Hello, and welcome to today's Journal Club Spotlight on Pharmacy webcast, which is supported by the Society of Critical Care Medicine CPP section. My name is Jamie Robinall Gray, and I'm a Transitions of Care Clinical Pharmacy Specialist at Temple University Hospital Mean Campus in Philadelphia, PA. I will be moderating today's webcast. A recording of this webcast will be available to registered attendees. Log into 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. You may also follow and participate in live discussions 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 presenter and do not necessarily reflect the opinions or views of SCCM. SCCM does not recommend or endorse any specific test, physician, 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 a Q&A. Our first presenter today is Melissa Cole, a PGY-2 critical care resident at ProMedica Toledo Hospital, Russell J. Abebe Children's Hospital in Toledo, Ohio. Our second presenter is Abe Karimi, a PGY-2 critical care resident at Avira McKennan Hospital and University Health Center in Sioux Falls, South Dakota. And our third presenter is Gabriella Young, a PGY-2 critical care resident at University of California San Francisco Health in San Francisco, California. And now I'll turn things over to our first presenter. My article will be Therapeutic Anticoagulation in Critically Ill Patients with COVID-19. I have no disclosures for this presentation. The objectives of this presentation are to describe the pathophysiology of hypercoagulability in COVID-19 and to evaluate the effect of therapeutic anticoagulation in organ support free days. To go over a little bit of background information, hypercoagulability, specifically in COVID-19, starts with a dysregulated immune response. And that immune response can cause inflammation and release of cytokines like IL-6 and tumor necrosis factor alpha, which trigger a release and overall activation of the intrinsic pathway within the coagulation cascade. And so what we can also see from this is that we will see an increase in fibrinolysis as well as an increase in fibrinolysis inhibition. And so overall, we're actually going to see an inhibition of fibrinolysis. And that will also be shown up in labs as an increased fibrinogen and an increased D-dimer. When we look at the risk for thromboembolic events in non-ICU and ICU patients with COVID, we can see that it's going to be more common in the ICU population. And if we're thinking about COVID specifically, there was a report that said that PEs occur about twice as much in COVID patients if we were comparing this to the influenza population. In terms of guidelines that they have out right now, the National Institute for Health and Care Excellence in the United Kingdom, they make recommendations specifically for treatment dose in patients who are requiring supplemental oxygen and doing that treatment dose for 14 days or until discharge. And if the treatment dose, while on treatment dose, these patients have increased oxygen requirements, they recommend to reduce the anticoagulation to intermediate or standard dose prophylaxis. In the United States, they do have a little bit of a different recommendation. And unless a patient requires therapeutic anticoagulation for a different indication, it is recommended that COVID-19 patients receive prophylactic intensity anticoagulation. And this will bring us to our first polling question. So how does your institution handle anticoagulation for critically ill COVID-19 patients? Do they do A, low-dose prophylaxis, which is something like an oxyparin 40 milligrams daily? B, intermediate-dose prophylaxis, which could be an oxyparin 1 milligram per kilogram daily or 0.5 milligrams per kilogram twice daily? Do you use treatment-dose anticoagulation or other? So it looks like we have kind of a tie between using low-dose prophylaxis or intermediate prophylaxis in our patients. And so what this study had wanted to go ahead and look at, the hypothesis was that therapeutic anticoagulation is associated with improved outcomes in critically ill patients diagnosed with COVID-19. This was an open-label, adapted, multi-platform, randomized controlled trial. And they actually, this specific study is a subset of the critically ill population from a larger trial, which is the ACTION trial that was just published in the Lancet. And so they took patients from the REMAP-CAP trial, the ACTIV-4A trial, and the ATTACK trials who had anticoagulation and randomized those patients to look at treatment-dose versus usual-dose thromboprophylaxis. And so the inclusion criteria for this was confirmed COVID-19 and severe COVID-19, which was defined as requirement of organ support with high-flow nasal cannula or invasive ventilation or non-invasive ventilation, vasopressors, or inotropes. And some key exclusion criteria were patients admitted to the ICU for greater than 48 hours or admitted to the hospital for greater than 72 hours, a high risk for bleeding or receiving dual antiplatelet therapy, a clinical indication for therapeutic anticoagulation, and a history of heparin sensitivity, including HIT. The statistical analysis for this included monthly interim analyses, and so at that point they randomized the patients that they had and determined superiority and futility. And so superiority they defined as a greater than 99% probability that the odds ratio was greater than 1, where greater than 1 was favoring therapeutic anticoagulation. And futility was defined as a greater than 95% probability of an odds ratio less than 1.2. And so again, an odds ratio greater than 1 would show a better outcome with therapeutic anticoagulation. And specifically for the primary analysis, they adjusted this for age, sex, site, and time period. The primary and secondary analysis, they did utilize a Bayesian cumulative logistic model because they did have