Lots of excellent literature out recently. We would heartily recommend a look at all the parallel releases from the SCCC 2023, including the CLOVERs trial and others (1). There has also been some great work on thromboprophylaxis in trauma which we will try and get to shortly (2).
But first on the list, was this randomised controlled trial comparing conventional rapid sequence induction in patients with aspiration risk using neuromuscular blockade (NMB), to deep analgosedation with remifentanil (3). What’s this all about then? RSI has undergone enough changes recently thankyou very much. Most centres are still barely catching up. And now someone wants to drop neuromuscular blockade? Sounds bonkers…..
Or does it? Many remifentanil proponents have long argued for its rapid onset efficacy in achieving adequate muscle relaxation to facilitate most things. The rapid offset also solves all those awkward problems of what to do if the view is not as expected, or you run into problems, or you just don’t like the cost of an emergency dose of sugammadex….
Very interesting. And the authors should be congratulated for studying such a tricky question in critically ill patients, rather than simply enacting a test of change. Like all papers we would encourage you to read the detail within the manuscript and formulate your own thoughts. But here are some of ours:
This was a trial you said?
Yes – a multicentre, randomised, open label, non inferiority trial delivered across 15 sites in France and Europe. Open label is probably reasonable here, given the risks of the procedure and the now established but expensive antidote for the control (NMB) group. We need to remember the subjectivity this introduces however, especially around outcomes such as ‘best laryngoscopic view’. Multicentre is helpful and would hopefully ensure adequate balance of perioperative risks between groups, whilst also maintaining allocation concealment for investigator teams. Non inferiority probably warrants a bit of discussion however. The principal rationale for running this study as non-inferiority can only be that NMB is known to be harmful; therefore, if we could find a way to safely perform RSI in patients without using NMB, then we wouldn’t need it to be any better, just no worse than our current method? I have questions here on costs, familiarity, awareness, and longer term outcomes. I personally think there are multiple advantages to NMB during the period of intubation/stabilisation in the critically ill patient that would not necessarily be achieved with remifentanil. So I think this should have been a superiority trial. But that’s just me. Maybe you have different thoughts.
Who were the patients?
In 18 months, the authors recruited 1150 adult patients undergoing tracheal intubation in the operating room who were deemed to be at risk of aspiration (thus requiring RSI). Impressive stuff. Age, gender, comorbidities and ASA grading were evenly matched between groups. In fact, pretty much all baseline characteristics were well matched. Overall, the patients were perhaps a bit younger and a bit fitter than I was expecting (mean age 50, 80% ASA 1 or 2), but table 1 would suggest that randomisation did its job.
What did they do?
On inclusion, the patients were randomised to receive either NMB (control) using suxamethonium or rocuronium at conventional RSI dose (1mg/kg), or remifentanil at a dose of 3-4mcg/kg alongside a hyponotic of their choice. Tracheal intubation was initiated 30-60 seconds following induction in both groups. Rescue therapy (with either remi or NMB) was allowed in the event of desaturation or unplanned difficult tracheal intubation.
All RSI procedures were led by anaesthetic teams and efforts were made to ‘standardise’ RSI procedure, in order to ensure fidelity to the intervention and limit other between group differences. However, some of the standardisation was not, I would suggest, universal practice. The authors argue that it is not common practice in France to do any of the following: use of a stylet or bougie (boo), cricoid pressure (hurrah) and BVM ventilation during the apnoeic period (boo). Laryngoscopic approach was predetermined prior to randomisation, to limit the risk of bias.
What were they looking for?
The authors chose a primary outcome measure of successful tracheal intubation on the first attempt without any major complications. All well and good. Major complications were defined as visible pulmonary aspiration, desaturation below 95%, major haemodynamic instability, sustained arrhythmia, cardiac arrest or anaphylaxis occurring up to 10 minutes from induction. There were loads of secondary outcomes which you can peruse at your leisure. All outcome data was reportedly collected by independent observers and/or research assistants, to limit missing data and ensure accuracy of reporting.
What about the stats?
The non-inferiority assumptions are always worth chatting about. The authors give some reasonable justification for their choice of non inferiority margin, which was an absolute increase in the primary outcome of 7% and a relative increase of 8.75%. They report these assumptions in the context of an expected first pass success rate of 80%.
