I’ve recently been spending a bit more time in the pre-hospital environment and it’s been interesting to compare practice there as compared to the emergency department. One theme has been that in the prehospital environment clinicians are seemingly more cautious when it comes to drug doses in procedures such as RSI and sedation. Perhaps this is to be expected as in the prehospital environment there is less back up and support than that available in a warm and dry resus room. In Virchester ED we have previously taught a standard regime for ED RSI using Fentanyl, Rocuronium and Ketamine in a 3:2:1 regimen for stable patients and 1:1:1 for the unstable patients. However, in prehospital care we have largely adopted a 1:1:1 for all patients and less than that in the unstable (including the omission of fentanyl all together). This was also the teaching on the excellent PHEA course (Prehospital emergency anaesthesia) run by GNAAS (Great North Air Ambulance Service) when I completed it in 2020. Both local and course based instructors have emphasised that the supposed cardiovascular stability of fentanyl is not always the case in practice.
This is an interesting challenge as in my ED training I have been encouraged to use fentanyl as part of of an approach to reduce the stress and impact of laryngoscopy/intubation. Something of particular concern in the many head injured patients that we see (Ed – though I’m not always sure that I see colleagues leave it sufficient time to work before giving the other drugs).
Is this approach evidence based though, or simply anecdotal? What is the evidence base for reduced/omission of fentanyl in prehospital patients requiring RSI?
A recent paper from our friends in Sydney HEMS may address this issue. The abstract is below, but as always please read the full paper and make up your own mind.
What kind of paper is this?
As a randomised controlled trial this is the design that we want when testing an intervention. In this case a comparison between patients receiving fentanyl as part of an RSI vs. those who do not. Patients were randomised using block randomisation (in groups of 8). It’s unclear whether these blocks were stratified by location (ideally they would be). Syringes were labelled in pharmacy which allowed the clinicians to be blinded to the treatment allocation (fentanyl or saline).
Tell me about the patients and setting
The study was conducted in the emergency departments of five hospitals in Australia. All patients requiring an RSI were screened. The only exclusions were for service pressures, allergy or paralysis only regimen.
What about the intervention?
The authors tried to keep the regimen for RSI as standard as possible in terms of process, but the drug doses of other drugs such as ketamine was left to the discretion of the treating physician. For example this did leave a range of available ketamine doses from 0.5-2mg/Kg. This was presumably necessary to match induction dose to the clinical status of the patient. In order to account for this with regard to fentanyl dose they matched the dose of fentanyl to that of the chosen amount of ketamine in a 1:1 ratio. Thus a patient receiving 80mg of ketamine would receive 80ug of fentanyl. This seems a pragmatic approach which reflects the common practice of reducing both drugs in patients who are cardiovascularly compromised, and also accounts for changes in body size/age.
The study drug was given first and then the ketamine, then rocuronium. The authors state that there was no pause between fentanyl and the other drugs.
What about the outcomes?
The primary outcome was whether the patient’s systolic blood pressure fell outside of the 100-150mmHg at any recorded time point up to 10 mins post induction. Patients with a pre-RSI SBP>151 were considered abnormal if the SBP rose by more than 10% of fell outside the 100-150mmHg range. For those with a pre-RSI SBP of <99 then a 10% fall or fall outside the range 100-150 was considered abnormal.
A range of secondary outcomes were also recorded.
Tell me about the results
302 of 476 screened patients were randomised into the study. 290 made it through to analysis. The groups were pretty similar at baseline, and they were seemingly pretty stable from a cardiovascular perspective. Of note just 20 patients had an SBP <99, and 35 had vasopressor use at intubation. 92% had an SBP in the 100-150mmHg at induction.
In terms of the primary outcome 66% of patients in the fentanyl group met the criteria vs. 65% in the placebo group. Clearly this represents no difference and is the main conclusion of this study. The primary outcome is that fentanyl does not affect the pre-determined outcome that this study was designed and powered to find.
There were of course a number of secondary outcomes that we should consider as hypothesis generating. The authors highlight that 29% of fentanyl patients vs. 16% of placebo patients became hypotensive. This is a difference of 13% (95% CI of 3-23%) and is statistically significant, but we should be cautious as it is not the primary outcome and is one of a number of comparisons that have not been adjusted for multiple analyses. Similarly there was less hypertension in the fentanyl group.
What do these results mean?
The headline conclusion should be that this study did not show a clinically significant difference in SBP in RSI when fentanyl is compared to placebo. I say this because the authors defined clinical significant before the study and then tested it, finding no difference.
The secondary outcomes relating to hypotension and hypertension are interesting and arguably chime with my anecdotal experience in ED and Prehospital care. However, they were not primary outcomes for this study and as such are more fragile. It is likely that fentanyl does cause a decrease in blood pressure on the basis of these results, but there is still some uncertainty.
What is not clear from the primary outcome result is how this works for individual patients. This is a function of how the analysis has been performed. When we look at average changes this often hides very significant changes in a small group of patients. I am arguably more interested in whether fentanyl causes a very significant decrease in blood pressure for some patients rather than whether a population stays within a particular band, or whether there is a small ‘average’ change in blood pressure. Outliers, are interesting from a clinician perspective as those patients in whom there is a very significant change in blood pressure, may arguably suffer the most harm. In this study a patient could potentially reduce their SBP from 149 to 101mmHg (a very significant change) and not register as an adverse result. This is not a criticism per se, but a just a reminder that summary statistics ‘summarise’. Figure 3 in the paper goes some way to address this, and whilst it’s tricky to identify if there are individual exceptional results, it does look as though there are a number of patient in whom SBP changes are clinically significant (though for individual patients we cannot tell if they are going up or down in that figure).
It’s also worth reiterating that there are very few patients (20) who started off with low BP, and that’s the group pf patients in whom I am most concerned with giving fentanyl to in clinical practice.
A larger number of patients started with blood pressures over 150 (24% and 30%), and within that group a reduction in SBP may be beneficial (pathology dependent).
I’m also interested to know why SBP was chosen over mean arterial pressure, which in my practice is more commonly used as a resuscitation goal.
Fentanyl is thought to have benefits which are not tested in this study and we must be mindful that the outcomes here are not patient outcomes, but (as Ken Milne would say), they are MOOs (monitor related outcomes). On that basis these are not definitive, but they are interesting. They do chime with what we have seen in clinical practice and subsequent changes to clinical practice too. I was also taught that fentanyl is most likely to be of benefit if given as a pre-treatment as opposed to being just prior to induction (as in this study). Thus the potential benefits of fentanyl, and even the elements of cardiovascular stability may be masked or enhanced by the method of administration.
Should we stop using fentanyl?
Remember that the primary outcome in this study showed no difference. The secondary analyses are interesting, but should be considered hypothesis generating, although they are also in keeping with other observational studies (see refs). From my perspective I will continue to use fentanyl on a case by case basis based on the clinical picture, CVS parameters and resuscitation goals.
Fentanyl versus placebo with ketamine and rocuronium for patients undergoing rapid sequence intubation in the emergency department: The FAKT study-A randomized clinical trial https://pubmed.ncbi.nlm.nih.gov/35064992/
Fentanyl Pretreatment in Rapid Sequence Intubation https://www.tomwademd.net/fentanyl-pretreatment-in-rapid-sequence-intubation/
Association of fentanyl use in rapid sequence intubation with post-intubation hypotension https://pubmed.ncbi.nlm.nih.gov/29653790/