ordinal data, and we'll discuss that data a little bit further. And they did do a protocol analysis based on the patients, whatever they received for their initial treatment within the first 24 hours was which group that they were assigned to. The primary outcome for this study was organ support-free days to day 21, and as I mentioned, this was an ordinal marker. And so this scale was composed of survival to discharge, and in survivors, it was the number of days free of organ support up to day 21. So a patient who died during hospitalization got a negative 1, and then that could go up from there, assuming multiple days of organ support free. The secondary outcomes include survival to day 90 and major thrombotic events or deaths through day 28 or hospital discharge. And thrombotic events include myocardial infarction, pulmonary embolism, ischemic stroke, and systemic arterial embolism. Some safety outcomes that they looked at was major bleeding during treatment period, which was defined as ISTH for non-surgical patients, and the incidence of heparin-induced thrombocytopenia. When we look at some of the background characteristics for these patients, we saw that their age was not significantly different, around 60 years of age, with a total body weight around 80 kilos. The APACHE score in both groups was 13, which was not very high as what we might see in some critically ill populations in terms of severity of illness. And also interestingly, when we look at D-dimer, we see that in both of these groups, these were in the 800s. And this is something that we'll discuss a little bit further as we move on into the critique of the trial, but at our institution, we had guidelines where we would consider immediate or therapeutic anticoagulation in patients with D-dimers closer to 1500 nanograms per milliliter. And there is some other literature trials, for instance, the INSPIRATION trial, where their baseline values were a little bit closer to 1000 for D-dimers on presentation. In terms of the primary outcome, organ support-free days to day 21, we saw that there was no significant difference, and actually there was a 99% chance probability of futility in this, saying that therapeutic anticoagulation was not useful in this population in terms of organ support-free days. And when we look into secondary outcomes for these patients, we also saw a non-significant difference in terms of survival to discharge. And interestingly, when we look at the major thrombotic events or death, we see that that number is mostly impacted by mortality because there was a relatively low instance of thrombotic events at 5% and 10.3%. And the most common in terms of thrombotic events was pulmonary embolism for each group, followed by ischemic cerebrovascular events. The author's conclusions for this study was that therapeutic anticoagulation did not improve outcomes in severe COVID-19 patients, and they did also speculate that perhaps initiation of therapeutic anticoagulation at patients who were this critically ill, it might be too late, and perhaps utilizing therapeutic anticoagulation earlier when patients are not as severely ill might be more beneficial. Looking at the strengths of this trial, as I've mentioned, this is a large multi-centered randomized control trial that used three of the large COVID studies to utilize for their patients. They did also utilize inclusion and exclusion criteria to avoid patients with suspected COVID because the REMAP-CAP trial did utilize patients with suspected COVID, and they did also exclude patients who had other co-founders that could contribute to increased thrombotic or bleed risk. And they did also, in terms of their primary outcomes and secondary outcomes, include outcomes from all three trials, which was really helpful in determining the significance of anticoagulation. Some weaknesses to consider for this is it was an open-labeled trial, and they did also utilize per-protocol analysis. So a patient could receive usual care thromboprophylaxis of anoxaparin 40 milligrams daily for 24 hours, and then they could be changed to an alternate therapy, but they would still be considered in the usual care prophylaxis group. They also allowed the dosing practices specifically based on the local institution. And so a large proportion of the patients were from the United Kingdom, and based on the NICE guidelines that we had previously, intermediate dose prophylaxis was used much more prevalently in this. And so intermediate dose in the therapeutic anticoagulation group was actually 10 percent of that population, and in the usual care group, 51 percent of those patients received intermediate dose. So it's a little bit more difficult to say where treatment anticoagulation does that better compared to intermediate dose or usual care. We also did have a variability in terms of duration of therapeutic anticoagulation. And since there were a lot of guideline recommendations and changes as the COVID-19 pandemic has gone through, the authors did try their best to try to compare patients based on the time in which they received COVID, but there are inherent changes in practice throughout that they could not control for. We did see with major bleeds that they were numerically higher in the treatment group, but not statistically significantly different. But when we compare this to previous literature, it was about similar to what we had previously seen. Additionally, we saw that thromboembolic events were numerically higher in the usual care group, which was similar to previous literature, but it's unclear how much this truly impacted organ support-free days, since it was a combined outcome of mortality as well as thromboembolic events. And finally, as I had mentioned previously, the D-dimer levels were numerically lower than what we saw in previous studies. As I mentioned, the inspiration trial, their