Two previously published trials (ARDS et Curarisation Systematique [ACURASYS] and Reevaluation of Systemic Early Neuromuscular Blockade [ROSE]) presented equivocal evidence on the effect of neuromuscular blocking agent infusions in patients with acute respiratory distress syndrome. Host Kyle B. Enfield, MD, FCCM, is joined by Matthias Eikermann, MD, to discuss the different sedation regimens between these two trials as well as the different regimens within the ROSE trial between treatment and control groups. They will also discuss the hypothesis that the proportion of deeper sedation is a mediator of the effect of neuromuscular blocking agent infusions on mortality (Wongtangman K, et al. Crit Care Med. 2021;49:1137-1148; Jabaley C, Crit Care Med. 2021;49:1211-1213).
Dr. Eikermann is a professor of anesthesia at Harvard Medical School and an attending anesthesiologist and vice chair of faculty affairs in the Department of Anesthesia, Critical Care and Pain Medicine at Beth Israel Deaconess Medical Center. This podcast is sponsored by Sound Physicians.
Estimated Time: 23:44 min
This podcast is sponsored by Sound Physicians, the employer of choice for critical care physicians, where we seek to transform acute episodes of care. At Sound Physicians, we ensure physicians have the time and resources needed to deliver compassionate care that measurably improves quality and lowers the cost of healthcare for patients in the communities we serve. For more information, please visit careers.soundphysicians.com.
Kyle B. Enfield, MD, FSHEA, FCCM (Host): Hello, and welcome to another edition of iCritical Care. I’m your host, Dr. Kyle Enfield. Today I am speaking with Matthias Eikermann, MD, who is the chair of anesthesia at Montefiore Medical Center and professor at Albert Einstein College of Medicine. Today we’re going to be discussing Dr. Eikermann’s paper that was released in the July 2021 edition of Critical Care Medicine, “Optimal Sedation In Patients Who Receive Neuromuscular Blocking Agent Infusions for Treatment of Acute Respiratory Distress Syndrome—A Retrospective Cohort Study From a New England Health Care Network.”
Dr. Eikermann, before we begin, do you have any disclosures for the audience?
Dr. Eikermann: Yes. I have been receiving investigator-initiated grants from Merck. Merck produces rocuronium and reversal agents for steroidal neuromuscular blocking agents. There is no direct relation to the use of neuromuscular blocking agents in critical care medicine.
Dr. Enfield: Thank you. The challenge for most intensivists in the era of COVID and ARDS is around the use of neuromuscular blockade. Two very large trials, ACURASYS and ROSE, led us to this conundrum of what should we do with neuromuscular blockade in patients with refractory hypoxemia and severe ARDS. You undertook a trial to try to understand this. What was your impetus for wanting to investigate the impact of sedation on these outcomes?
Dr. Eikermann: I have long had an interest in the clinical use of neuromuscular blocking agents. Most of my work is related to the operating room, but since I’m a critical care physician, I was always curious as to whether or not it would be reasonable to use neuromuscular blocking agents in the ICU. Initially I was very critical because I was focused on the downsides of long-term use of muscle relaxants on muscle strength. And I was very interested in early recovery and making sure that our ICU patients are getting strong as they recover from critical illness. I was initially critical and gave talks years ago on the negative effects of using muscle relaxants in the ICU. Then, years ago, I saw the ACURASYS trial by Papazian and colleagues, where they clearly showed an advantage of neuromuscular blocking agent use in the ICU.
At this time, I had identified some methodologic issues related to that trial. And I was looking forward to seeing the results of the ROSE trial. There was basically a contradiction of the results at first glance. In the ROSE trial, a larger sample sized trial, the investigators could not show a beneficial effect of neuromuscular blockade in ARDS. So then we looked closer into the methods and compared the methods of the ACURASYS trial and the ROSE trial and found that the ACURASYS trial was maybe a little more methodologically rigorous because it used the same sedative regimen for all patients whereas in the ROSE trial, the patients in the neuromuscular blocking agent group received deep sedation. The patients in the control group who did not get neuromuscular blocking agents received light sedation, and there’s more and more information available on the downsides of deep sedation in the ICU. So we hypothesized that some of the differences between the ROSE and ACURASYS trials could be explained by the differences in the sedation regimen. And that’s what we were able to analyze, based on our data.
