Critical Care Medicine CPP section. My name is Matthew Gurka, Adult Critical Care Pharmacy Specialist at Spectrum Health in Grand Rapids, Michigan. I will be moderating today's webcast. A recording of this webcast will be available to registered attendees. Log in to mysccm.org and navigate to the My Learning tab to access the recording. A few housekeeping items before we get started. There will be a Q&A after each of today's speakers. To submit questions throughout the presentation, type into the question box located on your control panel. You'll have the opportunity to participate in several interactive polls. When you see a poll, simply click on the bubble next to your choice. You may also follow and participate in live discussion on Twitter, following hashtag SCCM CPP JC and hashtag PharmICU. SCCM provides the following disclaimer. This presentation is for educational purposes only. The material presented is intended to represent an approach, view, statement, or opinion of the presenter, which may be helpful to others. The views and opinions expressed herein are those of the presenters and do not necessarily reflect the opinions or views of SCCM. SCCM does not recommend or endorse any specific test, position, product, procedure, opinion, or other information that may be mentioned. I'd like to introduce our speakers for today. Each will give a 15-minute presentation followed by a Q&A session. Our first presenter is Anastasia Ingeleit, PGY-2 Critical Care Resident at the University of Illinois Hospital and Health Sciences System in Chicago, Illinois. Our second presenter is Karen Ng, PGY-2 Critical Care Resident at Stanford Children's Health in Palo Alto, California. And our third presenter is Elena Etter, PGY-2 Critical Care Resident at Grady Health System in Atlanta, Georgia. And now I'll turn things over to our first presenter. All right, thank you, Dr. Gurka, for that introduction. Today, I will be presenting on the effect of continuous infusion hypertonic saline versus standard of care on six-month neurological outcomes in patients with traumatic brain injury. This is the COBE randomized clinical trial. Just to give some background on TBIs, it's estimated that 69 million patients experienced TBIs in 2019. Though the mortality in this patient population has decreased as our management of them has improved, incomplete recovery is still extremely common. This is kind of referring to their lack of return to their functional baseline. Most of this incomplete recovery is secondary to secondary brain injuries, such as elevated intracranial pressure, intracranial hypertension, and cerebral edema. The focus of TBI management is preventing these. So recently, literature has been examining the use of hypertonic fluid solutions to prophylaxe against these secondary brain injuries. There are these two trials looking at hypertonic solutions that have shown a decreased risk of intracranial hypertension, and hypertonic fluids were associated with higher hospital survival in a meta-analysis. So that brings me to the research question from the COBE clinical trial. So in patients with moderate to severe TBI, what is the impact of continuous infusion 20% hypertonic saline compared to the standard of care on neurological outcomes at six months? This was an investigator-initiated, multi-center, unblinded, parallel group, randomized control trial. It was unblinded to the investigators as the effects on serum sodium from a hypertonic solution would be very obvious at bedside. However, the people assessing the primary outcomes at the six-month mark were blinded to what group the patients were allocated to. This was conducted across nine French university hospitals and ICUs that treated at least 50 TBI patients every year. So level one trauma center is very comfortable working with this population. They were randomized to either the intervention group or standard of care in blocks of six patients randomized one-to-one. So the intervention here was a continuous infusion of 20% hypertonic saline given at 0.5 to one gram an hour for at least 48 hours. And this was titrated based off serum sodium levels that were monitored every eight hours for dose adaptation. The intervention was stopped when all specific therapies were stopped for at least 12 hours. So these were the specific therapies that were used in the standard of care arm alongside isotonic crystallides for maintenance and resuscitation fluids. These therapies included things like barbiturates, hypothermia, and decompressive craniectomy as needed. So this study included 18 to 80-year-old patients, so adults with moderate to severe TBI. This was defined as a Glasgow coma score of 12 or lower with CT findings consistent of TBI. And they had to be enrolled within 24 hours of their admission. They excluded pregnant patients for legal reasons. They excluded people who were already dependent for their ADLs at baseline, as well as patients with cervical spinal injury as these would affect the Glasgow outcome score which was the primary outcome. They also excluded people who had presented with imminent deaths. And they called fluid retention, which was pulmonary edema or ascites a contraindication to hypertonic saline. Their primary outcome here was the Glasgow outcome score at six months post-injury. And some secondary outcomes that we'll be talking about included mortality at six months, favorable neurologic outcomes at six months. These were defined as a Glasgow outcome score of six to eight as well as the evolutions of sodium, blood osmolarity, and intracranial pressure. They also looked at safety events including severe thromboembolic events and the incidence of central pontine myelinolysis. So they, the authors of this study looked at some previous smaller studies that showed hypertonic solutions improved Glasgow outcome scores by about 20%. So assuming that 70% of the patients in the control arm had a poor outcome, the authors needed to enroll 370 patients to detect a relative difference of 20% in that outcome score at six months. So that was their power analysis. The patients were analyzed based on randomization groups, so intention to treat, and some primary outcomes used proportional odds models. They had three a priori subgroup analysis I wanna talk about today, including severe versus moderate TBI, hyperosmolar therapy prior to randomization, and then time from the trauma to inclusion. They also identified a post-hoc subgroup analysis and looked at patients who either developed or did not develop intracranial hypertension. Secondary outcomes were also analyzed using logistic regression. So starting with the patients that were enrolled into this trial, 370 patients were identified and randomized and followed for six months. However, primary outcome data was not