Subscribe
Share
Share
Embed
Conversations in Pulmonary, Critical Care and Sleep Medicine by the American Thoracic Society
Eddie: [00:00:00] Hello, and welcome. You're listening to the ATS Breathe Easy Podcast with me, your host, Dr. Eddie Qian, also the host of the ICUeda Podcast. Each Tuesday, the ATS will welcome guests who will share the latest news in pulmonary critical care and sleep medicine.
Whether you're a patient, patient advocate, or a healthcare professional, the ATS Breathe Easy Podcast is for you. Joining me today is Scott Micek, who will be discussing hospital-acquired and ventilator-associated pneumonias. Scott is a professor of the University of Health Sciences and Pharmacy in St.
Louis, and the clinical pharmacy specialist for the medical ICU at Barnes-Jewish Hospital. Welcome, Scott.
Scott: Eddie, thank you. Nice to be here.
Eddie: I'll, I'll give a, a little quick disclaimer before we begin. As a, as a primary ICU-focused provider, and based off of your titles too, a primary ICU-based pharmacist I will be focusing a lot of my discussions mostly on the ventilator-associated part rather than the hospital-acquired [00:01:00] piece.
But I... A lot of the, the parts of the discussion I think will apply to both realms. So just, just to give that disclaimer up front.
so Scott, l- let's, let's set the stage a little bit first. D- tell me what is ventilator-associated pneumonia? What is hospital-acquired pneumonia? Why, why, why do I care as a clinician?
Scott: Well, Eddie, I think it's one of the most common infections that we face from a healthcare-acquired type of infection, whether it's a skin infection, whether it's an abdominal infection.
Pneumonias are certainly the most common that we see in the medical ICU, and I think that translates across to all ICUs. And not only is it common, but it's associated with a lot of morbidity and mortality. When you're talking about ventilator-associated pneumonia, it could be longer times on the ventilator in these patients that develop these infections.
For patients that aren't on the ventilator, it could lead to that with respiratory compromise. So you're talking a lot of, a lot of morbidity, and depending on [00:02:00] what study you read or what epidemiologic review you read, the mortality can be significant as well, anywhere from 10 to 20%. So, you know, we have our work cut out for us in preventing these infections, but also when we have to treat them.
Eddie: Yeah. Wh- when we're, when we're talking about these different types of infections, you know, infections is such a broad umbrella term. What, what kind of pathogens am I worried about when I'm talking about hospital-acquired and ventilator-associated pneumonias?
Scott: Yeah. I think as clinicians, we all worry about these difficult to treat types of pathogens.
So from, you know, a gram-positive standpoint, I think Staph aureus and particularly methicillin-resistant Staph aureus i- is at the top of our list. But then I think more troubling is sort of gram-negatives in this day and age and all the different types of drug resistances associated with those. So you think about Pseudomonas aeruginosa and just all the different types of resistance it can [00:03:00] have, whether it's to cephalosporins or carbapenems.
And then the Enterobacterales, I think, you know, probably less common than Pseudomonas, but also we'll probably get into different resistances associated with those. And then I think, you know, we deal with these problem pathogens that are collateral damage from the treatment choices we make, treating those Pseudomonas and those Klebsiellas.
I'm talking about things like Stenotrophomonas or Acinetobacter, those more, you know, resistant pathogens with regard to ventilator-associated pneumonia. And I, I think it, you know, we try to fit everything into a bucket, but we have to know where these patients came from before they came, were placed on the ventilator.
Were these patients that were in our ICU through the ED, or were these patients that came from a hospital? Were, were these patients that were transferred in from other hospitals or LTACs? I think it all sort of breeds different types of pathogens and resistance.
Eddie: Yeah. It's really interesting that you bring up like [00:04:00] Stenotrophomonas, Ace- Acinetobacter, and these type of organisms 'cause I, I'll find clinically, like I'll recover them from my, my sputum cultures and otherwise, and sometimes we decide to treat, sometimes we say, "Maybe this is just a colonizer."
What, what kind of things push you one way or another when you're trying to make that call?
Scott: Yeah. I, I, I think especially in this era of rapid diagnostics where you get such sensitive information and delineating whether this is a colonizer versus a true pathogen is, is sometimes difficult And, you know, I think it is the clinical condition of the patient.
Are, are they i- is there some level of decompensation associated with that respiratory specimen? Obviously, something's triggered the clinician to obtain those specimens. But, you know, in the big picture, is, is this more of an oxygen requirement? Is there a septic shock component to this? I think that points you down the pathway of, yeah, this is real.