Dr. Enfield: I think a lot of us have had those same challenges. I know many people are looking forward to reading this study. One of the challenges that I had reading it is that this is a retrospective study of exposure and mediator variables, which is not something many of us have been comfortable looking at in trying to understand the complex interactions between two things. Can you briefly describe the system you used to analyze these impacts on your cohort?
Dr. Eikermann: First I must say that I really believe that well-conducted observational trials can teach us things that sometimes even randomized controlled trials (RCTs) cannot teach us because we study all patients who present to an institution rather than focusing only on a very small, well-defined group of patients who might not be a good reflection of all patients being treated in the ICU. Also, whenever you look at your observational research, you need to convince readers that you have used adequate confounder control. And patients who get neuromuscular blocking agents are the ARDS patients who are sicker, who have a lower P/F ratio, and maybe more severe hypoxic respiratory failure.
It’s always the investigator who needs to convince, first the reviewers and editors and then the readers, that we have addressed those concerns of inadequate confounder control. We did that. If you look at the paper and the supplements and sensitivity analysis, you will probably agree that all differences in case severity between the different groups—those with and without neuromuscular blocking agent treatment—were taken care of. You also mentioned the mediation analysis, which is pretty cool; it gets you a little bit closer to arriving at causal interferences. Basically the mediation analysis can be explained in three steps. I’m happy to walk you through those.
First, you look at whether the exposure—neuromuscular blocking agent—is associated with the end point, that is, a mortality. And that was the case. In our cohort, the patients who received neuromuscular blocking agents, even when controlling for confounders, had higher mortality. The next step is to add in the variable that you believe could be a mediator, which is a variable that explains the association between the exposure and the outcome. In this case, it was a high proportion of deep sedation. When we then added that variable—higher proportion of deep sedation—to the model, the association of the exposure, which is muscle relaxants, and the end point, which is mortality in our study, disappeared. In other words, muscle relaxants only explained, in our cohort and the primary analysis, variance in mortality when we did not control for deep sedation.
The third step is basically an analysis of the effects of deep sedation on mortality. Deep sedation was a strong predictor of mortality. If you have those three regression analyses, you can get a variable, which is percent mediated through the mediator, here, a high proportion of deep sedation. And in this case it was almost unbelievable. We had to run this analysis several times, and several statisticians confirmed those findings. A hundred percent of the effect was mediated by deep sedation. In other words, in our study, the muscle relaxants were not associated with bad outcome, but they were associated with a higher proportion of deep sedation. And that in turn was a predictor of mortality. So those are the findings in a nutshell.
Dr. Enfield: That is fascinating. I think one thing we miss so often is the importance of observational research along with RCTs in critical care. Could you elaborate a little bit more on this? I find the same thing to be true. We have to control for so much in our RCTs that I feel we don’t always truly represent the patients we’re caring for in the ICU.
Dr. Eikermann: Yes, that is certainly true. And that’s particularly true for drug trials, particularly for drug company-sponsored trials. Sometimes you have more exclusion criteria than number of patients included in the trial because the sponsor of the study wants to be 100% sure that the results are in line with the expectations. And therefore we have good confidence that in this subcore of patients who meet all criteria, for example, in ARDS, you could say it is pneumonia, for example in patients who are younger than 35 years, and the onset of the respiratory symptoms should not have been longer than five days, and the patients cannot have renal failure, liver failure, and so forth. And you add exclusion criteria and then you find a positive effect of your intervention. That is important to demonstrate proof of concept, but that cannot be the only piece of information that we look into.
We actually treat patients who are older than 35 who may have some relevant comorbidities. Therefore observational research studies are important. Particularly if we find that the observational studies can be validated by some of the observations in subgroups in RCTs, then I think it’s extremely powerful. For example, there was surprising overlap in the findings of the ROSE trial, which is an RCT, probably the highest level of evidence on ARDS and muscle relaxants, and our study on neuromuscular blocking agents. I can talk more about that if you like.
Dr. Enfield: Yes, I would love for you to expand a little bit on that. I have one follow-up question, but I’d like to hear your thoughts on that first.
