Thank you, appreciate that introduction, appreciate the opportunity to speak today. I have disclosures as shown on the screen that should not really relate to the talk. So we're going to go over a few challenges with pharmacotherapy optimization and selection in critically ill elderly patients and hopefully describe some medication specific strategies to consider in this patient population. We expect, as we likely all know, that the makeup of the overall population of the world as well as the population in the ICU will grow in the percentage of patients that are geriatric and considered elderly. By 2050, it's expected to be 30% in the overall population as well as the ICU. And we know that patients who are elderly have increased likelihood of a lot of comorbidities and about 25 to 50% of them at any time can be considered frail, meaning that they have a decline in various body functions and systems leading to an increased vulnerability and a chance of poor outcomes such as delirium, disability, and death. And those can be greatly increased even with small minor changes such as medication changes, elective surgeries, and things like that. So they're increased vulnerability. This study is a portion of the data coming from the Delhi study about two years ago, looking at the relationship of delirium and mortality in the ICU. And specifically in this slide, looking at the impact of frailty. And you can see that there is quite a relationship. Overall, about 40% of patients in this study who are 50 and older were considered frail, which is a pretty significant number given that that's a pretty low age cutoff. And you can see that there was increased delirium in patients who were frail, and that led to an increase in hospital and ICU length of stay. Additionally, there was an increase in mortality in patients who were frail in the overall cohort, as well as those who had delirium, as you can see on the bottom line. And this emphasizes and continues the thought that was shown by previous studies, especially a study shown three years prior that similarly showed an increased ICU and hospital length of stay in patients who were frail, and a decreased likelihood of discharge that they would be able to be fit and be able to self-care. So we see the impact of frailty in our older ICU population. This study published last year emphasized these results even further, looking at different geriatric ICU phenotypes. And what they looked at was phenotypes based on their risk factors, as well as the outcomes associated with that. And they broke the patients into seven different phenotypes, with phenotype A being a general control group of patients who are older with low SOFA scores, and phenotypes B through E having different characteristics, such as respiratory failure, increased age, moderate SOFA, and renal failure, with phenotypes F and G both being comprised of patients who are frail with low and high SOFA scores respectively. So you can see that the mortality was the highest in phenotypes F and G, despite especially particularly phenotype F having a low SOFA score, again, emphasizing the role that frailty and the increased vulnerability that these patients, these elderly ICU patients have that we're treating. And this, like many things, was replicated with patients with COVID-19. So this relatively small study looked at the different patient characteristics in those that recovered from COVID without being transferred to the ICU compared to those who were either transferred to the ICU or died. And as we would imagine, patients who had a poor outcome of mortality or being transferred to the ICU had increased age, increased comorbidities, as well as increased medications and a lower ability to self-care. And that resulted in an increase in the frailty index of these patients who died or were transferred to the unit. Additionally, that we know that there are changes and things to consider, many factors in pharmacotherapy in the ICU in general. So, and these things are changing constantly. So for example, patients who have changes in fluid status or hemodynamic instability may see changes in their serum concentrations of drugs that could be increased or decreased depending on the actual pharmacokinetics of that individual agent. On the flip side, you can see changes in, acute changes in renal or hepatic function leading to either an increase or decrease in excretion and metabolism, further complicating the care of patients in the ICU. And this can probably point to a bunch of studies to highlight this, but I think these studies from the mid-2000s highlight this idea really well. And that patients who are on low molecular weight heparin had anti-TANA levels that were significantly reduced in patients that were either on vasopressors or who were edematous compared to those who were not on vasopressors and not edematous. Looking at two really common clinical scenarios in the ICU and seeing that there's a clear changes in drug concentration and clinical effects for routine meds that we give. Additionally, we have just in addition to the ICU factors and the kind of the elderly overall outcomes that we looked at earlier, we have some pharmacokinetic and dynamic considerations in the elderly patients themselves. So across the board, you're mostly looking at decreases in absorption, metabolism