collected for 11 of these patients. Four patients refused their medical data after randomizations, and seven patients were not completely followed to six months. Two had guardianship problems, one patient was determined not to have a TBI, and four were lost to follow-up. You can see here that the average patient in this trial was a male in their mid-40s with a severe TBI, over 70% of the patients in both arms had a severe TBI. The average, or the median Glasgow Coma score for the patients was seven in both arms. Most of the patients did not present with hypotension or hypoxia at the time of their trauma, and over half of the patients in both arms received some sort of bolus hyperosmolar therapy before being randomized after about 12 to 13 hours in each arms. The average length of the intervention in that arm was 2.7 days, so these patients stayed on the hyperosmolar therapy for over 48 hours. Here's the primary outcome on this very nice bar graph, and I also wanted to take a second to talk about the Glasgow outcome score a little bit. So this is a functional outcome scale utilized in TBI research that assesses functional baseline. So a score of nine is upper good end of recovery, basically a full return to baseline, and a score of zero is dead. And for the sake of the primary outcome, they actually lumped a score of zero and one together as functionality-wise, the authors determined that this was similar, so they kind of analyzed that point together. You can see based on just visually looking at the graph that there's not a lot of differences between the intervention or the control group, and this statistically was not significant either. They also looked at the distribution of the scores between the two groups at six months and determined that there was no significant difference in the distribution of the scores between the intervention or the control. Some secondary outcomes included the intervention's effects on sodium and blood osmolarity, and no surprise here, the intervention was associated with higher blood osmolarity and serum sodium concentrations. So you can see in these graphs on the first 48 hours when the patients were receiving the hypertonic fluids with these orange boxes that they had higher serum sodiums and higher blood osmolarities, and this normalized about 48 hours after stopping the intervention. The intervention was also associated with a reduced risk of intracranial hypertension. Other secondary outcomes I wanted to highlight included mortality at six months. There was no significant difference here, as you can see, and then about 33% or 35% of patients in both groups experienced one or more episodes of intracranial hypertension, and moving down to favorable neurological outcomes at six months, and about 30% of patients experienced this in both arms, and this was not statistically significant. So those are sort of the main outcomes that I wanted to highlight as well. And then safety outcomes, looking at a severe adverse drug reaction of any type, you can see more patients experienced that in the intervention arm compared to the control arm. And then when you broke it down as to what the adverse drug reaction was, there was severe hypernatremia being kind of the most common in the intervention arm. This was defined as a serum sodium of greater than 160 MEQs per liter versus severe thromboembolic events were also slightly higher in the intervention arm compared to the control arm. And the incidence of central pontine myelinolysis was actually one patient in both groups. They did not report out their chloride values in this study, but they did highlight acute kidney injury, which was defined as a KDGO stage two or higher. Four patients experienced this in the intervention arm, eight patients experienced this in the control arm. So overall, the authors concluded after presenting their results that in a multicenter randomized clinical trial involving patients with moderate to severe TBI, continuous infusion of 20% hypertonic saline solution for a minimum of 48 hours did not significantly improve clinical outcomes as assessed by the Glasgow outcome score measured at six months. So some strengths of this study, I think the broad inclusion of multiple types of TBI is a strength in this trial's design and an attempt to make it more generalizable. Most of the studies that looked at hypertonic solutions in this population included only severe TBI. So this was an attempt to make it more generalizable. And I think it was appropriate to include these patients. I also think the scales that they utilized were appropriate to assess the outcomes. The Glasgow outcome scale is the sort of preferred way to measure functional outcomes in TBI patients. So I think that that was an appropriate choice of primary outcome. I also think that the other outcomes were pragmatic and appropriate. There were no increases in mortality seen in this study. The other studies that the authors referenced at the beginning looked at reported out mortality in the patients who received hypertonic saline and saw similar rates of mortality ranging from 11% to 25%. And I also do have to give the authors credit. They had some pretty small attrition in this study. They only lost four patients to follow up throughout the six months that they followed these patients. So I think that is a strength in their design. However, some limitations, though they lost 3% of their total study population. So they're underpowered for the primary outcome. And though it makes sense that this trial was on blinded, there's really no way to blind the intervention if you're giving someone hypertonic saline. It does weaken the strength of the study. And when you look at the confidence intervals of the outcomes, you can see they're extremely wide, which also suggests its lack of power. I also think it's a big limitation that they didn't report serum chloride values, even though they were collected. We know that hyperchloremia is associated with poor outcomes, including AKI, the need for renal replacement therapy, as well as mortality. And over 70% of the patients had a severe TBI. So we know that they tried to include as many patients as they could, made it more generalizable, including the moderate group. We did some subgroup analysis to see that whether or not the type of TBI severity was impacting the outcome. And the subgroup analysis did not find that difference significant. But it could be said that the inclusion of multiple types of TBIs clouded the primary outcome. And we didn't see the benefit that we would have seen if we just looked at moderate or severe TBI separately. So before I share my conclusions with the audience, I did just want to ask a couple polling questions here. So would you consider hypertonic saline via continuous infusion for prophylaxis of secondary brain injuries in patients with TBI? Please select yes or no. All right, so a majority of the audience said no, 76% said no, 24% said yes. I tend to agree with the majority of the audience here. I will share my conclusions in a couple of slides, but after reading this article, I'm not sure if this is an intervention that I would utilize in TBI patients. So moving on to my second polling question. If you are currently working with patients who experience traumatic brain injuries, what's your institution's practice surrounding hypertonic solutions? Do you bolus hypertonic solutions when intracranial pressure is elevated? Do you utilize continuous infusion hypertonic solutions when intracranial pressure is elevated? Or do you use prophylactic hypertonic saline for moderate slash severe TBIs at your institution? I'm curious. Okay, so the audience majorly answered the first option, you bolus hypertonic solutions when intracranial pressure is elevated. 