Dr. Eikermann: For example, even in the context of the ROSE RCT, patients who received neuromuscular blocking agents also had lower Richmond Agitation-Sedation Scale (RASS) scores after termination of the neuromuscular blocking agent infusion than the controls. So even in this RCT, the clinicians gave the sedatives at a higher dose for a longer period in patients who received neuromuscular blocking agents compared with those who did not.
This is one of the most important findings of our study: we think that, often when clinicians are used to looking at patients who are subject to controlled ventilation with a set respiratory rate and peacefully interacting or not interacting with the ventilator, they want to keep up the status quo, whereas in patients who did not receive neuromuscular blocking agents, maybe the clinicians tried earlier to track the sedatives to allow for some spontaneous ventilation. That was an example of how we found, in a subgroup of the ROSE trial, some evidence for the validity of the finding that we found important in our observational study.
Dr. Enfield: Yes, one of the things that I found fascinating about your paper was the prolongation of deep sedation. Do you have some thoughts about what was causing that? We’ve seen in some other small observational studies, particularly in the past year, that patients were getting very high doses of medications for sedation to achieve a RASS goal and there seemed to be more medications being used than necessary in other patients to achieve that same goal.
Dr. Eikermann: Yes, I think that often in ARDS treatment, the sedation and neuromuscular blocking agent regimen is sometimes an anxiety trigger for patients, clinicians, and nurses. Also, junior staff might be so afraid of a progression of the hypoxic respiratory failure if they change anything that they just don’t do it. And they don’t apply rigorously what we know is important from other areas of critical care not directly related to ARDS, which is the daily interruption of sedation, spontaneous awakening trials, and so forth. I have seen this in several areas where I’ve worked, in Boston and also in Germany.
Anxiety is certainly a factor that sometimes immobilizes the clinicians to try to see whether the patient tolerates a lower level of sedation and maybe some level of spontaneous breathing. After the ACURASYS trial was published, there was a reason for keeping patients paralyzed for a long time: because the data were so positive.
Now, after the ROSE trial, and maybe also some meta-analysis on RCTs, on observational data, also in light of the current recommendations published in Intensive Care Medicine as a guideline, we conclude that it’s just not safe to keep patients paralyzed and deeply sedated long term. It adds a specific risk. And we should not forget about the basic principles of trying, every shift, to decrease the patient’s level of sedation and increase mobility level.
Dr. Enfield: Tagging onto that, what are your thoughts around the specificity of the RASS as a way of monitoring sedation? We know that the clinical skill of determining whether the patient is responsive at different levels of sedation varies among clinicians. What impact do you think the individual’s anxiety around making sure people are truly in a deep sedative state before paralyzing them may be influencing some of these results?
Dr. Eikermann: The RASS is a validated instrument and, if clinicians follow instructions closely, I would say we should use the data for clinical decision-making. I’m not aware of a better scale that is more valid and less prone to bias. But I agree with you that probably the expectation of the clinician and what the sedation level should be has an effect on the assessments. As you know, during neuromuscular blockade, you cannot make assessments of sedation level.
During the period of actual neuromuscular blocking agent infusion, the sedation level cannot be assessed based on a rating scale. The question is, what should we do during that time? In our institution in Boston, where we conducted the study, our policy was that, once the neuromuscular blocking agent infusion was started, we do not decrease the sedative dose to be on the safe side and to make sure that we don’t have this horrible scenario of a patient who is awake and paralyzed. The question is, should we use other strategies to monitor sedation, such as EEG-based technologies? I think those are interesting questions.
Dr. Enfield: I think we’ve all had the same idea about whether there is an EEG method we could use to approach sedation in the paralyzed patient, because that does create a lot of challenges. Hopefully someone will do that research at some point. As you concluded this study, you said you were really surprised about the degree that deep sedation influenced the outcome or the modifier effect there. Were there any other findings that particularly surprised you?
Dr. Eikermann: Well, personally, my bias was that probably neuromuscular blocking agents do not have beneficial effect on mortality in ARDS. That is just a bias because it was not really based on valid data. When we looked into the subgroup of patients who were deeply sedated, we found a significant beneficial effect of neuromuscular blocking agent infusions on mortality. I was surprised by it. But others were probably not surprised because that basically supports the validity of the ACURASYS study, which means that if you really need to sedate the patient deeply, then neuromuscular blocking agents might improve outcomes. I personally found this surprising, but it makes sense. In conjunction with the other data supporting the use of muscle relaxant infusions in ARDS, it will probably also change my personal clinical management if patients need deep sedation.