and elimination, as well as changes to distribution, most likely due to increase in adipose tissue, decrease in albumin concentrations and other factors leading to alterations in their distribution. Similarly, you'll see changes to pharmacodynamics. So things like variations in receptor sensitivity, increased permeability of the blood-brain barrier, so increasing effects of certain medications and we'll get into that in a little bit, as well as changes to neurotransmitters are also going to play into this as well as these patients are just more likely to be on more medications at baseline with more potential drug-drug interactions. And so it's clear how the variation in effect and duration of medications that we give in the ICU can be changed quickly. So I want to use just a few scenarios or topics to kind of highlight some of these principles. We'll start with some of the sedatives and analgesics that we know can impact outcomes both in the unit as well as outside the ICU. So there are many risk factors for again, changes to the patient's mental status and psychological outcomes within the unit and outside. And many of these are non-modifiable and patient-specific as you'd imagine. But some of them we think are potentially modifiable, including the degree of sedation that we administer in the presence of factual versus delusional memories. That's highlighted by this study here, which looked at trauma ICU patients and found that patients who had delusional memories were much more likely to have worse quality of life scores post-ICU discharge. And while we know that medications aren't the only reason and the only mechanism by which that can happen, we think that they play a significant role in that. So this study by Girard and colleagues combined patients from the Brain ICU and MindUSA studies and looked at different phenotypes of delirium and showed that one, as we all know, delirium is very prevalent. You can see over 70% of patients had at least one day of delirium and the vast majority of these days, actually, there was more than one phenotype expressed by individual patients. Sedative-associated delirium was the most common type. And you can see that after adjustment for covariates, it was the most impactful for their global assessment of their cognitive function at both three and 12 months. And so we know from other data that medication-induced delirium can actually be reversible and can be rapidly reversible if we withdraw or at least minimize the offending agent or agents that are believed to be causing it. And so the combination of it actually being reversible, but having long-term cognitive outcomes really moves us or should move us to wanna minimize these medications, reduce doses with or have the goal of getting these agents off altogether, especially if we can not only impact the outcomes that we know are in the ICU, getting people off the vent, getting people out of the unit and minimizing other secondary outcomes like ventilator-associated pneumonia and other things, but actually improve their long-term cognitive function when they are discharged. This study, there've been a lot of studies, a lot of data published in the last 10, 15 years looking at the degree of sedation that we administer patients overall, but focusing on some early window. So the early part of their ICU stay. So this study was looking at their first 48 hours and described their sedation intensity using a sedation index tool and found that sedation intensity was significantly lower in survivors compared to non-survivors. And additionally, after adjustment, increased sedation intensity, age, and APACHE-2 were significantly associated with increased mortality at 180 days. And this persisted, not only actually was more emphasized, not when the sedation intensity was higher, not just in the first two days, but in the first five to seven days. So early sedation, deep sedation is bad, but as that continues, you can see this kind of dose time relationship for both of these outcomes. And as you would imagine in patients with COVID-19, this is also true and maybe even more so. So I think anecdotally, we can all say that patients with COVID-19 probably in our units had some changes to their sedation practices compared to those without COVID-19. That was probably even the most true early in the pandemic. And you can see here, this is one study highlighting that the sedation burden index is seen on the right, really indicating that these patients got much higher doses of sedatives and analgesics than those without COVID-19. And what that resulted in, as shown on the left, is about a doubling in the time a patient spent in coma within the first 10 days, leading to, after mediation analysis, an increase in mortality. So the actual presence of COVID wasn't really shown to be related to COVID-19, but simply the sedation burden index, the amount of sedation these patients have received. And that's gonna be even more prevalent in, again, our geriatric patients, as we talked about. They're much more likely to have prolonged exposure to these meds after they're administered, given their reduced metabolism and elimination, as well as the increased permeability to those receptors. And while there's not really a simple, real straightforward solution to this, we think application of best practices is