baseline values were anywhere from 900 to 1,000 nanograms per milliliter, whereas some other studies have baseline presentation D-dimers upwards of 2,000. So these patients did not necessarily have as elevated D-dimers as we might think in our practice in terms of needing to utilize anticoagulation therapeutically. And now we have our second polling question. So based on the results of this trial, would you change your practice for anticoagulation in critically ill COVID-19 patients? Okay, and so it looks like the majority of them said no. And I think based on this information, I would agree that we're seeing that therapeutic anticoagulation is not helpful for this population and kind of more well-aligned with what we're having in terms of the current American guidelines. Some key takeaway points to take from this trial is that therapeutic anticoagulation does not improve organ support free days in critically ill COVID-19 patients. It's still very unclear how hypercoagulability in COVID-19 patients affects organ function, specifically in the critically ill population. They had noted that there are some studies out there that they had done postmortem autopsies which showed more hemorrhagic bleeding. And so kind of seeing the fine line between having thromboembolism versus bleeding is still not clear and fully clear in the COVID-19 population. Unless there is an indication for therapeutic anticoagulation, patients with COVID-19 can receive prophylactic dosing but don't necessarily need to receive treatment dosing for COVID-19 specifically. But further literature is warranted to better understand the pathophysiology of coagulation and how anticoagulation can affect outcomes in COVID-19 patients. I thank you all for listening to my presentation and I'll take any questions at this point in time. Thank you for the presentation, Melissa. So from, I guess, your perspective at your institution, will this change your practice there at all? Or where would you see therapeutic anticoagulation play a role in these COVID-19 critically ill patients? Typically, what we have seen in our practice is that we won't necessarily give them treatment dose, might consider doing intermediate dose anticoagulation in patients with elevated D-dimers, but more often than not would not. So I would definitely not include therapeutic anticoagulation for our COVID-19 patients. Again, just give them therapeutic anticoagulation if they require it for other indications. Thank you. So that concludes our Q&A session. Thank you, Melissa Cole, for your presentation. Before moving on to our next presenter, we would like to ask a brief polling question regarding today's attendance to gain a better understanding of our overall attendance to ensure continued support of this Spotlight on Pharmacy webcast. How many attendees are you viewing this webinar with? Please select one, just me, two to five people, five to 10 people, or greater than 10 people. Now, I'd like to introduce our second presenter, Abe Cremey. Thank you for the introduction. My journal club article that I chose is, Effect of Phenylephrine Push Before Continuous Infusion of Norepinephrine in Patients with Septic Shock. And I have no disclosures to disclose at this point in time for this presentation. So before we get started here and jumping into the background here, when it comes to phenylephrine utilization in septic shock, it's good to kind of review some of the previous surviving sepsis guidelines. So looking at the 2012 surviving sepsis guideline, when it comes to phenylephrine, they actually recommended to not use it except when norepinephrine is associated with arrhythmias, cardiac output is high and blood pressure is low, or as a means of salvage therapy. Later on in 2016, when the newer guidelines were published, they actually removed the phenylephrine utilization due to the lack of evidence with its benefits in the septic shock population. Despite what previous and current guidelines have stated in the past, we all kind of understand that vasopressor pushes are still utilized and administered in practice due to the rapid availability of the drug at bedside, usually with the goal of attenuating acute hypotension and kind of bridging that patient to a more continuous infusion vasopressor. Situations in which this method of push dose is utilized can vary, which include within the peri-intubation period, within specific critically ill patients, patients that are post-cardiac arrest, or undergoing critical transport, or overall utilized in the emergency department, and lastly, sometimes in septic shock patients. Looking at previous literature review, I pulled seven studies that were done in the past that looked at different vasopressor push utilizations, and it ranged from epinephrine, phenylephrine to ephedrine. We can see that mostly all of them kind of lack looking at specifically septic shock patients. There was a study done by Rotondo et al., which is third from the bottom there, and it looked at ICU patients. This patient population that they looked at received push dose vasopressors, but it varied for their diagnosis. Their diagnosis ranged from respiratory failure, trauma, cardiac issues, or sepsis. Therefore, no data truly exists looking specifically at septic shock. And then quickly looking at phenylephrine itself as a specific vasopressor we're focused on, we all kind of know that it's a selective alpha-1 receptor agonist that causes vasoconstriction, resulting in increased systematic vascular resistance and pulmonary vascular resistance with the goal of increasing mean arterial pressure and overall improve perfusion. Therefore, it is thought to be useful in emergent situations to raise mean arterial pressure, which is why we utilize it sometimes. But because of the increased systemic vascular resistance, phenylephrine may decrease stroke volume and may