By Tom Heaton
“Just when you thought it was safe to return to your anaesthetic room, the makers of ‘Etomideath: An Anaesthetic Horror Story’, bring you the terrifying sequel - Etomideath 2: The Heartbreaker!! No RSI is exempt, no adrenals are safe, and this time it’s after your heart!”
After our recent look at etomidate and its fall from favour as an induction agent, my eye was caught by this recent study on etomidate in the March 2014 publication of Anesthesiology. It primarily caught my attention as it appeared to be trying to look at some of the questions that we had discussed around the last post; namely where do we stand with etomidate in non-septic patients. The evidence of adrenal suppression seems to be there even if the clinical consequences aren’t clear, but most of this examination of clinical effects seem to be focused on the septic patient needing intubation where perhaps this adrenal suppression could be seen as being more problematic. There seem to be (to me at least) a good selection of other clinical scenarios where having a cardiovascularly gentle induction agent might prove advantageous e.g. a multitrauma patient with head injury where ketamine would definitely raise the ICP (little something for you there John), but even brief peri-induction hypotension is bad. In this paper the authors look at the use of etomidate in another group of patient where it might be advantageous; those undergoing cardiac surgery who probably have knackered hearts to start with. Hopefully this preamble is enough to get your interest stoked too whilst we go in for a closer look at the paper. If you haven’t been able to have a look at it for yourself yet then the link is below.
After our recent look at etomidate and its fall from favour as an induction agent, my eye was caught by this recent study on etomidate in the March 2014 publication of Anesthesiology. It primarily caught my attention as it appeared to be trying to look at some of the questions that we had discussed around the last post; namely where do we stand with etomidate in non-septic patients. The evidence of adrenal suppression seems to be there even if the clinical consequences aren’t clear, but most of this examination of clinical effects seem to be focused on the septic patient needing intubation where perhaps this adrenal suppression could be seen as being more problematic. There seem to be (to me at least) a good selection of other clinical scenarios where having a cardiovascularly gentle induction agent might prove advantageous e.g. a multitrauma patient with head injury where ketamine would definitely raise the ICP (little something for you there John), but even brief peri-induction hypotension is bad. In this paper the authors look at the use of etomidate in another group of patient where it might be advantageous; those undergoing cardiac surgery who probably have knackered hearts to start with. Hopefully this preamble is enough to get your interest stoked too whilst we go in for a closer look at the paper. If you haven’t been able to have a look at it for yourself yet then the link is below.
What's The Paper About?
As I’ve covered so far, this paper is trying to give more information to the debate about etomidate’s place as an induction agent. The authors point out the controversy about the significance of this relative adrenal suppression, noting that there is little information about outcomes in this subgroup of patients (those undergoing cardiac surgery) where etomidate’s profile could be considered desirable.
The paper is a retrospective analysis of 3,127 patients undergoing some form of cardiac surgery at a single centre, looking at whether the use of a single dose of etomidate at induction had an impact on certain outcomes. These outcomes included: in-hospital mortality, length of hospital stay, length of mechanical ventilation, and incidence of severe hypotension.
The paper is a retrospective analysis of 3,127 patients undergoing some form of cardiac surgery at a single centre, looking at whether the use of a single dose of etomidate at induction had an impact on certain outcomes. These outcomes included: in-hospital mortality, length of hospital stay, length of mechanical ventilation, and incidence of severe hypotension.
What Did They Find?
The results were in line with what the authors (and myself) were expecting; no significant difference in any of the primary outcomes. 61.7% of the patients had received etomidate as the induction agent and this was mostly compared against propofol with a splash of midazolam occasionally thrown in (all patients received fentanyl).
The adjusted odds ratio for in-hospital mortality with etomidate use was 0.75 (95% confidence intervals 0.45 - 1.24) and for severe hypotension was 0.80 (CIs 0.58 - 1.09). In addition, the adjusted hazard ratio for length of hospital stay was 1.07 (CIs 0.97 - 1.18) and for time to cessation of mechanical ventilation was 1.10 (CIs 1.00 - 1.21).
Overall then it seems there is little in the way of instant death from giving etomidate to these patients. These outcomes were pretty much maintained even after the variety of statistical fiddling and adjusting that is generally done in these retrospective studies and the authors played around with the data a bit more in other ways to look at a few other areas (I won’t go into any more detail than that though).
The adjusted odds ratio for in-hospital mortality with etomidate use was 0.75 (95% confidence intervals 0.45 - 1.24) and for severe hypotension was 0.80 (CIs 0.58 - 1.09). In addition, the adjusted hazard ratio for length of hospital stay was 1.07 (CIs 0.97 - 1.18) and for time to cessation of mechanical ventilation was 1.10 (CIs 1.00 - 1.21).
Overall then it seems there is little in the way of instant death from giving etomidate to these patients. These outcomes were pretty much maintained even after the variety of statistical fiddling and adjusting that is generally done in these retrospective studies and the authors played around with the data a bit more in other ways to look at a few other areas (I won’t go into any more detail than that though).
Is It Any Good?
