Good afternoon and welcome to today's Journal Club, Spotlight on Pharmacy webcast, which is supported by the Society of Clinical Care Medicine CPP section. My name is Leslie Jurado, Surgical Trauma and Neuro ICU, Clinical Pharmacist at New Hanover Regional Medical Center in Wilmington, North Carolina. 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 will also have the opportunity to participate in several interactive polls. When you see a poll, simply click the bubble next to your choice. 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 during 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, assistance, product, procedure, opinion or other information that may be mentioned. And now I'd like to introduce your speakers for today. Each will give a 15-minute presentation followed by Q&A. Our first presenter today is Makayla Huffer, PGY-2 Critical Care Resident at Barnes-Jewish Hospital in St. Louis, Missouri. Our second presenter is Elaine Chang, PGY-2 Critical Care Resident at Oregon Health and Science University Health in Portland, Oregon. And our third presenter is Justina Gerges, PGY-2 Critical Care Resident at Beth Israel Zichnes Medical Center in Boston, Massachusetts. And now I'll turn things over to our first speaker. Good afternoon and thank you for the opportunity to present to you all today. The article I will be looking at discusses the development and validation of a medication regimen complexity scoring tool for critically ill patients. In 2004, a screening tool called the Medication Regimen Complexity Index, or MRCI, was developed to identify patients in the outpatient setting that may benefit from pharmacist intervention such as medication therapy management. However, as I mentioned, this tool was created for and used in the outpatient setting. When we look at the ICU setting, scores such as the APACHE score have been validated for physicians and nurses to help determine appropriate patient ratios, but this is not true for pharmacists. In fact, no such tool has been validated in the ICU setting to identify which patients may benefit from further pharmacist intervention. We know that pharmacists are justified based on quality improvement measures, improved safety measures, and cost effectiveness measures. However, the exact ratio and how to allocate these services is unknown. When it comes to pharmacist allocation, ratios can vary significantly between institutions based on which method is used for justification. The gap in critical care pharmacy practice that this study identified is a need for a tool to help objectively assess medication regimens to therefore guide justification to further ICU pharmacist positions and allocation of pharmacist services. I'd like to now begin with a poll for the audience to get a better feel for what institution's typical approach is to creation of pharmacists to patient ratios in their settings. What does your institution primarily use to justify pharmacists to patient ratios in the ICU? Is it number of orders verified, number of rounding teams, the number of patient beds, the intervention activity, or a medication complexity scoring tool? Okay. It looks like we do have kind of a nice spread here with the majority being number of patient beds. My institution typically does number of rounding teams, which looks like probably the most common, and then intervention activity and medication complexity scoring tool, the least common. As I mentioned to meet the gap previously, the objectives of this study were to develop such a tool and validate it internally, then secondarily to attempt initial external validation and examine if this tool is associated with important clinical outcomes. It was a single center retrospective observational chart review that examined 130 medical ICU patients at Augusta University Medical Center from November 2016 through June 2017. Patients were excluded if they were in the ICU for less than 24 hours, if they were in hospice care, or if they have active transfer orders by 24 hours. Once these patients were identified, their medication regimens were evaluated starting on their first full day after admission. As I walk us through the methods, I will briefly mention the statistical test used. Unfortunately, due to time constraints, I will not be able to go into depth on the statistics, but in my assessment, all tests used were appropriate. The development of the MRC ICU tool utilized similar methods as the development of the previous MRCI tool. Initially, a scoring tool was developed that included 71 items based on literature search and the clinical knowledge of three pharmacists. After this tool was developed, it was applied to all 130 medication regimens by one investigator in a randomly selected subset, rescored by the other two investigators. At this point, statistical analysis was performed to ensure internal consistency using the Chromebox alpha. Changes were then made to the tool, including content and weighting of items. This new tool was then used to rescore 50 of the 130 regimens by one investigator and a randomly selected subset was once again rescored by the other two investigators to determine internal consistency and inter-rater reliability using an inter-class correlation coefficient, or ICC. Then all 130 regimens were rescored by one investigator once again to determine convergent validity using the Pearson product moment and discriminant validity using the two-sample t-test and the Pearson product moment. Two additional pharmacists not previously involved scored 25 regimens twice with a two-week gap to determine the inter-rater and test-retest reliability once again using ICC. For internal validity, six pharmacists from the institution rated six regimens based on complexity and then the MRC-ICU tool was provided for feedback and agreement was assessed with Kendall's coefficient of concordance. For external validity, five critical care pharmacists from outside the institution ranked those same six regimens, then the tool was provided for feedback and consensus assessed using Spearman's rank correlation. For clinical outcome correlations, they used the two-sample t-test and Pearson product moment. The endpoints chosen match the study objectives quite closely, as you will notice. In