30% of the audience does utilize continuous infusion hypertonic solutions when intracranial pressure is elevated, and about 5% of the audience is currently using the approach from the study to manage TBI patients. Thanks for sharing your institution's practice where I'm at. We're doing bolus hypertonics for elevated ICP as well. So just to wrap up here, in conclusion, I agreed with the author's conclusion that there were no significant impacts on the Glasgow outcomes score, therefore functional outcomes in patients with moderate to severe TBI that were given this 20% hypertonic saline solution compared to current standards of practice. So personally, after reading this trial, I would not change my practice and start utilizing 20% hypertonic saline for patients without confirmed intracranial hypertension. And with that, I would love to open it up for questions. Thank you so much for your time, and I appreciate you listening. Thanks, Anastasia. So one of the questions we had coming in here is, can we retire doing any more studies on using hypertonic saline infusions in TBI? So as you said, can we retire, or? Right. I'm assuming they meant, I'm assuming they mean, can we stop doing the trial? Yeah, so it kind of, yeah, when I did my background research on this, it kind of reminded me of the like vitamins and sepsis thing. It sounds like there's been a lot of studies on this showing different things. This was a very pragmatically designed trial. It was prospective. I'm not impressed. I think it's awesome that they published negative results, but I agree. I don't know if there's really any evidence that this is helping people. So thanks for your question. So I have a question for Anastasia. Can you maybe speak to, you know, you called out the chloride and that they didn't necessarily report out on that. I guess, is there any steps that we could take to potentially minimize the amount of chloride these patients are receiving when they're getting hypertonic saline infusion? Yeah, I personally, I haven't had this happen to me in my very brief experience in practice, but I am aware that you can do hypertonic solutions with sodium acetate as well. So that might help reduce chloride, but then you, I think that's a strategy that I've seen utilized before in my review of the literature. I'd be curious what other institutions who work more with TBI patients also do with that as well. Yeah, that's definitely an option that comes up quite a bit, you know, on the message boards of, you know, can we do it, do other centers do it? But you do have to worry about shortages, obviously, with all the services that we're seeing currently with the pandemic. So sodium phos is a shortage item, but then sodium acetate is as well. So you kind of have to worry about that. Yeah. And it looks like we have another question here. Can you comment on the choice of fluid that they use? So the 20% versus, you know, any of the other hypertonic solutions, maybe why they pick that one? Yeah, so I also thought that was an interesting choice to use the 20% as well. I think it's because the studies published previous to this utilized 20%, so they were using the methods utilized in these smaller studies showing improved benefit and tried to match that. I have that actually, as one of the references here as well. And I don't think I have control of the screen anymore. But that was the trials that this study is based off of use 20% hypertonic saline. So they did their best to match that. I think it was an interesting percentage choice, though, based on my experience. Yeah, definitely. So it looks like about all the questions that we have. Thank you for your participation, Anastasia. Yeah, absolutely. Thank you, guys. Okay, so before we move on to our next presenter, we'd like to ask a brief polling question regarding today's attendance in order to gain a better understanding of our overall attendance to ensure continued support on the Spotlight on Pharmacy webcast. So how many attendees are viewing this webinar with? So is it just you by yourself, is it two to five people, five to ten people, or more than ten people? Okay, with that, now I'd like to introduce our second presenter, Taryn Ng. Hello, everyone. Thank you for that introduction, Dr. Gurka. As you mentioned, my name is Taryn, and I am a PGY2 at Lucille Packard Children's Hospital at Stanford, and today I'll be presenting on the treatment of multisystem inflammatory syndrome in children, also known as MIS-C. What are we missing? I apologize for the pun. So for a little bit of background, MIS-C is a hyperinflammatory response that is presumed post-infectious complication of COVID-19. It has prominent cardiovascular presentation that's similar to Kawasaki's disease and cytokine storm. If you're not familiar, Kawasaki's disease is a febrile illness that can occur in childhood, involving inflammation of blood vessels that results in coronary artery aneurysms, and also has an unknown etiology for the most part. There are several reasons why it's pretty difficult to manage MIS-C, one of the first reasons being that there are variable definitions and criteria that define MIS-C. Both the World Health Organization, CDC, the Royal College of Pediatrics and Child Health all have slightly varying definitions of what constitutes MIS-C, mostly because the data for the case definitions of MIS-C is limited, and the clinical description is only represented in those case theories. There's also the clinical symptoms of MIS-C are seen in several other types of disease states, including different infections, malignancies, and other rheumatic diseases that can occur in childhood. So oftentimes, it may end up being a diagnosis of exclusion, depending on what the child is presenting with. And the suggested treatment strategies have really