We need to treat this [00:05:00] Stenotrophomonas acetetabactor, right? Versus, okay, this patient is, you know, somewhat stable. They're-- they've been on a ventilator for, you know, a week plus, or maybe they have a tracheostomy at this point, and they have a fever. But there's really nothing else that, you know, is pointing towards decompensation.
That's where you get stuck with it, right? And I, I think that's always the conundrum that we face in, in deciding are we, are we gonna put on empiric broad-spectrum agents that could target some of these more resistant pathogens, or based on their microbiologic history that we have, do we, do we target those?
So, yeah, those are difficult choices. But I, I think to summarize it, it really is, you know, thinking hard about where is the patient in their, you know, sickness journey, so to speak.
Eddie: Yeah. No, that's really interesting. It's really interesting you bring up, you know, talking about empiric antibiotics. I feel like a lot of, a lot of clinicians are familiar with how to treat Staph aureus and [00:06:00] Pseudomonas and otherwise.
W-what, what are your recommendations for empiric treatment considerations for, like, before maybe before I have some of this information or data back, empiric antibiotics for treating what I presume is a ventilator-associated pneumonia. Are you starting off and reaching for the, the Bactrim or the fluoroquinolone to treat Steno-trophomonas right off the bat or something else?
So can you talk me through how you think about that?
Scott: Yeah, I think there's a couple things that we take into consideration before we get the, the microbiologic reports back. I think, one, what is their microbiologic history? How distant is that history? So obviously, the more recent stuff i- is more significant in the decision-making.
You know, what could they be colonized with, like you had mentioned, could now be a pathogen. I'd say... So I think that helps with empiric decision-making. When it comes to things like Stenotrophomonas, what is their antibiotic history recently? [00:07:00] You know, if, if patients have been on carbapenems I, I think that increases your risk of Stenotrophomonas, amongst other things, but I think that's an important one in many of these patients.
And, and then I think the third thing, you know, for us, before we get the results of that, is what is the patient host like? Are they immunocompromised? You know have they received immunologics or different biologics? Are they on chemotherapy even? Those types of things, while the literature might not necessarily say you're a higher propensity, I think using your local antibiogram and knowing what sort of that pathogens are in those particular patients, if you have that available, I think is important, too.
Eddie: Yeah. That's, that's, that's really interesting and great that you bring it up. Y- you know, when I look at, you know, resources up to date, otherwise, than in talking about the empiric treatment for [00:08:00] venil- ventilator-associated pneumonia, I can't help but realize s- a lot of places recommend, you know, we've ta- talking a lot about these gram-negatives, double covering your gram-negatives.
Now, I'll tell you u- up front, that's not my practice. I'm, I'm curious what your practice is over at Barnes-Jewish. But I know that different providers across the country and across the world have different views on that. W- what, what are your kind of ta- what takes on that, views on that? When, when or when do you not cover, double cover your gram-negatives empirically right off the bat?
Scott: Yeah, I think that's been a, a question that we've tried to look at from an observational study standpoint over the last couple of decades and because guidelines do promote this. But our, our general practice for, I would say, the, the, you know, seventy-five percenters, those patients that, you know, aren't totally complex, but it's a, it's a ventilator-associated pneumonia that you know their history.
You've been part of that treatment team all along. [00:09:00] We'll use monotherapy upfront, you know, it's any of those different antibiotics. We're, we're a cefepime-based hospital. I know others are piperacillin-tazobactam. Cefepime would be our upfront choice if they, if they don't have those histories that we had mentioned.
Where we will use combinations of therapy, whether it is say, an aminoglycoside in these patients, is particularly when they're also presenting with septic shock or there's new vasopressor requirement. It, it may be severe sepsis where they need vasopressor, you know, for a few hours. I think that points us in that direction to use tobramycin in our hospital.
We may use a combination, like you said, targeting Stenotrophomonas. Maybe it's minocycline, may- maybe it's levofloxacin. Something along those lines has, has traditionally been where we go to i- in those realms. And then, and then maybe, maybe they've received [00:10:00] those before, and this is a second episode of this, or maybe that's where you go into the carbapenem monotherapy upfront.
So that, that's sort of how we think about the combinations of therapy. I think, you know, we've studied it, not necessarily in ventilator-associated pneumonia but in those more complex patients where you may have a, a, a se-- a bloodstream infection along with that, a concomitant bacteremia where aminoglycosides may be of some benefit.