increase ventricular afterload in the setting of myocardial dysfunction or intravascular volume depletion, which overall is kind of a detrimental effect that we don't want. In addition, it could have an impact on microvascular blood flow. So, in light of the lack of clinical data with phenylephrine usage in septic shock patients and the potential pharmacologic concerns with its use and continued use of it as a push-dose vasopressor, researchers at Cleveland Clinic Health System performed this retrospective study that I stated earlier, the topic of this discussion. Overall, the goal of this retrospective study was to evaluate the hemodynamic and clinical effects of using phenylephrine pushes before continuous infusion of norepinephrine in adult patients with vasopressor-dependent septic shock, and overall looking at its hemodynamic stability and its use. This, like I stated before, this was a retrospective observational study. They had a multi-center cohort. The study ranged from January 2012 to November 2017, and they focused on those patients with septic shock initiated on norepinephrine. For the inclusion-exclusion criteria, patients had to be 18 years or older, admitted to an ICU such as medical, surgical, neurosciences, or mixed, receiving continuous infusion of norepinephrine, or meet a United States Centers for Disease Control and Prevention definition of adult sepsis event. This event was defined as the presence of sepsis, was persisting hypotension, mapped less than 65 milliliters of mercury requiring continuous infusion of norepinephrine. Also, they needed blood cultures obtained within 48 hours before or after norepinephrine initiation, and four days or more broad-spectrum antibiotics within 48 hours before or after the first blood culture specimen was obtained, and also a serum lactate of greater than two within 24 hours of norepinephrine initiation, and that was their definition. Exclusion-wise, they excluded patients that received the phenylephrine push 60 minutes before or 120 minutes after initiation of norepinephrine, received a vasoactive medication before transfer facility, and then received continuous infusion of phenylephrine. So, interventions-wise, patients were divided into two cohorts for evaluation, those who received phenylephrine push before norepinephrine initiation, and those who did not receive the phenylephrine push. Patients had to have documentation of phenylephrine push within 60 minutes before or within 120 minutes after initiation of norepinephrine continuous infusion. This time window was utilized because often within their system, there was a lag from when the phenylephrine push was administered to when the time of the electronic medication record ordering and documentation had occurred. The average dose is listed there, a phenylephrine of 394 micrograms of phenylephrine, on average receiving about two phenylephrine push doses with a range of 1 to 12 pushes. The primary outcome looked at achievement of hemodynamic stability at 3 and 12 hours. Hemodynamic stability was defined as the time at which MAP reached greater than or equal to 65 millimeters of mercury and continued to be greater than or equal to 65 millimeters of mercury for six hours while the dose of continuous infusion vasoactive agents was not increased. They chose 3 and 12 hours mainly to look at the more immediate phase of the phenylephrine duration and then more of the after effects of the phenylephrine, choosing 12 hours. That's based off of the PK of the drug. Secondary outcomes, they looked at ICU and hospital mortality, ICU and hospital length of stay, duration of mechanical ventilation, vasoactive duration, so other vasoactives that may have been used during the duration of phenylephrine and norepinephrine, and change in SOFA over 72 hours, change in or trend of lactate values over 24 hours, and then MAP or mean arterial pressure over 24 hours. So kind of getting to the statistical analysis, to start off, this study kind of did a little bit of more fancy type of statistical analysis for a retrospective study. They utilized a propensity score matching for their way of developing the match cohorts. So to kind of give a little bit more of a background on propensity score matching, it's a Cisco matching technique that attempts to estimate the effect of a treatment by accounting for the covariates that predict receiving the treatment. So the covariates that are listed there are ones that the authors chose based off of their collaboration and determining what would affect these patients to needing phenylephrine. Overall, propensity score matching's goal is to reduce the bias due to confounding variables. In randomized experiments, we kind of know the treatment groups will be balanced on average by the law of large numbers. However, in observational studies like this, the assignment of treatments for research subjects is not random. Therefore, matching, which is what they did here, attempts to reduce the treatment assignment bias and mimic randomization. So like I said before, these are the covariates that the authors here chose. Other Cisco analysis they utilized was propensity score plot and love plot to examine the balance between propensity score and covariate distribution between groups. They also utilized a GE logistic regression or a generalized estimating equation model to predict outcomes, and I'll touch on this when we look at the results. And they also used multivariate Cox regression model to account for concern of immortal time bias. They used this Cisco equation to calculate time of hemodynamic stability. So overall, when we utilize propensity score matching, it's important to have a large cohort to begin with. And so that's what they did here. They had a cohort of about 4,000 patients or 3,797. After exclusion and matching, we had about 141 patients match the phenylephrine