relied on extrapolation from other inflammatory and rheumatoid diseases with similar clinical presentations. So for example, since MIS-C does follow after Kawasaki's disease in terms of presentation, the initial treatment in Kawasaki's is IV immunoglobulin, so that's why we use IVIG in MIS-C patients. And because of that, there really is limited evidence in terms of what the most effective and safe treatment is for MIS-C patients. And the myocardial dysfunction that comes with this presentation is a major source of morbidity and mortality. So there's an urgent need to provide guidance to healthcare providers in terms of evaluating and managing MIS-C patients. In terms of previous literature, there was a study published in December of 2020 by Belhalger et al. It was a single-center retrospective review done in France. And they looked at their own institution's practice of previously using IVIG alone as the initial therapy in MIS-C patients, compared to when they switched to IVIG and medical partners alone. As for science, their primary outcome was the incidence of recovery, defined as less ventricular ejection fraction, greater than 55%, and isovolumic relaxation time greater than 90 minutes. They found that IVIG and steroids, compared to IVIG alone, had reduced time to recovery in both of these aspects. Another study published by Oldali et al. in February of 2021 was a retrospective cohort study also done in France. And they also looked at IVIG and medical partners alone versus IVIG alone. Their primary outcome was treatment failure, defined as persistent fever two days after initial therapy or occurrence of fever within seven days. For their results, they found that IVIG and steroids had a lower risk of treatment failure, and they also had a significantly lower risk of use of adjunct therapy, hemodynamic support, acute left ventricular dysfunction, and pediatric ICU length of stay. So they concluded that IVIG and steroids had a more favorable fever course and cardiovascular outcome. Okay. And that leads us to the study that we'll be discussing today, which is the Multisystem Inflammatory Syndrome in Children, Initial Therapy and Outcomes. It was published in July of 2021 in the New England Journal of Medicine. And the hypothesis for this study was that IVIG and steroids compared to IVIG alone in the initial treatment of MIS-C is associated with improved cardiovascular outcome. For this study, they used the following six criteria as their definition for MIS-C. So serious illness leading to hospitalization, age less than 21, fever, lab evidence of inflammation, multisystem organ involvement, and confirmed COVID-19 infection. This was a retrospective multicenter disease surveillance study that was done in the United States from March to October of 2020. They used propensity score matched groups in an inverse probability-weighted analysis with adjustments to confounding factors, and they reported their results as risk ratios and 95% confidence intervals. They included patients who were less than 21 years of age and diagnosed with MIS-C and who had received either IVIG alone or IVIG and steroids as their initial treatment. They excluded patients who had received a biologic as initial treatment or had any underlying cardiac condition. Their primary outcome was a composite for cardiovascular dysfunction, composing of less ventricular dysfunction or any shock requiring vasopressor use. And their secondary outcome was the individual components of the primary outcome. In addition to receipt of adjunctive treatment and persistent or recurrent fever on or after day two of treatment. This is their inclusion flow diagram. So they started off with 518 patients who met MIS-C criteria. They then divided them into four different groups based on the initial treatment that they had received. So the first being IVIG only, the second being dual therapy with IVIG and steroids, the third being triple therapy with IVIG, steroids, and a biologic. And then the fourth category was just any other combination of treatment on day one. They then narrowed it down to patients who had only received IVIG or IVIG and steroids. So 169 patients were excluded. With the patients they had left, they did propensity score matching based on demographics including age, race, sex, any pre-existing conditions, inflammatory lab markers, vasopressor use, cardiovascular, respiratory involvement, and mechanical ventilation. This left them with 103 patients in each group. In terms of baseline characteristics, since they were propensity matched, it was pretty comparable between both groups. The majority of patients were male in the IVIG and steroid groups. The median age was 8.8 years versus 7.6 years in the IVIG alone group. Most of the patients were black or Hispanic and previously healthy. Most had been admitted to the ICU and over 40% in each group did require vasopressors. And what the study found was unsurprisingly, patients with more severe illness did receive more immunomodulatory treatments. So looking at panel A here, we can see on the X axis is the month of admission and the number of patients. On the Y axis is the percentage. So in the month of August and September specifically when we know we were having some COVID spikes within the U.S., we can see that more patients were receiving either triple therapy with IVIG, steroids, and a biologic or even dual therapy as opposed to just IVIG alone. And then in panel B, we can also see that patients who required more immunomodulatory treatments up front did have higher markers for severity of illness including ICU admission, vasopressors, left ventricular dysfunction, mechanical ventilation, and coronary artery aneurysm. For the primary outcome, the study found that IVIG and steroids was associated with a lower risk of cardiovascular dysfunction. So looking at the fourth plot here on the first line, our cardiovascular dysfunction had a risk ratio that was less than one and did not cross one mark, meaning that IVIG and steroids had a lower risk of cardiovascular dysfunction, almost half as likely than the IVIG alone group. Similarly, for secondary outcomes, the IVIG and steroid group was less likely to have shock resulting in vasopressor use or require any adjunctive treatment. They did not find a significant difference in terms of incidence of persistent or recurrent fever. And the inverse probability-weighted analysis just confirmed the above results. They also looked at the median ICU length of stay and found no significant difference between the groups. It was two days in the IVIG and steroid group versus three days in the IVIG alone group. So the authors for the study concluded the initial treatment for MIS-C with IVIG and steroids was associated with a lower risk of short-term outcomes