Eddie: Yeah. That's really interesting, and thanks for kinda talking me through all that. W- w-- Moving, I think-- So we talked a little bit about the treatment. There are two things that border the ventilator-associated pneumonia pathway when you, like, look at some of the recent literature and, and how to treat, how to manage.
You mentioned one of them. You s- mentioned rapid diagnostics. There's been more and more papers talking about rapid diagnostics, their utility, how could they help, how are they maybe not helpful. And then also, there's been a lot of data about de-escalation of [00:11:00] antibiotics. And n- and now this is not necessarily just in the pneumonia space, but, you know, empiric broad-spectrum antibiotics, very common in the ICU.
But w- when do you stop? How do you stop? Where do you go to? Do you have any comments on, on either the rapid diagnostics or de-escalation or both, to kind of taking you to both sides there?
Scott: Yeah, I, I, I think the beauty of the rapid diagnostics, you know, as someone who's trained at a center that sort of was at the crux to broad-spectrum antibiotics up front and then de-escalation, you know, I've been fortunate to be part of MICU teams with Maren Khalaf that, you know, sort of coined this approach and, and really as, as it's evolved over the years.
I think rapid diagnostics adds just another great layer to this from the standpoint that, You know really early if you're missing something in your coverage. And I think the, the more important part of that is, are there potential resistance [00:12:00] genes that are present in this patient? Can you identify those?
Gi- just give you an example. I was rounding today, and we had a patient on a respiratory specimen. Actually, it was a blood specimen as well that quickly identified as Klebsiella pneumonia and identified that it was an ESBL producer. You know, five years ago when we were doing this, this patient would've been on cefepime and probably been on cefepime for twelve to twenty-four, maybe longer, hours.
Now we're able to make that transition very quickly, and that's probably more appropriate. And that may have it, an influence on this patient's outcome, the timing of it. Probably more so if they were septic than anything. I think problem with rapid diagnostics, as I alluded to earlier, was how sensitive they are.
So not only would you get that Klebsiella that has ESBL, but what if you also have Acinetobacter with it, then group A strep and Haemophilus? You know, you're, you're probably not worried about the, the group A strep or the [00:13:00] Haemophilus in your coverage, but what about the Acinetobacter? And do I cover all these things, and do I broaden it out?
And that makes it very difficult. And then what if those patients have no growth on their, you know, culture and no susceptibility? So I think it's almost a double-edged sword with this. I think it can really help in terms of upfront if you, if you need to use a carbapenem or even some of these more novel antibiotics upfront empirically.
But then you can get into trouble if, you know, what if this is polymicrobial, and what am I really covering then? What if we don't identify anything on the respiratory specimen or the blood cultures and growth to get susceptibility? So both ways on that.
Eddie: it really is so, it really is so hard to clinici-- we as clinicians really need to truly understand, like, what, what's included in the pathogen panel, what's, like, what's being probed for and what's not.
And I feel like a lot of that, at least for, for us, is not super transparent. So we'll be in the situation, as you described, where the probe will tell [00:14:00] us one thing or another, and we're like, "Okay, we got it." And it's not common, but it does happen that the culture then comes back with something else. And for-fortunately, many of the times, the antibiotic choice we chose covers that even if we didn't know, but we're only a hop, skip, and a jump away from, from not from a situation where that's not the case.
Scott: Yeah. And I think the specimen that you obtain when we're talking about ventilator-associated pneumonia is important in this too. So say you just get a tracheal aspirate, which is generally recommended, right? And that, you know, you might get more of that proximal airway you know, bacteria, as opposed to do-- if, if you do a, a bronchoalveolar lavage and you you know, get rid of that wa- maybe it is a little bit more sensitive and specific as well.
So I, I think that, that has something to do with it as well. We, i-in the MICU that I practice in, it, it's variable based on attending physician if they're doing the tracheal aspirate versus the BAL, and it'd be a nice thing for us to look [00:15:00] at and compare. We haven't done that yet.
Eddie: Yeah. Sounds like, sounds like a great a great comparison to do, and then potentially a future, a future evaluation and trial.
I think there are some groups that are looking at that question actually-
Scott: Mm-hmm ...
Eddie: Interestingly enough. we've talked a little bit around this, but there's a... We've talked about gram-negative organisms. There's a categorization within them that's very microbiological. Non-fer-- lactose-fermenting gram-negative RAs include Pseudomonas, Acinetobacter, and some of these stickier bugs that we've talked about.