the second objective, the clinical endpoints chosen to be evaluated included patient outcomes, length of stay in the ICU, patient mortality, and patient acuity scores. In general, these patients match our general ICU population well, with an average age of about 60 years old, a little over-majority female, the average length of stay in the ICU of six days, an Apache 3 score of about 68, and the mean number of medications at 24 hours of about 17. As you can see, the MRC ICU demonstrated a high degree of validity between different users and multiple testing. Additionally, it was significantly associated with number of medications and orders, as one would expect a medication complexity regimen to be. Even further, it was not significantly associated with age or sex, demonstrating a high discriminant validity of the tool. Now that I've shown you the development of the tool, you're probably all wondering, well, what does the tool actually look like? It includes seven different categories, including high-priority medications, ICU medications, TPN, prophylaxis, analgesia and sedation, antimicrobial agents, and devices. Within these seven categories, there are multiple items, and each item is carried a weight from 1 to 3, with 3 representing a higher degree of complexity, down to 1 indicating less. Due to time constraints, I did not list all 39 items. However, if anyone would like more information, I would refer them to Table 2 in the study. The authors not only wanted to create and validate a tool, but also to look to see if this tool was at all related to clinical outcomes. They found the MRC-ICU tool was correlated with length of stay, mortality, and patient acuity. Additionally, it was not related at all to number of home medications. So, in my opinion, these patient outcomes are not only statistically significant, but clinically significant as well. The authors therefore concluded that the MRC-ICU tool is a validated tool that is associated with important clinical endpoints, as mentioned on the previous slide. They do agree, though, that further validation of this tool is necessary. So, when we look at the strengths of the study, I think that it attempts to answer an important gap in pharmacy practice. Additionally, in my assessment, the authors used appropriate statistics to create and validate their tool. When looking at application, the population studied was a general MICU population, and patient characteristics matched what one would expect, thereby increasing external validity. Additionally, the authors also took it to the next step to examine if this tool is correlated with important clinical endpoints, things that we really care about as pharmacists. However, like every study, there are significant limitations as well. It was a retrospective study, and perhaps a prospective study could have examined more fine-tuned patient outcome data, things like adverse drug events or costs. Additionally, when looking at the tool itself, there are several questions arise. One such being, does the antibiotic indication matter? Prophylaxis versus treatment, as this is not addressed in the tool, and in my opinion, could make a clinical difference. Additionally, the tool scores medications based on restricted antibiotics, and we know that these can vary significantly based on institution. While the tool does address dialysis, it does not address modality, and so it's intermittent hemodialysis equivalent to continuous renal replacement therapy when it comes to medication complexity. The tool does not specifically address the definition of medication orders, and one could ask if something like a saline flush is truly a medication order, and if it would carry the same weight as other medication orders. Finally, it is a static tool that looked at one time point, 24 hours after admission. We know our critically ill patients are anything but static. Also, while the study looked at the general MICU population, it was single site, and very limited baseline characteristics were reported, making further generalizability challenging. Along these lines, this study does not address other critically ill populations, such as surgical or cardio. A major limitation of this study is the external validation. Kudos to the authors for attempting. However, in my opinion, I would not say that five critical care pharmacists from an outside institution is robust enough to count as true external validation. Because of the study design, there's also a high risk for bias. And then finally, this study gives no guidance on how such a tool should or even could be implemented in the clinical setting. So, I believe the MRC ICU tool attempts to answer a novel question, how to justify and allocate critical care pharmacist positions. Additionally, I agree with the authors and that their tool is promising as it showed a high degree of validity and is associated with important clinical outcomes. However, identifying its role in practice and application may be difficult. One challenge is due to different pharmacy practice models, as these can differ immensely throughout the country. And such a tool as the MRC ICU may lend itself more beneficial to certain models over others. For example, at my institution, which is a clinical specialist model that justifies pharmacists based on number of rounding teams, I see this tool as less helpful. Score threshold would also need to be investigated as it would be hard to interpret acuity with this score without associating that score with an outcome such as mortality. So with this, at what score does mortality increase and at what score does it not matter? In my view, I see this tool as potentially useful for pharmacist time allocation over position justification due to its limitations and challenges in application. However, this is model specific as well. Additionally, it is not a simple tool to apply and therefore would need to be built into an EHR or some sort of computer technology as this tool is complex in itself and not something that could be done by the bedside. Finally, I do think it needs further external validation as this is a major limitation of the current study. So while I think it is an intriguing tool, there are significant limitations to transitioning it into