including newer persistent cardiovascular dysfunction compared to the initial treatment of IVIG alone. This leads us to the critique of this study. So some of the strengths of it were that it was a multi-center. It had a pretty robust study design with objective clinically relevant outcomes. There was a rigorous adjustment made for baseline disease severity via the potency score matching. And based on the Odaly et al. study where they predicted a 50% treatment failure in the IVIG alone group versus a 20% treatment failure in the IVIG and steroid group, this study did meet. I should have put 90% power on that. They only needed 86 patients in total to be enrolled to make that power. And they even were able to do the potency score matching and they took out 150 patients just to do that and still met power. So I think that's a strength. It's also the first study done within the U.S. which is another large epicenter for COVID cases just like France was. So it was kind of the first done of its kind in this country. As for some of the limitations, it might be difficult to determine the effect of initial therapy versus adjunctive treatment mostly because a lot of these patients in the IVIG alone group, about 50% or over 50% of them did require secondary treatment with adjunctive steroids. However, this would bias the results more towards the IVIG alone group and we did still see a significant difference for the IVIG and steroid group. Additionally, there is a well-known side effect for IVIG which is fever and it has been reported to occur up to 36 hours after a dose. And about 30% in each group did require a second dose of IVIG. So the secondary outcome of persistent or recurrent fever might be skewed by the fact that these patients were getting more adjunctive treatment. There may also be a variable disease time course in these patients. They defined day zero in this study as the first day of treatment, not necessarily the first day that patients had received or patients presented with symptoms. There was no data for any of the vasopressor duration or dose, no data for any treatment related adverse events or long-term outcomes. Additionally, it's unclear what the best dosing strategies are for these patients as well. The majority of patients did receive two grams per kilo of IVIG as their first dose, but the second dose ranged from one to two grams per kilo. And for steroids, methylprednisolone was the most commonly prescribed at two mgs per kilo per day. There were about 20% of the patients who did receive pulse dose steroids at 10 to 30 mgs per kilo per day. And then finally, there was the benefits for other treatment regimens couldn't really be analyzed. For example, biologics, there were 34 patients who had received biologics as their initial treatment. And then up to 23% of patients did receive a biologic at some point throughout their treatment time course. For clinical application in patients with concern for MIS-C, it may be appropriate to use IVIG and steroids for initial therapy. As we saw, the more sick that these patients present with, they will need more immunological immunomodulatory therapies up front. The number needed to treat in this study was eight. So the benefits may outweigh the risk for a lot of these patients, especially if they are needing steroids for their initial treatment. But there are still further studies that are needed to determine optimal dosing strategies of IVIG and steroids, as well as long-term outcomes for MIS-C. That leads me to our first polling question, which is, what is the dose of IVIG that your institution uses for your inflammatory syndromes? And is there a maximum single dose? Thank you. So it looks like the majority will use two mgs per kilo, and most do have a max dose. At our institution, we also tend to use, oh, it should be two grams per kilo, I'm sorry. And we do have a max single dose of 100 milligrams, or grams. And that leads me to our second question. Oh, so it looks like most people use ideal body weight. We actually start with using actual body weight, and then for patients who have a BMI that's in the 95th percentile or higher, we'll use ideal body weight. Thank you for your answers. Those are my references, and I will be happy to take any questions. Okay, thanks, Karen. First question you have here is, based on the evidence to this point in COVID-induced MS-MIS-C, would it be ethical to conduct a corticosteroid trial of corticosteroid alone versus IVIG plus steroids? And sort of a secondary question or thought is, is it confirmed that IVIG is better than placebo in this disease state? Because of the severity of how some of these patients are presenting, I don't think any prospective trials looking at one treatment or the other would be that ethical, especially if they are requiring more immunological therapies upfront. And I have not seen studies looking at any treatment, initial treatment versus placebo, if that was the question. Okay. And I know this trial specifically, was more focused on IVIG versus IVIG and steroids, but I guess, can you speak a little bit to who you think would qualify for using the biologic? Yeah, so patients who probably have really poor cardiovascular presentation, so really poor ejection fraction, and who have maybe shock and a lot of end organ dysfunction, may require more intensive therapy. So a biologic might be used more upfront. At our institution, our first line tends to be anakinra. And if patients still are refractory to that, then we'll possibly add on teslazumab as well. But it would be definitely the more sick patients who have end organ dysfunction or something like that. Okay. And you would pick, I guess, a biologic, as like the third line agent, instead of maybe a repeat dose of steroid, or if somebody wanted to repeat a steroid dose, I guess, what would be your dose suggestion? Yeah, so we start up front with methylprednisolone at two migs per kilo, divided Q12 hours. And then if we need to intensify, we'll do like pulses with 30 milligrams per kilo for up to three days, I believe. And then if they need it after that, we can add on a biologic. Okay, great. It looks like those are all the questions that we have here. So that will conclude our Q&A session for Taryn. Thank you very much for presenting. And then now I'd like to introduce our final presenter, Elena Eder. Thank you for that introduction. Today, I will be reviewing the prescribing practices of valproic acid for agitation and delirium in the intensive care unit. I do have an overview slide here, but for the sake of time, I will go ahead and jump into the background regarding this topic. So we know that delirium and agitation are estimated to impact over 70% of ICU patients. Over 70% of ICU patients. And the most recent PADIS guidelines recommend