push cohorts and then 282 match to the no phenylephrine push cohorts. This was a 1 to 2 ratio cohort matching that they did here for this study. And so we see that there was more in the no phenylephrine push cohort here. Looking at baseline characteristics, initially before matching using the propensity scores, there was actually a vastly large difference between the cohorts. However, after matching, there was similar for most of the baseline characteristics except for the ones listed below. So for patients that had COPD dialysis or AIDS and then those receiving epinephrine, these are the ones that had a difference from baseline characteristics. Looking at standardized mean difference, a cutoff was 10%. So if there's a 10% difference in standardized mean difference, that was their threshold for having a difference within the baseline characteristics. So now we kind of get into the results. Here with the primary outcome of hemodynamic stability at three hours versus 12 hours, we can see that more patients who received the phenylephrine push achieved hemodynamic stability at hour three than those who did not receive a phenylephrine push compared to a 12-hour. You see the percentages listed there. They used a risk difference within their calculation. And with that 95% confidence interval, it not including zero shows that it was statistically different compared to the 12-hour since it's a difference data that they collected here. Looking at multivariate adjustments. So first off, with the GE multivariate logistic regression analysis, they showed that phenylephrine push was independently associated with higher incidence of hemodynamic stability within three hours with that p-value given there at 0.02. But while utilizing the multivariate Cox proportional hazards model, they showed that they did not detect association between phenylephrine push administration and time to hemodynamic stability within three hours. In addition, phenylephrine push receipt was independently associated with higher ICU mortality, but showed no independent association with time to death at 28 days. Oops, did not mean to go that fast. So looking at the table here, you kind of see the p-values for the ICU mortality of less than 0.05, so showing a difference. And then time to death at 28 days was no difference there. Other secondary outcomes they look at was heart rates. They looked at reduction in heart rate and for the first 24 hours after norepinephrine initiation for seven beats per minute versus six beats per minute in phenylephrine push versus no push, they showed a significant change according to their statistics that they provided. As far as serum lactate concentrations and MAP trends, there was no differences between them throughout the study. So overall, the author's conclusion stated that phenylephrine push was associated with early but not sustained hemodynamic stability in patients with septic shock. In addition, phenylephrine push independently associated with higher ICU mortality. Because of what they came up with results here, they cautioned the use of the phenylephrine push prior to norepinephrine initiation in patients with septic shock that are hemodynamically unstable. So kind of moving into my critique here for strengths and limitations and overall thoughts on the study. So starting with the strengths, we all can kind of agree that this was a relevant research question. A lot of times we get patients that are septic, even to the ER or into the ICU that are hemodynamically unstable and do not respond to fluids. And we try to resuscitate them as far as getting their MAPs greater than 65. And the question does come up of what to use in the emergent phase. So I thought this was a pretty relevant research question. The statistics that they did was pretty fancy, I thought, and tried their best to kind of remove bias. So the propensity matching and GE regression that they utilized, I thought, was good on their part. In addition, their other baseline characteristics, such as SOFA scores, APACHE-3 scores, lactate values, fluids, steroids, and left ventricular ejection fraction that they followed within the baseline characteristics were all fairly similar to one another between the two cohorts. So I thought that was a well-matched there between them. However, kind of transitioning into the limitations, starting with the fact that they discussed in the beginning that this is a multi-center study throughout the Cleveland Clinic, which included the main campus and other rural areas outside of the main campus. However, when you look at the matched portion, so once propensity scores were determined and patients were actually matched, about 80% of the patients were from main campus and not really from any other rural sites or outside of the main campus area. In addition, 50% of the patients were in the medical ICU compared to other ICUs stated in the beginning of this presentation. So overall, once matching had occurred, there wasn't too much diversity as far as location and what type of ICU these patients were in. In addition, even though they utilized propensity matching to reduce bias, the fact that authors collaborated to create the covariates used to create the propensity matching criteria, I still think there is bias still there, mainly because it was based off author collaboration in determining the covariates and not really based on previous studies, mainly because this is kind of one of the first studies to look at this phenylephrine push before norepinephrine initiation in septic shock patients. Also, they discussed the fact that phenylephrine and its use was related to decrease in heart rate. However, the results that they showed, kind of looking at it, was not really clinically relevant. In addition, they did not really comment on timing of steroid therapy initiation. We all kind of know the timing of steroid