clinical practice. I have one more polling question then for the audience. How helpful would a medication complexity scoring tool like the MRC ICU be at your institution? Helpful with ICU pharmacist justification? Helpful with allocation? Helpful with both justification and allocation? Not helpful or not sure? Okay, interesting. We have a nice spread here as well. It looks like most people see that there could be some benefit in both justification and allocation. I kind of like to see that, is that it was an intriguing tool and maybe there'll be more that come from it. It looks like the minority is just helpful with justification. I want to thank you all for having me. Before I take any questions, I just want to thank my mentor, Dr. Adrian Wong, for all of his help in putting this together. I'll take any questions you have now. Michaela, you mentioned that this tool would be difficult to apply to other ICUs. However, how do you think that this could be applied to, let's say, in your ICU or a surgical ICU? Is there anything that you would like to incorporate to the tool? If I get your question right, you're asking. Maybe not every ICU, but even in just like a surgical ICU, is there something that I could take from it in my practice to apply? Yes. I do think one thing I found very interesting about this is trying to very objectively determine what value we have in the sense of the complexity of medication. Looking at what's actually on the tool and where pharmacists are most beneficial. If anything, I think I could look at this tool and see what aspects of it come into my day-to-day practice. Is that where I see the focus of my work? Again, I'm looking at Table 2 now in the paper, if anyone's following along, reading through that. Things like the antibiotics, most of us know aminoglycosides are trickier dosing. That comes up in my day-to-day practice. Obviously, I spend more time on those. Things like prophylaxis obviously gets attention too. I think, if anything, it's more of a way for me to communicate what it is that I'm doing, more than for me justifying what I'm doing or allocating my time. Thank you. I have another question. How do you think this tool would compare to the PCATCH score in terms of allocating clinical pharmacy personnel? That's a good question. Actually, if you look in the discussion of this article, they do a little bit of the comparison with the PCATCH. I know one thing that they talked about being slightly different was the PCATCH score actually also incorporated a teaching component into it. I think that's interesting, but again, very institution-specific as not every institution does have necessarily as big of a teaching component. That is a part of a lot of our pharmacy practices. I think it should be accounted for in our day-to-day practice when it comes to physician allocation. Did you see any time frame or time period that they had to use this tool? For example, would you be able to do this while you're rounding or while you're verifying orders? You asked a really good question. I don't think that's been determined. They looked at it or used this tool just at that 24-hour mark. The question is, why are we using the tool? If it's for justification, obviously, that would be more of a retrospective look back, but if it's for allocation, that would be, in my opinion, a current limitation of this tool because you'd have to somehow have it built to, in real time, evaluate a patient's regimen and somehow spit you out a number that means something. That would have to, again, have a large technology component to it, either in the EHR itself or some other separate program that could take that information and very quickly give you an interpretation. I'm not sure if that's the author's original intent, but in my opinion, if that were able to be done, I think it could be useful. I just see a lot of hurdles in actually getting that to the bedside. Great. Thank you. Thank you. That concludes our Q&A session. Thank you, Michaela Hoffer. Now, I'd like to introduce our second presenter, Elaine Chang. Elaine Chang. Hi, everyone. I'm excited for the opportunity to present today. The article that I'm looking at is seeking to answer, is early chemical thromboprophylaxis in patients with solid organ injury a solid decision? Jumping right into the background, over the last decades, nonoperative management has become the standard of care for managing hemodynamically stable blunt trauma patients who sustain a solid organ injury with spleen, kidney, and liver being the most common injury patterns. For those who may not be familiar with nonoperative management, it can range from simply just observation to angioembolization of the site of bleeding with coils. Therefore, these patients still carry a risk of bleeding. Conflictingly, blunt trauma patients are also at a high risk of VTE and may benefit from early initiation of chemoprophylaxis. As with any use of chemoprophylaxis in trauma, this requires balancing the risk of recurrent bleeding and failure of nonoperative management against the risk of thrombotic events. Data in TEG-based studies have shown that trauma patients with blunt abdominal solid organ injuries transition to a hypercoagulable state around 48 hours after injury. And to date, several small retrospective studies in the last 17 years have shown safety of early administration of VTE prophylaxis within 48 to 72 hours of admission. Yet there are still no widely accepted guidelines for an optimal and safe timeframe for the initiation of VTE prophylaxis. As a result, the aim of the study was to investigate the optimal time to initiate chemical thromboprophylaxis in patients who have undergone nonoperative management of their blunt solid organ injuries. And for the rest of this journal club, I'll be abbreviating chemical thromboprophylaxis as CTP. We hypothesized that early initiation within 48 hours of admission of CTP does not increase the rates of bleeding complications or the need for operative interventions. Going into the methods, this was a two-year 2013 to 2014 retrospective analysis of patients within the ACST QIP database, which includes data from more than 800 institutions. They included trauma patients greater than or equal to 18 years old with at least