against the routine use of most pharmacologic agents for the treatment of delirium. However, antipsychotics are still frequently used for this indication, despite a lack of clinical evidence and known adverse effects. There is an urgent need for safe and effective pharmacologic options to help with the treatment of these patients. Underlying ICU delirium is an increase in some of our excitatory neurotransmitters, including glutamate, serotonin, and dopamine, as well as a decrease in some of the inhibitory neurotransmitters, including acetylcholine and GABA, and a general increase in the inflammatory markers. So given this proposed underlying mechanism, there's been an increased interest in the use of valproic acid. Due to its ability to antagonize due to its ability to antagonize NMDA receptors, decreasing glutamate activity, potentiate GABA activity, and decrease general inflammation, thereby possibly addressing multiple mechanisms of ICU delirium and agitation. Existing literature supporting valproic acid for this use are small and retrospective in nature. All of these studies include less than 50 patients with a total of only 117 patients in these studies combined. And largely these studies are limited to patients with hyperactive delirium or agitation, and those who have already failed at least one antipsychotic. So those with difficult to manage hyperactive delirium and agitation. When we look at dosing in previous studies, patients typically receive 1,250 to 1,500 milligrams daily, divided twice daily to three times daily, and patients have received a loading dose in most studies to acutely manage the agitation. And studies have shown that rates of hyperactive delirium and agitation have been decreased with valproic acid, and it also has allowed opioid and benzodiazepine sparing, so a decrease in these agents which we know may increase the risk of development of delirium. This leads us to our study we're discussing today, the prescribing practices of valproic acid for agitation and delirium in the intensive care unit, which was published in the Annals of Pharmacotherapy. And in publishing these studies, the study's authors were looking to address the characterizing current valproic acid prescribing practices to identify the benefits and risks of this therapy. This was a single-center retrospective cohort study in adult ICU patients, and it included a diverse ICU population, with patients from the medical, cardiac, and surgical ICUs all included. And they reviewed patients over an eight-month period from January to August of 2018. In designing this study, they also wanted to ensure that the patients who were included were receiving valproic acid specifically for the management of delirium and agitation. They included patients who received the intervention for at least 24 hours, and it had to be indicated specifically for the treatment of agitation and delirium per documentation in a provider or pharmacist's note. And they excluded patients with any anti-epileptic home medications or who were admitted to the neurosciences ICU in hopes of excluding patients who may have another clinical indication for valproic acid. Their primary outcome was to characterize the prescribing, and this included various aspects of dosing, as well as therapeutic drug monitoring, time to initiation, and duration of the therapy. They also had various secondary outcomes looking at efficacy. And so they would assess delirium with CAM ICU scores and agitation with RAS scores. However, the frequency of the assessment in each patient was not reported. Their secondary outcomes included delirium resolution at 48 hours, the overall duration of delirium, the recurrence of delirium, as well as adjunctive sedative medication utilization, including opioids, benzodiazepines, dexmedetomidine, propofol, and antipsychotics. For safety outcomes, they looked at the known adverse effects of valproic acid, which included thrombocytopenia, defined as platelets less than 150,000 or a decrease of over 50%, hepatotoxicity, so AST greater than three times the upper limit of normal, ALT greater than two times the upper limit of normal, a total bilirubin of greater than 2.6, or a doubling of any of these values from the baseline after starting valproic acid. Hyperammonemia was defined as a SAM ammonia level of greater than 60 or an increased 1.5 times baseline for those who had a baseline ammonia level measured. And these safety outcomes do closely mirror those in previously published studies. However, as you can see here, there were no clinical safety outcomes. These were all surrogate measures, which is what has previously been examined in previous studies. For statistical analysis, they appropriately controlled for type I error with an alpha of 0.05. The primary outcomes used descriptive statistics and secondary outcomes were analyzed using Wilcoxon-ranked sums of continuous variables. And for categorical variables, chi-squared or McNamara tests were used appropriately. Overall, 80 patients were included, which is the most that has been included in any valproic acid studies to date. 35 of these patients received valproic acid as monotherapy, but we can see that the majority of these patients were also on an adjunctive antipsychotic. Looking at the baseline characteristics for both of these arms combined, we can see that these patients were elderly, predominantly male, and admitted to the medical or cardiac, vascular and thoracic surgery ICUs. The majority of these patients were mechanically ventilated. And we can see that about a third of the overall population had a previous psychiatric diagnosis or substance use history. At baseline, these patients were receiving various other adjunctive sedative medications. So about half of all of these patients were receiving an opioid or propofol prior to the initiation of valproic acid. And the minority of patients were receiving dexmedetomidine or a benzodiazepine. And as previously discussed, over half of these patients were already on an antipsychotic medication. Looking at the timeline for the development of delirium and initiation of valproic acid, we can see that after ICU admission, patients developed delirium a median of four days after admission to the ICU. And valproic acid was initiated three days later on ICU day seven. Ultimately, valproic acid was discontinued on ICU day 13. So the median duration of therapy with valproic acid was six days. Looking at some of the prescribing characteristics, the most common dose of valproic acid was 250 milligrams three times daily. The second most common prescribing regimen was 500 milligrams three times daily. The majority of these patients received the medication enterally. And interestingly, there were no loading doses administered, which is in contrast