For starters with 3,127 patients it’s not a bad size of group, giving some strength to any conclusions we can draw. However, as the authors note they have still ended up with some pretty wide confidence intervals on areas of interest (notably mortality), going on to point out that they would need a pretty hefty sample size to detect a significant difference in outcome when the mortality was as low as it was (3.4%).
The most obvious limitations are evident when we start looking at the study design as the strength of the conclusions will be limited by the retrospective observational nature. A double blinded RCT this is not. The usual look at the possible confounding factors was undertaken and identified a few possibly important differences between the groups. This includes more patients in the group not receiving etomidate being described as in cardiogenic shock (4% vs 1.3%), patients who received etomidate being more likely to have a history of congestive heart failure (23% vs 18.3%) and different approaches to the way the CABG surgery was undertaken between the groups (off-pump vs on-pump).
This is the point in paper analysis that I always start to lose faith in what I am seeing in the conclusions. When the starting groups are notably different from each other like this, even with all the fancy statistical adjustment I start to lose faith that outcome differences between the groups are linked to the single variable we are interested in. The authors highlight that in these patients, factors such as pre-op creatine, cardiogenic shock, type of surgery, and history of CHF were all predictive of mortality. If these factors are, what other unmeasured ones could be too?
Before I start sounding too pessimistic there are several laudable points to this study. The authors have done their best to make the most of the nature of their data and provide information that we can try and base conclusions upon. The descriptions of the methods was clear and statistical analysis of confounders was set on an ‘a priori’ basis. Plenty of additional description was provided to explain certain quirks that cropped up in the data, such as some patient arising already intubated and therefore not receiving any induction agent, or the institute's tendency towards performing CABG procedures off-pump.
The most obvious limitations are evident when we start looking at the study design as the strength of the conclusions will be limited by the retrospective observational nature. A double blinded RCT this is not. The usual look at the possible confounding factors was undertaken and identified a few possibly important differences between the groups. This includes more patients in the group not receiving etomidate being described as in cardiogenic shock (4% vs 1.3%), patients who received etomidate being more likely to have a history of congestive heart failure (23% vs 18.3%) and different approaches to the way the CABG surgery was undertaken between the groups (off-pump vs on-pump).
This is the point in paper analysis that I always start to lose faith in what I am seeing in the conclusions. When the starting groups are notably different from each other like this, even with all the fancy statistical adjustment I start to lose faith that outcome differences between the groups are linked to the single variable we are interested in. The authors highlight that in these patients, factors such as pre-op creatine, cardiogenic shock, type of surgery, and history of CHF were all predictive of mortality. If these factors are, what other unmeasured ones could be too?
Before I start sounding too pessimistic there are several laudable points to this study. The authors have done their best to make the most of the nature of their data and provide information that we can try and base conclusions upon. The descriptions of the methods was clear and statistical analysis of confounders was set on an ‘a priori’ basis. Plenty of additional description was provided to explain certain quirks that cropped up in the data, such as some patient arising already intubated and therefore not receiving any induction agent, or the institute's tendency towards performing CABG procedures off-pump.
Final Thoughts
Well a speedy run through another paper that throws some more fuel on the ‘etomidate = death’ fire/debate that is still burning. I liked the paper’s approach of a look outside the popular ‘etomidate in sepsis’ arena where etomidate has taken the biggest battering; an approach which nicely follows some of the thoughts we had left over at the end of the last post. Just because they haven’t detected a difference doesn’t mean that there isn’t one but it provides more information for us to work with, despite its flaws.
I continue to remain sceptical about etomidate’s lethality (certainly outside of sepsis at any rate) so I may be looking at trials like this through a biased lens, but I keep finding myself coming back to applying such critical appraisal to clinical practice. What benefits does etomidate provide and what are the drawbacks? I'm sure many will argue that there are other induction agent options out there for a variety of situations, but even ketamine has its drawbacks and I haven't yet seen enough to convince me that etomidate has no place in the future of anaesthetics or critical care practice.
As always please let me know your thoughts on the topic and let me know of any good links or articles that you happen to have stashed away on the topic (we have a few good links at the end of the last etomidate post). Thanks for reading and hopefully some more posts on this topic soon (as well as on a variety of others of course). All the best.
Tom Heaton
I continue to remain sceptical about etomidate’s lethality (certainly outside of sepsis at any rate) so I may be looking at trials like this through a biased lens, but I keep finding myself coming back to applying such critical appraisal to clinical practice. What benefits does etomidate provide and what are the drawbacks? I'm sure many will argue that there are other induction agent options out there for a variety of situations, but even ketamine has its drawbacks and I haven't yet seen enough to convince me that etomidate has no place in the future of anaesthetics or critical care practice.
As always please let me know your thoughts on the topic and let me know of any good links or articles that you happen to have stashed away on the topic (we have a few good links at the end of the last etomidate post). Thanks for reading and hopefully some more posts on this topic soon (as well as on a variety of others of course). All the best.
Tom Heaton
Image courtesy of chrisroll/FreeDigitalPhotos.net
RSS Feed