probably our best bet. Reducing infusions, more bolus therapy, patient-specific selection, more analgo-sedation, less benzodiazepine-based sedation, all these things that we talk about. And this is just an example of a study that implemented that and did that. And I think that's something that we can all, I think, continue to try to do and take home to our ICUs is to not only develop and update, but make sure we're adhering to our local ICU pain, agitation, delirium guideline, and hopefully getting some of the results that they saw here. Again, more analgo-sedation, fewer benzodiazepines, fewer overall durations of their continuous infusions, which resulted in, at least in this study, reduction in their time on the bed and time in the ICU. To pivot to using some of these principles, focusing on our elderly ICU population in the, for some of our antimicrobials and our anti-infective therapies, I want to just think about some of these principles. We know we're trying to balance with really all of our meds that we give, but particularly in the anti-infective world, a balance between efficacy and toxicity and trying to get precision dosing and making sure that we get the right level of a drug without too much that's going to result in more toxicity. I think the difference between elderly and healthy young volunteers is shown by this study, looking at ertapenem AUC as given, and you can see there's about a 40% increase in ertapenem AUC in elderly patients as compared to young adults. And again, these are in healthy volunteers, so these results would probably be more dramatic in non-healthy elderly patients, but you can see 40% at baseline if you look at just free or unbound drug with ertapenem, that goes up to about a 70% difference. So while not the biggest deal with ertapenem, given its relatively favorable safety profile, you are looking at, I think, a principle that we know exists with other agents that we give, that these patients are going to get a much higher level of exposure. And this is highlighted, I think, by this study, looking at various antibiotics given to elderly hospitalized patients. You can see overall a relatively low percentage of patients having all of their concentrations of drug within their therapeutic range, with many of these patients having at least one, if not multiple serum concentrations outside the range. So again, challenging to not only to get levels high enough to maximize efficacy, but also not get into the world of toxicity, which studies like this show can be linked with high levels. So another reason for dose optimization and dose adjustments that are relatively frequent. So this study showed there was an increase in both neuro and nephrotoxicity with beta-lactams that had a high level. So these drugs were linked with these potential side effects and we think about that quite a bit with cefepime. That comes up a decent amount, at least in our institution, a drug that we give relatively commonly, especially empirically, given its spectrum of activity. And yet what we're frequently assessing is whether the patient's experienced cefepime-associated neurotoxicity, because there are always many reasons and it's a differential of exclusion. But what it seems to be is that there's a trend toward patients who are at doses higher than recommended for their age or renal function, as well as their actual renal function itself. So age and creatinine clearance, both were found to be predictors of cefepime neurotoxicity in this study. And there's an overall trend toward that being seen. So making sure that we are giving the right dose, I think is as crucial or rotating agents as needed. And so one of the things that we can do in addition to simply dose adjusting is take advantage of our pharmacodynamics and take advantage of how we administer a lot of these medications. So this is obviously not groundbreaking and new, but just thinking about ways that we can administer these agents safely by giving them over prolonged or continuous infusions, maximizing our pharmacodynamics and maximizing our time above the MIC while avoiding those peaks that we know are not necessary to achieve efficacy. And while we can't really, at least routinely, at least at Brigham, but I think across the board, routinely monitor a lot of the antibiotics that we give or a lot of the other drugs, we know that we can with vancomycin. As a pharmacist, I'll probably just give a plug that we monitor vancomycin levels too much, but we do at least have that option. And here we can see that, and we know that an AUC to MIC ratio of 400 or more is usually associated with improved outcomes, improved clinical cure, and more attainment of our pharmacodynamic goal. What this study shows, interestingly, I think, again, thinking more patient-specific pharmacotherapy is that elderly patients may not need those high serum trough levels and that those, while historically are associated with that pharmacodynamic goal of AUC to MIC of 400, that elderly patients may need only simply a level greater than 10, and they associate a level specifically of 12 with achieving that goal almost in all patients. So achieving levels of 18, 19, 20, probably not necessary in most of our patients and maybe putting them at risk of adverse effects. Finally, I wanted to kind of zoom out and look at just overall