Dr. Enfield: So where do we go next? What’s the next step in this research line of thinking that you are considering now?
Dr. Eikermann: In terms of research studies, it would probably be nice to do a higher sample sized trial, similar to the ACURASYS trial, that focuses on patients who clearly need deep sedation to control their respiratory effort. We are afraid of lung injury as a consequence of high inspiratory effort that cannot be controlled based only on optimal ventilator management. That sometimes happens in patients with severe ARDS and low P/F ratio. With rigorous criteria and inclusion only of patients who need deep sedation to control their respiratory efforts, and in both arms, the same sedation regimen is applied. That would probably be the final piece. I predict that, if adequately powered, it will lead to a positive finding. We are not for profit, but I think it’s definitely worth studying.
Other than that, I would say to make sure that we do not paralyze or deeply sedate our ARDS patients unnecessarily. Treat the ARDS patients clinically as other ICU patients who are mechanically ventilated and do the spontaneous awakening trials and spontaneous breathing trials, synchronize them, and make that attempt on every shift. We don’t know yet to which level we should wake up our patients. Obviously there is also a threat in 100% waking up a patient who is not ready for spontaneous breathing. I’m hoping that future studies give us some additional pieces of evidence that help clinicians make that decision on how to wake them up. What is the right safe level? What variable should we use? And when should we stop the spontaneous awakening and spontaneous breathing trial?
Dr. Enfield: I really like those two takeaway messages: to use paralytics on patients who are deeply sedated based on this information, and to remind us all of the importance of the ICU Liberation pathway of checking every shift to make sure a patient still needs that level of sedation. I do find that that is probably the paramount piece, that we like our patients comfortable, not interactive, but we do have to check all the time to make sure that they’re not ready to progress. Are there any other major takeaways for the audience?
Dr. Eikermann: No. I encourage everyone to read the wonderful editorial by Craig Jabaley, which was published alongside our paper. I think it really summarizes our findings and also the context and the statistical analysis-related considerations of mediation analysis nicely. Thank you for having me on and giving me a chance to explain our study.
Dr. Enfield: I really appreciate you coming on iCritical Care. For the audience, the July issue of Critical Care Medicine has both Dr. Eikermann’s study and the accompanying editorial. And I recommend them to you as well because they are great pieces of scientific literature. I think they’ll really open our thoughts about future studies, as well as the use of paralytics and sedation on our ICUs. For iCritical Care, I’m your host, Dr. Kyle Enfield.
This podcast is sponsored by Sound Physicians, the employer of choice for critical care physicians, where we seek to transform acute episodes of care. At Sound Physicians, we ensure physicians have the time and resources needed to deliver compassionate care that measurably improves quality and lowers the cost of healthcare for patients and the communities we serve. For more information, please visit careers.soundphysicians.com.
Kyle B. Enfield, MD, FSHEA, FCCM is an associate professor of medicine in the Division of Pulmonary and Critical Care at the University of Virginia, where he also serves as associate chief medical officer for critical care. He received his undergraduate degree, master’s degree in epidemiology, and medical doctorate degree from the University of Oklahoma. Before his internship, residency, and fellowship at the University of Virginia, he was an intern in communicable diseases at the World Health Organization. His clinical and academic interests include highly transmissible diseases, disaster response, emergency preparedness, and critical illness recovery.
Join or renew your membership with SCCM, the only multiprofessional society dedicated exclusively to the advancement of critical care. Contact a customer service representative at +1 847 827-6888, or visit sccm.org/membership for more information. The iCritical Care Podcast is the copyrighted material of the Society of Critical Care Medicine and all rights are reserved. Statements of fact and opinion expressed in this podcast are those of authors and participants and do not imply an opinion or endorsement on the part of SCCM, its officers, volunteers, members, or those of the podcast commercial supporter.
Some episodes of the iCritical Care Podcast include a transcript of the episodes audio. Although the transcription is largely accurate, in some cases it is incomplete or inaccurate due to inaudible passages or transcription errors and should not be treated as an authoritative record.