to previously published studies. And the majority of these patients had no dose titration or taper. So whatever dose they were initiated on was continued until discontinuation of the medication. Looking at the incidence of delirium and agitation, we can see that on the start of valproic acid on day zero, 93% of these patients were experiencing delirium, which decreased to 68% by day four. However, when we look at the red line representing agitation, on day zero, 64% of patients were agitated, and this was unchanged by day four of valproic acid therapy. And of note, the presence of agitation was defined as any positive RAF score. So this does not represent the severity of any agitation, just whether or not agitation was present. So we can see that while there was a significant improvement in delirium, there was not resolution of agitation in these patients. For a secondary outcome, delirium resolution, overall, 55% of patients did experience delirium resolution, and there was a significantly higher rate of resolution in the patients who received valproic acid monotherapy as compared to those who were receiving an adjunctive antipsychotic. However, interestingly, this did not persist over their time in the ICU. So we can see that delirium recurred in 34% of these patients, and this was similar across both arms. And interestingly, 40% of these patients who experienced delirium recurrence were still receiving valproic acid therapy. So though there was a reduction in the incidence of delirium by 48 hours, this was not sustained as the therapy was continued. Looking at the use of adjunctive sedative agents, we can see similar to previously published studies, the percentage of patients on each of these agents decreased following the initiation of valproic acid therapy. So notably, dexmedetomidine, we can see, also decreased by valproic acid day three. So though we discussed the fact that agitation was still present in these patients previously, dexmedetomidine use, which we know we can use for agitation and hyperactive delirium, actually decreased, just like the other sedative agents, as opposed to being increased to try to address this agitation. And then looking at some of the safety outcomes, overall, the rates of adverse effects were low. They did monitor drug levels, and they split these levels into those that were above and below 50 micrograms per milliliter, and these were not correlated with increased risk of adverse drug reactions. But only 24 patients had therapeutic drug monitoring conducted throughout the study. We can see that the most common adverse reactions were hepatotoxicity and thrombocytopenia, both occurring in less than 10% of patients. Notably, we know that these adverse effects can occur in ICU patients who are not receiving valproic acid therapy, and so due to the retrospective nature of the study, it's difficult to conclude if this was secondary to the valproic acid therapy. But 6% of patients were reported as discontinuing the medication due to an adverse drug reaction that providers believed was due to valproic acid. Ultimately, the authors concluded that valproic acid is currently utilized in the management of agitated delirious ICU patients, and it may improve delirium resolution. So we saw that it did not impact agitation in this study. And valproic acid may decrease the utilization of other adjunctive sedative agents. Revealing some of the strengths of this study, there is detailed characterization of the effects of valproic acid In this study, there is detailed characterization of the prescribing of valproic acid, which can help guide future providers who are looking to use this medication for the management of ICU delirium or agitation. They contrasted two different arms, those who were receiving antipsychotic adjunctive therapy or valproic acid monotherapy. Previous studies have largely only included patients who were already on an antipsychotic agent. So it's interesting to investigate this medication as monotherapy. The design was limited to those who truly were receiving valproic acid for delirium. So we're able to conclude that these patients were delirious and were receiving the medication for the indication at hand. And the collection of adjunctive sedative agents helps to paint a picture of how sick these patients were, what other agents they're receiving, and at what point in therapy valproic acid was considered for the management of hyperactive delirium. For limitations, as far as generalizability, the majority of these patients were experiencing hyperactive delirium. They were elderly and admitted to the medical or CV ICU. So we can't draw conclusions about patients who may be experiencing different types of delirium, or particularly one ICU population that was missing was the trauma population was not studied. Adverse effects were measured by surrogate markers rather than clinical markers. And we know that the thrombocytopenia and hepatotoxicity that was observed may have been to reason that side of valproic acid therapy. So it's still difficult to say how safe this medication is. They could not control for drug interactions due to the retrospective nature. And 10% of these patients were on an adjunctive carbapenem, which we know significantly decreases valproic acid concentration. So it's difficult with a retrospective study to control for interactions with other interventions. Delirium resolution could have been confounded by other sedatives. So we saw that opioids and benzodiazepine use both decreased as the study went along. And resolution of delirium could be attributed to decreasing the utilization of these agents rather than the addition of valproic acid. Additionally, there was no comments on duration of mechanical ventilation. And we know that extubation can also remove a risk factor for delirium. So again, the retrospective nature of this study, makes it difficult to conclude that valproic acid was the cause of delirium resolution. And finally, the definition of agitation only allows us to draw conclusions about the presence of agitation and not whether the severity of agitation was improved. So it's possible that for patients who are severely agitated, valproic acid may still be a useful agent, but we only have whether or not the RAF score was positive. So we only know that agitation was not completely resolved based on the results of this study. So in conclusion, valproic acid may be useful on the management of hyperactive delirium. However, this effect may not be sustained. So future studies may work to identify which patients experienced the sustained effect so that possibly we can be more specific about who we initiate this medication in. There's limited safety data that could restrict valproic acid to use as a second or third line therapy. So