But let’s just take a second to think about what that means. That means that if the lower limit of the 95% CI estimate for first pass success rate without complications using remifentanil (and no NMB) was above 73%, the authors would conclude non-inferiority. I think this is fascinating. Would we really accept an RSI algorithm resulting in a 25% failure or complication rate, if it meant we didn’t need to use NMB? I don’t know about that. It would make me feel very uncomfortable.
The authors planned an intention to treat (‘as randomised’) and a per protocol analysis.
Ok, and the big news?
Well, the headline is that first pass success without complication occurred in 66.1% of the remifentanil group and 71.6% in the NMB group. After adjusting for randomisation group and strata, the between group difference here was -6.1%. The lower limit of the 95% confidence interval was -11.6%. Hence the intervention was not deemed to be non inferior to control. Not only that, but the upper limit of the 95% CI was negative, meaning this was a statistically significant result demonstrating that remifentanil was inferior to the use of NMB. Keep on blocking everyone. Happy days.
There is more to unpick than that though. These are LOW rates of FPS without complication and it’s worth thinking about why that might be. This information is helpfully presented in table 3, as components of the primary outcome. By far the biggest driver of the primary was the difference in successful first attempt intubation; 5% lower in the remifentanil group. No other results look as interesting, but there were more recorded episodes of aspiration, hypoxaemia and haeemodynamic instability in the remi group.
Some of the information in table 2 shines a light on this as well. Although these patietns were well preoxygenated, the majority of patients were intubated by relatively junior members of the team, videolayrngoscopy was used only 15% of the time, apnoeic ventilation of any kind happened in only around 1% cases, >70% of NMB patients got sux and propofol was the hypnotic of choice in >97% RSI events. This does not neceesarily translate to how we would all undertake this risky procedure in our countries and environments.
Among adults at risk of aspiration during rapid sequence intubation in the operating room, remifentanil, vs neuromuscular blockers, did not meet the criterion for noninferiority w regard to successful intubation on first attempt w/o major complications. https://t.co/ENvKxI7Oef
— JAMA (@JAMA_current) January 3, 2023
But this data does remind us of the risks – Although successful intubation occurred on the first attempt in >93% cases in the NMB group, haemodynamic instability occurred in 20% cases and hypoxaemia in 6.7%. As I have said before, this is a risky business. You should make it easy for yourselves. Not harder.
Anything else?
There has been lots of evidence on RSI recently. It is really worth your time to look it over and see how you can influence practice locally. Data from this trial, the INTUBE study (4) and PREPARE II (5) all highlight the risks. And now we are starting to understand what does and doesn’t work, we can unpick some of the dogma and promote some evidence based practice in this area.
You may feel that this is not your area of expertise, or it is not within your scope to influence. I would argue that if you work in emergency medicine or critical care in any form, then these are your patients, this is your evidence and what happens in your Resus room happens because of you, not in spite of you, no matter who is holding the laryngoscope.
More on some of this at the upcoming RCEM CPD conference….
Until then – keep on blocking. And give Neil Young a listen for old times sake.
Dan
References:
1 – NHLBT ALI CTN. Early Restrictive or Liberal Fluid Management for Sepsis-Induced Hypotension. NEJM 2023; DOI: 10.1056/NEJMoa2212663
2 – METRC. Aspirin or Low-Molecular-Weight Heparin for Thromboprophylaxis after a Fracture. NEJM 2023; 388:203-213
3 – Grillot. Effect of remifentanil vs neuromuscular blockers on during rapid sequence intubation on successful intubation without major complications among patients at risk of aspiration. JAMA. 2023;329(1):28-38. doi:10.1001/jama.2022.23550
4 – Russotto et al. Intubation Practices and Adverse Peri-intubation Events in Critically Ill Patients From 29 Countries. JAMA. 2021;325(12):1164-1172. doi:10.1001/jama.2021.1727
5 – Russel et al. Effect of fluid bolus administration on cardiovascular collapse among critically ill patients undergoing tracheal intubation. JAMA. 2022;328(3):270-279. doi:10.1001/jama.2022.979