initiation can have an effect on reaching a map and things like that and holding blood pressure. So they tracked glucocorticoid use, but they did not discuss the timing of it and whether or not it fell within the three versus the 12-hour. In addition, they did not comment on the timing of other vasopressor initiation. They just looked at the vasoactive duration and not the timing of the initiation. So overall, the conclusion of the presentation here, I would say that caution is warranted when administering IV push phenylephrine in septic shock patients. Just with the results they had with IC mortality and things like that and other detrimental effects that phenylephrine can have on patients that we've kind of known from previous knowledge and the mechanism of the medication, therefore, it's important to kind of utilize readily available intravenous vasopressors on patient floors. At our facility, we don't necessarily have a specific protocol for providers to use phenylephrine pushes in a septic shock patient, but we do have a protocol that states that only a provider or a flight care nurse can push phenylephrine in an emergent situation. And usually, we have norepinephrine infusions readily available on our patient floors, and that's what we go to first. In addition, future research could look at push dose administration of other vasopressor medications and their utilization within this patient population. And therefore, I think that pretty much wraps up my presentation. We'll kind of go to our polling questions here. So the first one is, does your institution provide readily available norepinephrine or other vasopressor infusions on patient floors? So, I cannot fully see what the results are, but if somebody else can, they can comment on it. My screen is showing zero percent, but... It looks like 86 percent said yes. What was that? 86 percent said yes. And that's what I pretty much determined with the utilization of Pyxis or Omnicell or other forms of medication cabinets. We usually have them available on our floors. Trying to go to the next slide here. So the next polling question, does your institution have a written protocol for pushdose 111 or is it provider-driven? And this can be either for sept, or I guess here will be for like a septic shock patient or an emergent situation, but I think we'll just go with septic shock patients. Is there a written protocol or is it usually just provider-driven? Okay. It looks like 100% have stated that it's usually provider-driven. So overall, with those two polling questions, I think it's important to kind of understand our usual practice, now that the majority of you, about 89% of the attendees here have stated that they have vasopressor infusions available on the patient care floors. Probably makes more sense to have a pharmacist involved with these types of emergency situations for septic shock patients available so they can kind of anticipate the need for a vasopressor infusion and not need to really give any pushes of phenylephrine and just have the infusion available. But that concludes my presentation. Thank you for listening, and I'll open it up for questions from anybody. Thank you, Abe. So question, given the previous literature about potential harm with high MAP targets, do you think it would be have been helpful to have a little bit more of a description of the MAP values during the study period at the three-hour and 12-hour timeframes? Yeah, I think that you bring up a very good point. I think it definitely would, like, yeah, looking back at it, they did not track ranges of MAP targets that were hits. They just utilized the definition to include or exclude patients. So I obviously think that, yeah, you're correct there. And kind of comparing that and keeping that in the back of our minds when you look at the results that they had with IC mortality and things like that. Great. Well, thank you. This concludes our Q&A session. Thank you, Abe Karimi, for your presentation. Now I'd like to introduce our final presenter, Gabriella Young. Thank you so much, Jamie. Hi, everyone. I'm Gabby. I'm one of the PGY-2 Critical Care Pharmacy residents at UCSF. Thanks so much for staying tuned. I'm excited to present this medication and patient safety-focused journal club. The article I'm presenting is entitled, Descriptive Analysis of Unwarranted Continuation of Antipsychotics or ICU delirium during transitions of care. Apologies for the lag, but to provide a bit of background, ICU delirium affects up to 80% of critically ill patients and has been associated with worse outcomes. Antipsychotics are commonly given to these patients to alleviate agitation and delirium, but while these are commonly cited indications, a reminder that these are off-label uses and there is a lack of definitive data to support the use of antipsychotics for these purposes. The PADIS guidelines do advocate for routine assessment of delirium and optimizing non-pharmacologic approaches. Although PADIS discourages routine use of antipsychotics to treat delirium, there may be some benefit in short-term use of these agents in patients who experience significant distress because of their delirium, especially if they pose a harm to themselves or others. And while these might be appropriate during the acute phase of illness, antipsychotics may be continued after ICU and hospital discharge, both purposely and inadvertently. Transitions of care are known risk points for medication errors, which are experienced by up to half of patients transferring out of the ICU. A retrospective study found that antipsychotics account for 10% of the medications that are continued inappropriately upon ICU transfer. Long-term continuation of these unnecessary medications may pose a financial burden on patients and contribute to polypharmacy, potentially increasing the risk of adverse effects experienced by