one blunt abdominal solid organ injury who underwent nonoperative management. Patients were excluded if they had severe head injuries with an abbreviated injury scale of greater than or equal to three, had a spinal cord injury with an abbreviated injury scale of greater than or equal to two, transferred from another institution, burned, or had a penetrating injury. Patients were then stratified to an early CTP group of less than or equal to 48 hours, a late CTP group of greater than 48 hours, and a no CTP group. CTP was defined as a prophylactic dose of either low molecular weight heparin or unfractionated heparin. The primary outcomes included rates of failure of nonoperative management, which was defined as any surgical procedure for their solid organ injury after the initiation of CTP. Other primary outcomes included post-prophylaxis, packed red blood cell transfusion, and mortality. The secondary outcomes included the rates of VTE events, an ICU, and hospital length of stay. Continuous variables were analyzed using a Manu-Whitney test or student T-test. Categorical variables were analyzed using a Chi-square test. And finally, confounding variables were controlled using a multivariable logistic regression, and a p-value of less than 0.2 on the univariable analysis was included in the multivariable logistic regression analysis. In terms of the results, a total of 36,197 patients met inclusion criteria. 23,160 were in the no CTP group. 4,819 were in the early CTP group of less than or equal to 48 hours. And 8,208 patients were in the late CTP group of initiation at greater than 48 hours. The overall mean age was 50 years old. 68% of the patients were male and 71% white. The majority of the patients were hemodynamically stable on admission with a Glasgow Coma score of 15. And only 4.8% of the patients had combined injuries of two or more. As you can see in this table, there were no statistically significant differences between the three groups in regards to their primary outcome. However, when the authors did a subgroup analysis based on the type of solid organ injury and injury grades 1 to 3 versus greater than or equal to 4, they did find a significantly higher rate of failure of non-operative management and post-prophylaxis blood transfusions, specifically in spleen and kidney injuries, as well as in the higher injury grades of greater than or equal to 4 group. Jumping into their secondary outcomes, they found that CTP groups had lower rates of DVT and PE. And you can see that the prophylactic effects were more pronounced in the early CTP group on this table. There were no CTP group, the no CTP group had the shortest ICU length of stay, and the late CTP group had the longest ICU length of stay. When they sub-stratified the early CTP groups into 24 hours versus 24 to 48 hours, they found no difference in the primary and secondary outcomes. There was also no difference between low molecular weight and unfractionated heparin in DVT and PE prevention. In summary, the authors concluded that the administration of CTP within 48 hours in patients undergoing non-operative management of blunt abdominal solid organ injuries was associated with decreased rates of DVT and PE, with no significant difference in PAC red blood cell transfusion, failure of non-operative management, and mortality. In terms of critique of this study, despite the growing body of evidence to support early CTP after solid organ injury, many practitioners are still hesitant to initiate early pharmacologic prophylaxis after solid organ injury. And this hesitation is reflected in the disproportionate study groups in which the majority of the patients did not get any CTP. And within the 36% that did, more than half received CTP in greater than 48 hours. To my knowledge, this is the biggest study looking at the safety of CTP in blunt abdominal injuries and further adds to the current body of literature to further support the safety and efficacy of early initiation of CTP. The strengths of this study are the large sample size from a multi-center database contributing to its generalizability. It was well-designed and included a control group of no CTP as a reference group, which previous studies omitted. The inclusion-exclusion criteria were appropriately designed to minimize confounding factors influencing the decision to start CTP, such as severe traumatic brain injury and spinal injuries, as both these injury types are known predictors for failure of non-operative management as well as DVT. And finally, the primary outcomes align with the research question and were consistent with those discussed in previous literature on this topic. There are, of course, limitations to this study, the biggest one being its retrospective non-randomized unblinded nature, introducing a considerable selection bias. The patients entered respective treatment arms based on the preference of the treating physician and presumably timing was based on the severity of injury. Furthermore, I was not able to evaluate each institution's protocol for evaluating injury severity grade and decision for non-operative management versus surgical operation, as well as, given its retrospective nature, I was not able to assess the parameters each institution used for PAC red blood cell transfusion. There was also surveillance bias, as no standardized VTE screening was performed, and we are unable to determine if all symptomatic and asymptomatic VTEs were captured. Next, the retrospective nature also introduces questions of what is the optimal CTP dosing of low molecular weight heparin and infractionated heparin to prevent DVT or PE, which the study does not answer. The study also did not report what kind of non-operative management patients received, as the risk of bleeding in observation alone may potentially be different from angioembolization. I was also unable to determine the reasons for failure of non-operative management in each case, so it is unclear if the need for surgical intervention is due to bleeding, a potential side effect of enoxaparin, or for other reasons. Finally, the authors found a difference between the lower injury grades 1 to 3 group, compared with the injury severity grades of greater than or equal to 4 group, in terms of failure