we have more robust data surrounding agents like dexmedetomidine or antipsychotics than we do surrounding valproic acid. So people may be hesitant to use this agent in practice when we're not sure what the long-term outcome or impact could be. And of notes, there was about 20% of these patients were continued on valproic acid at discharge from the hospital. There are low rates of dexmedetomidine utilization that may not reflect current practices. So in general practice, it may be more common to titrate up dexmedetomidine while the patient is still in the ICU rather than starting a new agent. And then standardization of delirium management may guide optimization of other therapies prior to valproic acid initiation. So while we know that patients may have been on an antipsychotic, we're not sure if this dose was optimized. This institution did not have a delirium management protocol, but if one had that in place, say it may have guided providers to increase the dose of something like say, clatiopine from 25 milligrams to say 100 milligrams before adding a new adjunctive agent to try to manage ICU delirium. So standardization of delirium management and institutions may ensure that the therapy that's already in place is optimized before adding another agent such as valproic acid. This moves me into my first polling question. What is the current role of valproic acid in the management of delirium? At your institution, is it used for hyperactive delirium, hypoactive, mixed delirium, delirium prevention by weaning other sedative agents or none of the above? Valproic acid is not used. So it looks like 58% of patients say that they utilize it for hyperactive or 58% of responders said that it's used for hyperactive delirium. And the next most common answer is E, none of the above. So it's either used for hyperactive delirium or not at all, it seems. In my experience, the patients who received valproic acid did experience hyperactive delirium. I would agree that in hypoactive delirium rather than adding another agent, at that time we might consider taking agents away. So I would say that would fit what I have seen at Grady as well. Okay. And then on to our next polling question. When are you most likely to recommend valproic acid for ICU delirium? Would it ever be as a first-line agent, a second-line agent after failure of antipsychotics, a third-line agent after failure of antipsychotics, clonidine, or other agents, or you would not recommend valproic acid for the management of ICU delirium? Okay. So it looks like the majority said they would utilize it as a third-line agent. And then it looks like the second most common answer is as a second-line agent after failure of antipsychotics. So I would say that this reflects most of the currently published literature. It seems like valproic acid is very rarely used as a first-line agent. We wait until a patient has failed some of the other agents that we have more data about. But I do think that this question also depends on the presentation of the patient. So those patients who did have a previous psychiatric disorder or substance use disorder, about a third of the patients in this study did receive valproic acid sooner. And so I think that based on this study, it is something kind of food for thought that in those patients with a psychiatric disorder, should we be considering valproic acid earlier? Something for future studies to look at. And that concludes my presentation. I can take any questions from the audience at this time. Great, thanks, Alina. So your first question is, specifically with hyperactive delirium, would you suggest monitoring your levels for that indication and then if so, what level would you aim for in these patients? So that's a great question. As far as the need to monitor levels, I think that for this indication, I would specifically be titrating to a clinical response. So the resolution of delirium or agitation rather than a therapeutic level. I think the only time that I may get a therapeutic level is if a patient maybe had a benefit from valproic acid and experienced some kind of possible adverse drug reaction. We could see if it was secondary to valproic acid or possibly just due to their critical illness, for example, thrombocytopenia. But as far as what levels to aim for, the majority of patients in the currently published literature had levels less than 50 milligrams per deciliter and did experience resolution of delirium. So I don't really see any reason to push levels any higher than that. I think at that point, you may need to consider alternative agents rather than valproic acid. Okay, great. Next question is, it's relating to the dosing that they use. So you highlighted that they didn't necessarily have a set protocol, but I guess, can you speak to maybe why they use such low doses as compared to some of the previously studied or even some of the studies using loading doses? Yes, it is interesting that there was a lower rate of loading doses, particularly given that they don't have a protocol or policy. I'm not sure why it wasn't commented on, but it's possible that in these patients, they didn't see the need for a loading dose since we weren't targeting a certain therapeutic drug monitoring at a certain level. But I do think that we can see in this study that they had similar rates of delirium resolution without the loading dose. Perhaps the loading dose is more useful in the acute management of a severely agitated patient. But they did not specifically comment on why the dosing may be different. I will say that it seems as if the lower dose of 250 milligrams three times daily as opposed to 500 milligrams three times daily being much more common in some of the other studies has shown similar rates of delirium resolution. So it seems like by pushing the dose up, we may not get the effects that we want that may place patients at higher risk for adverse effects. So it's possible that they were trying to be conservative. They were trying to be conservative. But I do think it's interesting, they also did not titrate the dose. So they started at this lower dose. But it seems as if providers rather than titrating up are more likely to just discontinue the medication if it was not effective. Okay, great. Just about out of time. So thanks for your presentation, Lena. Yes, thank you. Okay, well, thank you to our presenters today and everyone for joining. Please join us on the third Friday of the month from two to 3 p.m. Eastern Standard Time for the next Journal Club Spotlight on Pharmacy. That concludes our presentation for today.

critical care medicine

continuous infusion hypertonic saline

neurological outcomes

traumatic brain injury

multisystem inflammatory syndrome in children

IVIG

steroids

valproic acid

agitation

delirium