our patients. This brings us to our first polling question. I'm curious to hear about which of the following describes your institution's practice regarding medication order processing at ICU transition. Options include all medications are discontinued, then reordered upon transfer, meds are continued upon transfer with a mandatory med rec, meds are continued upon transfer without a mandatory med rec, or other. And so a majority of patients, I'm sorry, a majority of your institutions have their medications continued upon transfer without a mandatory med rec, and I think this might highlight one of the issues we'll talk about in this article. There are several studies that evaluate the use of, or the continuation of ICU-initiated antipsychotics, finding that up to 50% end up being continued after ICU discharge. Ideally, we should be discontinuing these medications once delirium is resolved. So why is it that they are so frequently continued? A study by Kobanek et al. investigated some of the reasons, and they found that prescribers might be hesitant to change or stop antipsychotics for fear of resurgence of patient behaviors, or because the medication was recommended by a psychiatry consult at one point. Surprisingly, the authors found that of these antipsychotics that were continued on the floor, 41% of primary treatment teams were unaware of the indication for starting this medication. So now that we've established this issue a little bit further, we can get on to the study. The research questions that were intended to be answered by the study, sorry, I'll go back. How often are ICU-initiated antipsychotics continued inappropriately, and what are the risk factors associated with this? To address these research questions, the authors conducted a retrospective analysis of adult patients who received at least three doses of a scheduled antipsychotic for a clinical diagnosis of delirium. Notably, they excluded patients who died while in ICU, who had antipsychotics ordered as PRN only, patients who were previously established on these medications, and patients who were ordered for an antipsychotic for an indication other than ICU delirium. The primary outcome of the study was the rate of continuation of antipsychotics at two time points upon transfer from ICU to the wards and at hospital discharge. Secondary outcomes were the risk factors for continuation of these antipsychotics, including patient and hospital characteristics. For their analysis, the authors used descriptive statistics as well as bivariate comparisons of a cohort who was prescribed antipsychotics at discharge versus a cohort who was not, and the authors also used a multivariable regression to evaluate risk factors. Moving on to our results, of 450 patients that were screened, there were 300 that met inclusion and exclusion criteria that were included in the final analysis. The table on the right demonstrates baseline characteristics but I'll just briefly go over these since we will revisit the variables when discussing risk factors. So in this cohort of 300, median age was 69. They were mostly male patients admitted for neurologic, infectious, or respiratory diagnoses. Now looking at their antipsychotic use and outcomes data. Again, looking at the entire cohort, there were 15% of patients that had a documented CAM-ICU assessment, but all of these patients had a clinical diagnosis of delirium. For the primary outcome, for the rate of antipsychotic continuation after ICU discharge was 52%, and the rate of antipsychotic continuation at hospital discharge was 61%. Looking at the figures on the right, at the top we have the agents that were most commonly started in the ICU that included haloperidol, olanzapine, and quetiapine. And looking at the antipsychotics that were most commonly continued at discharge, quetiapine and olanzapine were the most frequently prescribed agents. That brings us to our second polling question. Which antipsychotic is most commonly used to treat delirium in your ICU? So it seems like a large majority of the responding institutions use cryptiopine most commonly, so a similar pattern of antipsychotic use with a few that are also using Halol. Moving on to our secondary outcomes, this slide evaluates or demonstrates the risk factors for continuation after ICU discharge. You can see that admission and subsequent discharge from a mixed ICU, and this is compared to a medical ICU, carried at least a four-fold increased risk of continuing these medications at ICU transfer. Additionally, ICU length of stay was also found to be a significant risk factor with a lower risk of continuation with longer ICU stays. And the authors lastly identified a potential protective effect associated with the use of cortiopine, olanzapine, and haloperidol. I'll actually move forward from this slide in the interest of time. And so this table, again, looking at the continuation of antipsychotics, but this time risk factors at hospital discharge, ICU length of stay was found to be, again, a significant risk factor. We have a longer ICU length of stay that conferred a higher risk of continuing antipsychotics at discharge, and this was after adjusting for treatment duration and medications discontinued in the ICU. Additionally, patients with a, apologies, that block should actually be around days of antipsychotic use, and patients with a prolonged duration of antipsychotic treatment were less likely to have their antipsychotic continued at discharge. And finally, the most impressive finding that didn't get a lot of attention in the published paper is in the last row that hopefully you can all still see, despite some reformatting on this platform, but antipsychotics being discontinued while in the ICU was associated with a 99% reduction in the likelihood of prescription at discharge. To summarize some of our main findings, 52% of antipsychotics