of non-operative management and PAC red blood cells transfused. However, attendings at our institution often find a severity grade of 3 to be the hardest injury grade to make a decision on whether or not to start CTP, so authors could have stratified the secondary outcome differently. Overall, there was a statistically and clinically significant difference in their secondary outcome of rates of VTE among groups. Although the sample size was large, no power analysis was reported. However, the rates of the primary outcome were similar among all groups. So in terms of my takeaway points for this study, it appears that CTP is effective in the prevention of DVT and PE when initiated within 48 hours in patients with blunt solid organ injury undergoing non-operative management without increasing the risk of non-operative management failure, mortality, or need for transfusion. Overall, PE and DVT are some of the most common preventable causes of hospital morbidity and mortality, and thus based on the results of this study and previous studies, I would suggest early initiation of CTP within 48 hours, especially in those with low-grade injuries, as long as the hemoglobin has stabilized with minimal change within 24 hours. With the strength of the study and its numbers, it generated some interesting sub-analyses that warrant more research. Further studies should focus on the risk and benefit of initiating CTP at different injury severity grades as well as by organ. Furthermore, the sub-analyses also found no difference in outcomes in patients started on CTP within 24 hours versus 24 to 48 hours, contrary to the expectation that the incidence of VTE would be lower in the population that received early prophylaxis. Thus, further research can give us guidance on these populations as well as the optimal timing of initiation of CTP and its efficacy and safety within 24 hours. So I would like to end with some polling. The first poll is, at your institution, when is CTP most commonly initiated in patients with blunt solid organ injuries who undergo non-operative management? Less than or equal to 24 hours, less than or equal to 48 hours, less than or equal to 72 hours, or greater than 72 hours. So it looks like we have a good spread, and most institutions institute CTP consistent with the results of this study in terms of less than or equal to 48 hours. I have two additional more questions. The next question is, at your institution, when, whoops. The next question is, in your practice, what is the most common reason why CTP is not started within 48 hours from injury? It may get discharged, higher-grade solid organ injury, fear of bleeding, or attending preference. So it looks like the most common reason is the fear of bleeding, and then attending preference and higher-grade solid organ injury come in a tie. And in terms of this study, one of the primary outcomes was PAC-RED blood cells transfused. So potentially, with the results of this study, the fear of bleeding, that concern can be supported by the results of this study. And finally, do you have an institutional protocol in place for the initial initiation of CTP within this patient population, stratified based on injury grade, yes or no? And it looks like 100% do not have a protocol in place. That is the same with our institution, and typically, the initiation of CTP within our institution is guided by attending preference. Thank you. I appreciate all your time today, and I am open to any questions. Other patients in the ED had a systolic blood pressure less than 90. Did you see anywhere in this trial where they mentioned those patients receiving MTP, and if that may have been a reason why they may either not started the CTP or delayed it? What does NTP stand for? I'm sorry, Massive Transfusion Protocol? Oh, Massive Transfusion Protocol. No, this study specifically included patients who were hemodynamically stable on admission. So, they didn't specifically comment on patients who received the Mass Transfusion Protocol, and most of the amount of PAC red blood cells reported were after CTP was initiated. So, unfortunately, based on most of the patients in this study, it's hard to make a recommendation in patients who have received the Mass Transfusion Protocol. Thank you. You also mentioned on your witnesses that they did not mention doses for chemical TTP or doses for obesity. Do you think that would have made a difference in the results of this trial? I think in terms of they didn't report any doses, potentially it could have made a difference. In terms of the original GEERT study, they compared low molecular weight heparin at 30 milligrams BID compared to heparin 5,000 milligrams BID, and based on that study, the low molecular weight heparin did better. So, it seems that if potentially patients were on unfractionated heparin at 5,000 twice daily, potentially that could have influenced the rates of VTE in this population, potentially, because based on that study, it seems like that dosing is inferior. In terms of overweight patients, I think it definitely, potentially could, definitely could contribute, especially if they had a higher population of obese patients in this study, especially since weight, you know, obesity is a contributor or a risk factor for the, for the, for VTE, and in terms of their BMI of this study, the BMI was 25, so the average BMI was 25, so there weren't many obese patients included in this study. Thank you. Then the last question is, there was no difference in rates of DVTRP in patients that received rates of DVTRP in patients that received early or late CTP. Do you agree with that, with those results? There, in terms of a difference in their study, they, in terms of early versus late, they did see a larger difference in patients who were initiated within 48 hours compared to patients who were initiated at greater than 48 hours, and so I do agree with the results. The confidence intervals for those, if you look at their Table 3 in the study, the confidence intervals did not cross, so there was a larger difference in within 48 hours. What they didn't find was if there was a difference within 24 versus 24 to 48 hours. Yes. I think this study, those results are hypothesis generating, and it would be an area to