article.
I’ll continue using apneic oxygenation in the patients I don’t want to ventilate with a bag valve mask (bowel obstruction etc). As an anesthesiologist I’m confident in hand ventilating most patients. If they desaturate I just use the BVM.
But, in the patients I absolutely don’t want to ventilate, buying some time with apneic oxygenation is golden.
[…] During Attempts at Larngoscopy) has been published online and Emcrit has just released a podcast interviewing the leader author. There is also a brief review of this paper at […]
Scott,
FELLOW Trial summary and review by The Bottom Line coming very soon.
Blame Paul Young @DogICUma for the delay. 2 major trials in one week?!
Fantastic podcast. Thanks for doing all the hard work for us
BWs
Steve
[…] out this podcast with one of the FELLOW trial authors at […]
Once again this trial shows that difficulties of using data obtained in one setting…say the ED or the OR or ICU…and trying to show effectiveness of and intervention or lack of.
I applaud anybody who does research since nowadays it is hard to even get the most simple study approved by IRB process.
I was happy to get to hear the dialogue you had Scott.
I am of the mindset, similar to some above, that it is just such a simple step to add and I cant think of anyway it can cause harm…unless of course you get your blade stuck and tangled in the tubing…kidding
does it add to to better outcomes….it is really hard to know since you cant sit around and wait for PEOPLE to start desaturating.
The way you really test it is to not intubate them and see who desats first…but wait that would rely on both patients having very similar physiology for the reason for intubation.
[…] Scott Weingart at EMCrit: Podcast 158 – The FELLOW Trial on Apneic Oxygenation in ICU Patients […]
Forget the studies. Anecdotally the nasal cannula in my experience definitely buys you precious time for laryngoscopy during a true RSI – and I find this time is particularly useful when training residents. Once patients start to desat, the situations always get more dicey. I don’t like desats during emergency intubations in my trauma bay or ICU, period.
Even if you don’t believe it works for the apneic period – everyone believes in pre-induction denitrogentation/oxygenation. The cannula increases the efficiency and time to do this….. So taking it off would be an extra step which just isn’t necessary. Leave it on and continue….
Pasting my comment from emlitofnote: My question about the PPV isn’t that it’s not usual care; rather, it’s a different, very sick population that required that much PPV (only 77% in control arm, 70% in intervention arm). So not only is this a small study to begin with, but only 40 patients weren’t getting active ventilation in during the (non)apneic period. One takeaway I’ve had from using nasal ox during laryngoscopy is that it really seems to separate out the truly hypoxic patients, versus the healthy ones that I just let the intern do DL a little too long on…. Where does that leave us? I dunno. It’s the sick patients who I’d love to be able to oxygenate the most, and I think this trial tells us it doesn’t seem to help in them. While I agree that it’s unlikely to be dangerous and it’s just a few simple steps, I also agree with Matt Semler’s argument on EMCrit re: the hidden harm of adding one more step to the stressful, busy time around laryngoscopy, which is exactly my argument against cricoid pressure. I think there are probably sick patients who don’t need PPV who still may benefit so… Read more »
Scott, Awesome job getting Matt on so shortly after news of this trial go out. My thoughts: I mentally break this down (similarly to Matt) into 3 PHASES: 1) Pre-Oxygenation prior to apnea 2) Apnea to Laryngoscopy 3) Laryngoscopy to intubation ApOx only comes into play during PHASE 2 & PHASE 3 If we implement BVM/NIPPV during PHASE 2, we chop out a huge chunk of potential benefit from ApOx, and this essentially becomes a trial of: PHASE 3 ApOx vs PHASE 3 No ApOx {Patients with significant shunt physiology aside}.. Even in patients W/OUT significant shunt physiology, if we were to hypothetically use BVM/NIPPV during PHASE 2, we probably would wash out and not see much added benefit from using ApOx during PHASE 3. (Keeping in mind PHASE 2 is a relatively long period of time to be apneic!) Therefore, it’s no wonder that in patients WITH significant shunt physiology (in whom we already know ApOx will have far less effect anyway), we don’t see much added benefit when, comparing(again): PHASE 3 ApOx vs PHASE 3 No ApOx; which is what I think this trial (all its limitations aside) seems to suggest. I think a key question is: if… Read more »
Sam, fantastic comments. Agree with all of that
Great interview on an insightful study. It seems to me the evolving discussion here is about the relative safety of bagging or Bipaping sick, hypoxic pts after RSI meds are given vs. avoiding PPV due to aspiration risk
For me this underlines (with exclamation mark) that any provider involved in invasive airway management, must have a very clear understanding of respiratory physiology and pathophysiology. (Urgh, getting chills when thinking back to my time spent with West. Well worth it in the end).
As to not have a false sense of security when using any adjunctive method to guard against hypoxemia. In this case ApoX alone is not going to help much.
Very insightful research and discussion.
Dwayne
Specialist Anaesthesiologist
South Africa
[…] Apneic Oxygenation? (REBEL EM), The last breath for apnoeic oxygenation? (St Emlyn’s: JC), Podcast 158 – The FELLOW Trial on Apneic Oxygenation in ICU Patients […]
[…] Podcast 158 – The FELLOW Trial on Apneic Oxygenation in ICU Patients from EMCRIT.org […]
[…] http://emcrit.org/podcasts/fellow-trial/ […]
Good analysis and good points.
The first thing this study proves is that NODESAT is safe, that is, it is not worse than the other therapies. Obviously, this study proves that NODESAT is not substantially better than the other approaches, but….and here is the catch….I don’t think the study’s hypothesis can be proved with a simple pulse oximeter. So what kind of monitoring can prove the hypothesis? That is the important question. Here is a potential list:
1. Sidestream gas analysis monitor (the thing anesthesiologists use in the operating room for capnographry)
(1) Analyze the pre-induction End-Tidal O2 with the post-intubation End Tidal O2. I’m doing that in a current quality study (which does not concern NODESAT, but perhaps next year I will do that).
2. Intraarterial PaO2 monitor
(2) Not available widely
Using the pulse oximeter to answer this question is a faulty method. The pulse oximeter is too flimsy a monitor to answer such an important question.
Anyway, great job on this podcast
[…] Scott Weingart on EMCrit: Podcast 158 – The FELLOW Trial on Apneic Oxygenation in ICU Patients […]