medication practices in the ICU and try to apply some of the principles that we've been talking about. So we know that there are many downsides of polypharmacy in any of our patients, but particularly in elderly patients, and a lot of it goes to noncompliance or nonadherence, but mostly adverse effects is what we're talking about. Increased drug-drug interactions, increased costs as well, but really looking at a function of adverse drug events as a function of age and the number of meds that we are giving. And we know that a lot of these potentially arise, some of these problems arise from the ICU itself. So this study out of Beth Israel in Boston showed that 20% of patients who were started on an antipsychotic in the ICU actually continued that antipsychotic inappropriately after discharge. And a study a few years after this saw that this number was upwards of 30%. So there's quite a bit of data showing that we know we're adding meds and maybe that's great and that's important in the ICU, but what we need to make sure of is that we're actively assessing to see this patient's leaving the ICU, do they still need this? This patient's leaving the hospital, do they still need this? Because we know there are a lot of short and long-term side effects of these meds that may be leading to readmissions and other problems. We know that this is not new. So this analysis looked at six areas in Europe over 10 years ago and found that when looking at 900 patients that inappropriate prescribing as well as inappropriate omissions were relatively common across the board. So these numbers are relatively alarming that this happens this often and we know this, but the challenge is really what to do about it and how to address it. And this meta-analysis similarly showed not only is it common, but this is something that is linked with adverse effects. So while it didn't show that the presence of inappropriate medications is linked with mortality, in the overall cohort, although there were some areas, regional areas where it did show differences, really what it's looking at is that when you add on inappropriate meds, you're going to get more adverse effects and you're going to get more hospitalizations and this risk increased as the number of potentially inappropriate meds increased. This prospective study looked at a little bit of a wrinkle on this, which was interesting. So not only the inappropriate medications on admission and discharge, but also that were classified as potential, but also the actual inappropriate medications as defined by a multidisciplinary panel that judged each medication based on the actual patient clinical status. So what they saw, as you can see on the left, is that potentially inappropriate medications were much more common at discharge than they were at admission. So again, these meds are being added and potentially inappropriate once a patient gets discharged, but also that the majority of patients, over 50%, had an actual inappropriate medication, again, as defined by that panel, and the vast majority of those originated in the ICU. So again, we come to like, what do we do about it? I don't necessarily have any concrete solutions, but this meta-analysis was interesting, looking at various interventions. I think the take-home point of this is that really almost any intervention is a good idea, and these interventions were quite different from each other and they varied in what they actually consisted of. Obviously, mostly involved chart review and discussions with providers as to what should be prescribed or what should be discontinued. Potentially, some of them had a role for speaking with the actual patient one-on-one, either pre- or post-discharge, but over patients who had some kind of intervention, had a 21% chance, 21% reduction in their adverse event incidents, and that number increased to 35% when it was a pharmacist-led intervention. And finally, I'll just end with this study, which was recently published, that looked at electronic decision support. I think that's one of the next questions, is how to do this more systematically and across the board, and this study did not find a decrease in adverse events based on their intervention, but what they did see was a consistent and change in the actual prescriptions and the medication that were ordered, and this lasted 30 days post-discharge. So these patients actually maintained their appropriate med list and had adequate deprescribing that was consistent, and there was a follow-up that showed it was effective. What it didn't show is that there were adverse events reduced, and that could probably be due to multiple things, like improvement overall in the control group, and they had other pharmacists or other provider-related interventions in the control group, so that may have lowered the overall adverse effects percentage, but I think we can see an effective intervention with things like this. And so for key takeaways, I think we know that these patients bring pharmacotherapy-related challenges and trying to optimize them through various strategies, dose minimizations, or adjustments that are frequent, patient-specific drug selection and deprescribing should be done as often as possible. Thank you, that's all.

pharmacotherapy

elderly patients

frailty

medication optimization

ICU

polypharmacy