How do you think about-- do you think about those bugs, if you recover those either in your probe or your culture, differently as you move forward through their course, not just their initial antibiotic selection?
Scott: Yeah, I th- I think you have to think about how long you're treating these patients.
Certainly, that's important, and I've never been a person to recommend, "Okay, we're gonna treat for a week. Let's put a stop date on this order, and, and that's gonna be it from a stewardship standpoint." Or, you know, we gotta treat all of these patients for [00:16:00] two weeks necessarily. I think you have to let the patient tell you, you know, in, in some capacity what is the right duration.
I, I think the literature, you know, whether it's the SHASTRA study going back almost 25 years now or the more recent evaluation that unfortunately, because of enrollment, didn't en- en- enroll as many, but looking at Pseudomonas in particular, look as many patients as they had designed, I should say.
Looked at Pseudomonas. But I think the important thing, probably no difference in, in mortality, but the risk of recurrence or prolonged infection with shorter courses, there seems to be, you know, some consistency between those studies there. So I always, you know, in, in those particular pathogens that you're mentioning, particular Pseudomonas you know, seven is, is sort of the, the very minimum.
But let, let's think about where is this patient, you know? If, if this is a Pseudomonas and this patient [00:17:00] is extubated in, in two days after we recover it, and trending on that trajectory, they probably need a shorter course, right? Than someone who remains intubated or goes to a tracheostomy or something along those lines that's where recurrence is gonna happen most likely anyway.
So they, they deserve, based on the literature what we know, a longer course because we wanna prevent that subsequent morbidity from happening.
Eddie: Yeah, that's really, that's really interesting and kind of talking about individualizing courses. There are a lot of recent trials and papers looking at procalcitonin, CRP, and some, some of these other types of inflammatory infectious markers.
What, what are your, what are your views on that? How do you use them, or why do you not use them?
Scott: Well, I think w-- at Barnes-Jewish, we've never been a procalcitonin practicing hospital. We can, we can draw it, and I think part of that is because of our interdisciplinary teams, including pharmacists and all the pr-- where we're [00:18:00] sort of discussing this on a daily basis.
But I think other tools like CRP or y-you know, those types of things, a-any additional data that you can get that sort of projects patient getting better or not. And, and, you know, we're still believers in VAP. I think the, the literature goes back, and particularly with acute lung injury, looking at PaO2 to FiO2 ratios or, you know, we don't get as many blood gases anymore, so SPF or SP ratios to O2 ratios, those types of things.
What is the trajectory of this patient along those lines? So you know, I, I think that helps inform the decision-making that we do. But you know, we don't, we don't draw serial CRPs. I know there's been literature in that necessarily, but i-it's more of a interdisciplinary discussion and really a commitment to minimizing the duration of antibiotics to the shortest course that's-- the patient is telling us is [00:19:00] possible.
Eddie: Yeah. I, you know, you know, we, we also aren't a procalcitonin institution using c-serial CRPs or otherwise. I think the way that I've always interpreted that literature, and I'm curious to get your thoughts on this, is that it's, it's, I think, helpful to raise the floor. And I think both of us practice at places that are, you know, thankfully we're very-- have, have a lot of support, well-supported, lots of staff, a lot of people thinking about a lot of systems to be kind of watch, watching antibiotic stewardship.
But if you were at a place that didn't have all the different types of support, then maybe it's something that could raise the floor so to speak.
Scott: Yeah, I agree with that. And particularly if you combine it with clinical decision support where, you know, those values, you're getting some help interpreting those things because just drawing labs and, you know, you know, you get the lab overload in, in these patients.
And so using that, you know, almost AI model to help you as well, I think in, in those hospitals, whether it's through their you know, it's probably through their EHR [00:20:00] system that those things can be built in.
Eddie: Well, the, the listeners of the show will know that I've, I've spent a lot of time talking about those two letters that you spoke.
And I'm just gonna say that I didn't say it. And I didn't- ... I'm just gonna say that at the front. Th- this is, this has been a, a really, a really great discussion for me, and I've, I've learned a lot here. But I'm curious to your, I guess, if to close out, y- what are your thoughts on, on the future of this?
So we, we, we know we have some rapid diagnostics. We know how to empirically treat. We talked about some de-escalations. But what, what are some unmet needs, areas of, you know, clinical uncertainty in treating hospital-acquired and ventilator-associated pneumonias? Is there any difference there between talking about your non-fermenting gram lactose fermenting gram-negative rods or anything from that perspective?