look at for future research studies as potentially a primary outcome, so we can ensure that it's adequately powered. Thank you. Uh-huh. That concludes our Q&A session. Thank you, Elaine Chang. Now I'd like to introduce our final speaker, Justyna Gerges. Thank you for the introduction. Without strong evidence to guide the management of status epilepticus, treatment strategies vary from one institution to another and has led to somewhat of a controversy regarding the optimal anti-epileptic for early termination of seizures, which is why today we'll be discussing the results of the long-awaited ESET study, which is a randomized trial of three anticonvulsant medications for the treatment of status epilepticus. Status epilepticus is a common medical emergency that affects roughly 12 to 36 per 100,000 people annually, based on the most recent epidemiological study that was published in 2019. Interestingly, some older literature actually suggests that it presents most commonly in patients without a previous history of epilepsy, but regardless, the overall associated mortality rate is 20 to 30 percent, and this is a value that's been repeatedly reported in the literature from the early 1990s till present day, indicating that we likely haven't made as many advancements in the treatment of status epilepticus. A majority of seizures are self-limited and resolve within two minutes, but as a general broad definition, status epilepticus is a single prolonged seizure or multiple seizures without a return to baseline or regaining of consciousness. And as our understanding of status epilepticus has matured, an increased emphasis has been placed in regards to early termination of seizures. Subsequently, our definition has undergone multiple changes throughout the years, in order to promote early identification and treatment. And while the Neurocritical Care Society utilizes a five-minute cutoff of continuous seizing, the American Epilepsy Society guidelines use a 30-minute definition. However, I think it's still important to keep in mind that even though their definitions might vary, both guidelines still recommend initiating treatment within five minutes of seizures. It's also important to recognize that status epilepticus has two distinct pathophysiological processes. There's the triggering event and then the failure of our endogenous mechanisms to terminate a seizure leading to its prolongation. The etiology of a trigger event is largely nonspecific. It doesn't differ from that of a seizure that lasts two minutes or less. While there are multiple biological processes, alterations in our inhibitory neurotransmitter GABA and its GABA-A receptor leading to an inability to hyperpolarize neurons, that seems to be the most widely accepted theory today. Regardless, a prolonged seizure lasting 30 minutes or longer has been shown to cause neuronal death, significant morbidity and mortality, and an increased likelihood of seizure prolongation. So, I think it's important to keep in mind that even though neuronal death, significant morbidity and mortality, and pharmacoresistance to our first-line agents, making early termination crucial. Both guidelines recommend a stepwise approach to the treatment of status epilepticus and they agree that benzos should absolutely be the first-line agent for emergent anti-epileptic therapy. However, there's really no consensus as to which second-line anti-epileptic to use next. The American Epilepsy Society recommends using phosphonatone, levotiracetam, or valproic acid, which I feel is a more accurate representation of what's used in current practice, and that's really based on international surveys that were published in neurocritical care. Currently, American Epilepsy Society guidelines have no preference between the three second-line anti-epileptics on the basis of safety or efficacy, but they state that the results of the ESET study were anticipated to kind of give a little bit more direction. Here we have the properties of the more commonly used anti-epileptics for the management of status. Phenytoin was historically favored, however, it's largely been replaced by phosphonatone, given it can be infused three times faster and it's been shown to have better tolerability. Similarly, phenobarb is often not used or typically avoided, given its potential to cause respiratory depression. Valproate and levotiracetam are also frequently used, given that they have minimal adverse effects that could potentially limit their tolerability. While it's been theorized that valproate infusions can have some mild hypotension and respiratory depression, some literature actually shows that rates of up to 10 milligrams per kilo per minute are both safe and tolerable. Looking at some of the previous studies, Agarwal et al. included status epilepticus patients that were refractory to IV diazepam, and they ultimately concluded that valproic acid is equally as effective as phenytoin, but with better tolerability. Then Alvarez et al. evaluated phenytoin, valproic acid, and levotiracetam, and while it was retrospective in nature, they found no difference between valproate and phenytoin, and they also concluded that levotiracetam is less effective than valproic acid, but I think something important to note about this study is that they use submaximal doses of both valproic acid and levotiracetam. I did also want to mention that there were two rather large trials that were recently published in entirely pediatric populations. The CONCEPT and ECLIPSE studies compared maximum doses of phenytoin to 40 milligrams per kilo of levotiracetam, and found no difference in termination of seizures, but better tolerability of levotiracetam, and I think even with these previous studies, it kind of still leaves the question of what the most effective anti-epileptic drug is. It leaves that question still unanswered, which leads us to the trial that we'll be discussing today, which again is the ESET study. The objective of the study is to determine the most effective and or least effective treatment of benzodiazepine refractory status epilepticus. The trial was a prospective, randomized, double-blind, comparative effectiveness trial with 57 participating centers, and it was funded by the National Institutes