were continued after ICU discharge. There was a higher risk associated with discharge from a mixed ICU and a trend towards higher risk in a surgical ICU. In this cohort, a longer ICU length of stay was associated with a lower risk of continuation after ICU, and this suggests that maybe a longer time spent in the unit allows for more time to discontinue or taper these medications by the ICU team, and this may also explain the trend towards a higher risk of continuation in a surgical unit, since this population might have more turnover and shorter stays. Looking out on the right, 61% of antipsychotics were continued at hospital discharge. This time, there was a higher risk with a longer ICU length of stay when, interestingly, this was protective against continuation of these medications on the wards. In this case, a longer ICU length of stay may speak to a severity of illness, which is a known risk factor for delirium. Additionally, those with a shorter total treatment duration had a higher risk of these meds being continued at discharge, and this might be explained by initiation or re-initiation of antipsychotics close to discharge, giving us fewer opportunities to de-escalate treatment. The discontinuation of these meds while in the ICU was highly protective against antipsychotic continuation at discharge. The authors concluded that a large proportion of patients had antipsychotic therapy continued beyond ICU and hospital discharge, and they advised caution at care transitions to prevent unnecessary medication exposure. Moving on to the strengths and limitations of this study, for the strengths, I thought this was a clinically relevant study question, certainly one that would affect a lot of our practices. This is also a multicenter study with a diverse patient population conducted over several years. The authors used multivariable regression to identify independent risk factors, and the findings are consistent with other studies and adds to the body of evidence demonstrating this issue. As for some limitations, some are inherent by the retrospective design. Authors were not able to comment on the reasons that antipsychotics were restarted, and adverse events may be underreported, especially in the long term. Additionally, this study excluded PRN orders, even though these still lead to a risk of therapy being continued on the floor and at discharge, and so the scope of this problem might be larger than stated in the study. I also want to point out potential limitations of the regression analysis, for example, considering collinearity of things like a hospital length of stay and a treatment duration that were both included in the analysis. There were a few patient risk factors that were evaluated. I think I would be curious to hear about opioid and benzodiazepine use, as well as severity of illness. And I think what is concerning is that the authors identified that the likelihood of patients continuing on these meds might not be entirely driven by the patient or the clinical needs of the patient, but rather logistic factors like the type of ICU. And then finally, there was a lack of routine delirium assessment in this study, which speaks to a greater issue with inconsistent or difficult documentation of things like a CAM-ICU score. As far as what I'd like to take away from this study, ICU-initiated antipsychotics are frequently continued beyond ICU admission, and I think these high rates of continuation really demand our attention, and we should focus on the reasons for why they are continued and how we can prevent it. The main takeaway I wanted to highlight from this study is that discontinuation of antipsychotics while in the ICU was associated with a nearly 100% reduction in the likelihood of prescription at discharge. And so I wanted to highlight our roles as ICU pharmacists to perform thorough med recs at ICU transitions and also to encourage routine delirium screening using validated tools. And this will help us evaluate the efficacy and maybe opportunities to discontinue antipsychotics once delirium has resolved. And that concludes my presentation. My references are on the next slide, and I'm happy to take any questions. Thank you, Gabriela. At your institution, are patients routinely screened on the floor for delirium? Thank you, Keena. That's a great question. I'm actually not aware of the practices outside of the units since that's where I've been primarily on service, but I know that in our ICUs, CAM-ICU, which is the tool that we use at UCSF, is performed at least every shift. But I will have to get back to you about evaluation on the floors. Okay. And then I guess along with that is if a patient is transferred from the ICU to the floor and these agents are continued, how do you ensure communication to the floor team that is taking care of the patient in order to potentially optimize that being discontinued before discharge? Great question. So as much as we try to advocate for discontinuation of these agents while in the unit, it doesn't always happen depending on whether or not delirium resolves prior to transfer. I think one thing we typically use while at UCSF is we have EPIC as our electronic health record system, and we make use of little communication tools called iVents to really convey to any covering providers that might take over covering this patient that we are trying to de-escalate this treatment. And so we're usually using those as a means of communication to convey those recommendations once they leave the ICU. Okay. Great. Thank you. That concludes our Q&A session. Thank you, Gabriella Young. 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. And that concludes our presentation today. Thank you.

critical care pharmacy

therapeutic anticoagulation

COVID-19

ICU patients

antipsychotics

ICU delirium

transitions of care

phenylephrine push

septic shock