W- w- where do you see the future? What do we need to do more? Or is there something that you are doing at Barnes-Jewish that that's coming down the pipeline?
Scott: Well, I think [00:21:00] a couple things in terms of nosocomial pneumonia, and we didn't talk a lot about hospital-acquired pneumonia or, or non-ventilated HAP.
But I think doing a great job to prevent that, I think is important. There's not much literature on HAP, right? It's all extrapolated from VAP. But we know it happens, and we know it's associated with morbidity. So I think, you know, getting a better grasp on not only-- I think epidemiologically, we, we sort of have a good handle on it, but, you know, maybe it's quality improvement efforts to prevent it.
And so much resource has been put into preventing VAP, but maybe it's a little bit of that. From a pharmacist perspective, you know, I, I think it's important to think about how we're dosing these antibiotics whether, you know, the, it-it's a critically ill patient where there's pharmacokinetic, pharmacodynamic alterations, you know, [00:22:00] they're profound.
Getting tools to help us dose. You know, is therapeutic drug monitoring going to become standardized? I know in some countries it is. We don't, we don't do much of it. Is that one snapshot of therapeutic drug monitoring an adequate picture to know? You know, there's literature to talk about, you know, do, do we give toxic doses?
You know, cefepime neurotoxicity, I think, you know, is one that you guys probably talk about a lot. But is that true, and is that a dosing manifestation as opposed to anything else? So are prolonged infusions needed in all these patients? I think, I think we're at the cusp. We sort of, you know, it's kinda like the eight versus fifteen.
We kinda make it those two buckets, and then we- we do all these other things in much the same way. You, you do this for everybody, but there's-- we could be harming patients in that regard, too. So I think we can get a little more precise. Even though beta-lactams, which we've primarily inferred to here, have a wide [00:23:00] therapeutic window, I think we can get better at that, too, particularly in these patients.
And maybe that leads to shorter courses and and less resistance development and those types of things.
Eddie: Oh, you, you definitely tou- touched on a core that is one of my kind of like my most common topics that I kind of like, u- un- unknown topics in the care of our patients is for our patients on continuous renal replacement therapy.
Do, do we really know the doses of the antibiotics? I know we have, we have dosing for that, but do we really know what we're doing for our really, our really sick patients who have varying renal needs or otherwise? I, I, I'm, I'm curi- You brought up, you know, there's, there's some data about extended infusions and continuous infusions of these antibiotics.
You also brought up therapeutic drug monitoring. which direction do you think we should head when you're talking about those things?
Scott: I think it's the same thing that we just alluded to.
I think it has to become personalized. You know, a patient who has an [00:24:00] estimated GFR of 25, which is many of our patients probably don't need a prolonged infusion of a beta-lactam. I know you like consistency from a nursing standpoint so that you don't make errors when you need to do it but, and it's-- but you probably don't.
But that might be a patient who you need to do therapeutic drug monitoring in particularly those whose estimated GFR is going from 25 to 45 or the other way around. And you wanna be quicker at either giving the right dose to prevent toxicity or to, improve that PK/PD risk. So I, I think those things are things that we gotta get better at.
And I think that, you know, like we have rapid diagnostics, that's a technology that's improved. I th- we'll probably have better technologies in real time of estimating GFR, and maybe we can correspond that with therapeutic drug monitoring at the same time. I, I, I think those, tho- probably aren't that far off.
And [00:25:00] then, then you, you mentioned with- within the dialysis circuits, is there some way to assay, you know, what are our concentrations that are being... You know, those types of things. I think it's kinda primitive how we do it, but I think we'll get better at it.
Eddie: Yeah. No. I, I, I completely agree. It's something that's always in the back of my mind.
It's, like, something that I, I currently, like, don't have the ability to look at further, but it's, it truly is always in the back of my mind. And so you-- so some of the things that you touched on here has really struck a chord. But th- thank you, Scott, for your time here. This was a really fascinating discussion.
I've, I certainly learned a lot. Thank you to everybody listening for joining us for today's ATS Breathe Easy episode. Please subscribe and share this episode with your colleagues. We, at the time of the recording, we had just finished the ATS International Conference. I had a lot of, I had a lot of fun.
Scott, I don't know if you were able to join us. But we'll be look, be on the lookout for further correspondence 'cause as we get ready for next year in New Orleans. So thank you, everybody.