of Health along with the FDA. Here I have the inclusion and exclusion criteria largely for your reference. However, there are a few key points that I did want to mention. The study included patients taking anticonvulsants for the treatment of epilepsy, and they were randomized without regard to their maintenance medication. Additionally, they also excluded patients that were intubated, had liver disease, or severe renal impairment. Patients were randomized one-to-one-to-one to receive fosfenitone, levotiracetam, or valproate over 10 minutes. The study medication had to be administered between 5 and 30 minutes after the last benzodiazepine dose. Additionally, rescue therapy was given as clinically determined if seizures persisted or recurred 20 minutes after the study infusion. The study initially aimed to include 795 enrollments in order to provide 90% power to detect a 15% difference. And then once 300 subjects were randomized, they then used a response-adaptive randomization method in order to maximize the likelihood of identifying the most effective treatment arm. Interim analyses began once 400 subjects had been enrolled, and then it was continuously repeated in order to screen for success or futility, and plans ultimately stopped early for success if any of the agents were identified as more effective with a probability of 97.5%. And then the study did end up stopping early for futility after 400 enrollments in 384 patients, since there was a 1% chance of finding a difference. Of note, the 16 patients that were randomized twice or had two events included in the study, only the initial event was included in the intent-to-treat analysis of the primary outcome while both enrollments were included in the safety analysis. Given the randomized nature of the study, there really were no significant differences between the three groups. However, there are a few important things to note. 39% of the patients enrolled were children and adolescents under 17 years of age. A majority of the patients had a history of epilepsy, which is interesting because it kind of contradicts the older epidemiological data that I previously presented. Also, 30 patients actually did not have status epilepticus but were ultimately diagnosed as having a non-epileptic spell or a psychogenic seizure. Additionally, the median lorazepam equivalent administered was 4.7 to 5 milligrams. But I think the most concerning data point here is that patients were seizing for roughly an hour prior to the initiation of the second line anti-epileptic. And lastly, as we can see, there was significant non-compliance to the inclusion criteria, as there were a good number of patients that did not receive a benzodiazepine prior to randomization. The primary outcome evaluated cessation of seizures and improvement in consciousness at 60 minutes without other anti-convulsant medications, and that includes medications that are used for endotracheal intubation. As you can see, none of the three agents met the pre-specified threshold for most or least effective, which was a probability of greater than 0.975. Also, the results remained non-significant in the per-protocol and the adjudicated outcomes analysis. Additionally, there was a post-hoc analysis done to evaluate the cessation of seizures at 20 minutes after the second line anti-epileptic was initiated, and again, they did not find a significant difference. Similarly, they found no differences in regards to secondary outcomes, which included ICU admission, ICU and hospital length of stay. And then their last outcome, was time to seizure cessation, which they really only had data available on 14 patients, so they were unable to find a significant difference. Safety data was collected for 24 hours after enrollment. All of the agents were found to be equally tolerable in terms of life-threatening hypotension, arrhythmia, intubation, and seizure recurrence. Of note, numerically more patients in the phosphenytoin group did have life-threatening hypotension, which is consistent with its side effect profile. The authors ultimately concluded that in benzodiazepine refractory convulsive status epilepticus, none of the three evaluated anti-epileptics were shown to be superior or inferior. Phosphenytoin, valproate, and levotriacetam were effective in approximately half the patients with benzodiazepine refractory status, and there's really no difference in the time to cessation of seizures, but again, just keeping in mind that that data was only available in a very small number of patients. The study did have several strengths. To date, it is the largest sample size evaluating benzodiazepine refractory status epilepticus, and it's probably the best quality evidence we have, and as an added bonus, it wasn't funded by a pharmaceutical company, but actually by the NIH and the FDA. I think it's also a major strength that they evaluated levotriacetam and valproate at maximum doses. It's really the first study we have that evaluates levotriacetam at a dose of 60 milligrams per kilo, which has been recommended by the American Epilepsy Society guidelines since 2016. They also included patients that were taking anti-epileptics for seizure control and were randomized without regard to their maintenance anti-epileptic, which while it does boast generalizability, it could be theoretically a double-edged sword given that phosphenytoin is a neurotherapeutic drug, but I think the generalizability ultimately was limited by their exclusion criteria, mainly because they excluded patients with renal failure and endotracheal intubation. Also, I think it's important to take note of the fact that 10% of the enrolled patients didn't even have a seizure, but were actually experiencing a non-epileptic spell. Additionally, while EEG data was a part of their definition for acute seizure recurrence, that data was only collected in 60% of patients. Ideally, we would have liked to see that required in all patients, since we know that while a seizure may terminate clinically, there could still possibly be burst activity on an EEG. There was also significant deviation from the protocol. Roughly 43% of patients didn't even receive a benzodiazepine prior to getting a second line anti-epileptic, and while this may not necessarily be a limitation, I think it's a point of controversy that patients only received 4 to 5 milligrams of lorazepam equivalents, while the maximum is 8 milligrams that we can give. The most concerning point to the study is that patients were seizing for roughly an hour prior to receiving a second line anti-epileptic, which wasn't really addressed. I think I would have really loved to know what the rate limiting factor was here, or have some sort of idea why patients were seizing for that long prior to getting another anti-epileptic. Lastly, they ultimately used submaximal doses of phosphenytoin dosing in patients weighing greater than 75 kilos, and this was limited by their 10-minute infusion time, and while guidelines recommend a maximum dose of 1500 milligrams, there is some literature published in the Annals of Emergency Medicine suggesting that an uncapped phosphenytoin dose is both safe and effective. From this study, we still don't know what the optimal second line anti-epileptic is for status, and as all of them were equally ineffective and had similar tolerability. Treatment of status epilepticus is time-sensitive, and it's reasonable to choose an agent that's effective and able to be administered rapidly without concern for infusion rate limiting adverse effects, and has the least amount of contraindications. If we do get any more studies that evaluate second line anti-epileptics, ideally we'd want to see them administered at the very least within the guideline recommended time frame, because if you think about it, if patients were seizing for an hour and upwards of 40% failed the primary outcome, that means 40% of patients were ultimately seizing for at least an hour and 20 minutes. So that kind of tells us that what we're doing right now for the management of status epilepticus isn't ideal for early termination of seizures. So before I conclude today, I'd like to ask a few polling questions. The first being, what is your institution's preferred second line anti-epileptic in the setting of status epilepticus? So this is probably what I anticipated that a majority would pick levotracetam. Currently at our institution, we typically favor levotracetam as it's easy to administer, it's extremely accessible, and it's very well-tolerated. My next question is, which continuous infusion do you prefer for refractory status epilepticus? Okay so I think for me I think it really depends on what the patient status is or what the clinical scenario is. I do tend to kind of prefer propofol especially for intubating and the patient's hemodynamics are okay but it's interesting to see that a majority are picking midazolam and I was actually curious to see if anyone had moved towards ketamine but it doesn't look like anyone has. Okay great so with that I'm going to conclude my presentation and open it up to any questions. Thank you Justina. The first question is have there been any studies comparing levatoracetam loading dose strategies to determine the optima loading dose? The AES guidelines were written by the ESET authors which may be why they recommended a 60 milligram per kilo loading dose. So I'm not familiar with any studies that have really evaluated different loading strategies however I am aware of the fact that the American Epilepsy Society guidelines were written that that 60 milligram per kilo load of Keppra was written like keeping in mind that the results of the ESET study would likely change practice but unfortunately no I'm not really familiar with any Keppra loading strategies. Thank you. The second question is given the low rate of seizure termination should we be recommending two concomitant AEDs for benzodiazepine refractory seizure? So I think that might be where we move next. I think that's where the treatment of status epilepticus might move. I don't think we're there yet. I think just keeping in mind that what we're doing is likely not the best for the management of status. We're still having patient C's for a really long time. I think that might be where we move to where we're being a lot more aggressive a lot earlier on. Thank you. Next question is from a logical logistical drug preparation delivery standpoint. Do you have a recommendation for most for the most timely second-line agent? Well I think that kind of depends on the institution you're at. I know for my institution we have Keppra loaded in almost every single omni cell. It's very easily accessible whereas phosphonatone and valproate are we have to get those from the main pharmacy so it will take much longer for that to get to us. So I think that really will depend on the institution. Again just for us Keppra is the easiest and most accessible. Thank you. Next question is if Keppra is not available what would be your next agent to use and why do you think not everyone in the study received a benzodiazepine? Okay so those two questions. I think so the first one if Keppra was no longer available. If Keppra was suddenly not available I genuinely don't think I would have a preference between phosphonatone and valproate. I think possibly maybe towards valproate since it isn't neurotherapeutic and if I don't know what the patient was on previously I'd have a little bit less concern since the therapeutic window of valproate is so large. So I don't necessarily have too much of a preference there easily they're both not as accessible to me so not too much of a preference there. In terms of why not everyone got benzodiazepine prior to getting the second line antiepileptic it unfortunately depends on how they were brought to the institution like depending on the ambulance I think. So I know that BLS ambulances cannot administer any medications while ACLS ambulances can. Thank you. That concludes our Q&A session. Thank you Justina Gerges. Thank you. Thank you to our presenters today and the audience for attending. Please join us on the third Friday of the month from 2 to 3 p.m. Eastern Standard Time for the next Journal Club Spotlight on Pharmacy. That concludes our presentation today.

Journal Club webinar

pharmacy practice

medication regimen complexity scoring tool

critically ill patients

pharmacist services

ICU

solid organ injuries

status epilepticus

anticonvulsant medications

ESET study