Good morning, SCCM members. We are pleased today to present the latest advances in diagnosing sepsis and pneumonia. We have a great panel of presenters today. I'm pleased to introduce my colleagues, Dr. Ed Septimus, a professor of internal medicine at Texas A&M College of Medicine, and he's located in Houston, Texas. Dr. Trevor Van Schoenenbeld, an associate professor of infectious diseases, the medical director of the Antimicrobial Stewardship Program at the University of Nebraska College of Medicine in Omaha, Nebraska. And I am Debbie Goff, a professor of pharmacy practice at The Ohio State University Wexner Medical Center and infectious disease specialist. So we have a great discussion today and our agenda includes the following, and as you can see, we're going to present cases both on the management of sepsis and pneumonia, and we have plenty of time for discussion and Q&A, which we certainly hope you have questions for us. So let's get started in this topic. Antibiotic stewardship during a global pandemic. What did the world do on getting the right diagnosis with the right drug, the right dose, and the right duration? Well, in my opinion, and I think as many would agree, stewardship policies and guidelines at the beginning of COVID simply flew out the window. We've had multiple studies now identifying about 72% received antibiotics, however, only 8% had a bacterial or a fungal co-infection. Just a new paper in Open Forum Infectious Diseases, a more recent one, showed 69% received antibiotics and only 1.2% actually had a bacterial co-infection. So we clearly can do better. So what did making the diagnosis of COVID-19 teach us? Well, we will continue to over-prescribe antibiotics when we can't make a rapid diagnosis. And that was the challenge. You know, it was such an emotional response as these patients were flooding our ICUs in critical condition, and you felt the need to do something. And early on, when we really didn't know anything about COVID, that something was antibiotics. One, two, sometimes three, and multiple studies, as you can see some of those listed there, confirmed we over-prescribe and over-utilize for many days. I think we're doing better, but that paper just published in Open Forum Infectious Diseases showed us we're still over-prescribing antibiotics in COVID patients. And what I really enjoy about this paper that was published by a collaborative of experts on antibiotic stewardship in the ICU, the lead author, Rich Wondering, is an ICU physician at Northwestern Hospital in Chicago. And I'm very proud one of my colleagues, Dr. Tony Gerlach, who is our surgical ICU pharmacist, is also a co-author. This paper truly represents a collaboration of U.S. physicians, clinical pharmacists, CDC physicians, and European physicians asking about stewardship in the ICU. And I think this really drives home some of the feelings during COVID. So when you look at antibiotic stewardship in an ICU, why do we over-prescribe? And the paper is written sort of in a question and answer. Why do we give antibiotics that ID people and stewardship will tell us later, well, that actually wasn't necessary? We do it out of fear. It's not out of a knowledge deficit. It is out of fear, that fear that we're going to be missing an organism. And the fear that I want to do something. And the idea that antibiotics are fairly benign, right? I would rather give them. And that's what drives all this excess antibiotic use. But when you look at their own literature, this published in the Critical Care Journal, it shows 30 to 60% of antibiotics prescribed in the ICU are either unnecessary, they're inappropriate, or they're suboptimal. They're not dosed properly. So there's a lot of work. And then the adverse effects of antibiotics. You know, the concept always what I hear a lot from surgeons, you know, well, just in case, Debbie, they're infected. That's why I prescribed this. I didn't get any cultures and I really don't know what I'm treating. But that fever they spiked last night, just in case, it can't hurt, right? Well, that's what we thought for decades, except we now know that isn't true either. One in five patients are actually harmed by giving antibiotics. So this was a really nice study from Johns Hopkins. 27% received antibiotics and 20% developed at least one serious adverse event. And 19% of the antibiotic regimens were actually not clinically indicated. So the most common adverse events are GI, renal, and hematologic. And then antibiotic stewardship in the ICU, really, to work, you have to empower the staff to speak up and ask about antibiotic therapy. So if you have a surgical resident, and this happens over and over, why did you start these antibiotics? I was told by the attending, start antibiotics. That's not acceptable in 2021. You have to understand why and you have to know what you're treating. So that hierarchy that sometimes occurs in the hospital setting, and I work globally, that hierarchy occurs all the time. You have to be able to empower the nurse to ask, doctor, I don't really understand why you're starting these antibiotics. You're supposed to get a culture result or a culture. So anybody should be able to question why, because that's what makes collaborative care work. And that's what makes it best. So individualization, you know, guidelines tend to reflect the bias of large academic medical centers. That's where most of the guidelines, the authors are from. And so smaller hospitals might have a different perspective. They might not have as much antibiotic resistance. They might not see as much canademia. So, you know, the application of some of the things in the guidelines do need to be individualized to the hospital that you work in. All right. So let's start now with our very first question for the audience. My hospital has rapid diagnostic tests for both bloodstream infections and respiratory infections. So do you have both available in your hospital? Yes, no, or you're not really sure. We'll give you a couple seconds here to answer that. And that will help us as we continue on in the presentation. Now, if we were live, we would probably have some nice music streaming at this moment. But since we're doing this remotely. All right. And so most said 41% yes, 38% no, I'm sorry, and 21% not sure. So that's important, that not sure, because that shows a lack of communication from the microbiology lab and the stewardship and the clinicians that are frontline providers. So clearly, there's more work ahead. Thank you. And we look at this slide from Dr. Kumar from several years ago about appropriate therapy in the management of sepsis is the key to early survival and how every hour counts in the management of patients with septic shock. And the further out you go, the lower the survival rate. So it really drives that urgent need to get appropriate antibiotic therapy administered rapidly. So I looked at this newer study on antibiotic resistant pathogens, looking at the prevalence of antibiotic resistant pathogens in culture proven sepsis, and looking at outcomes associated with inadequate and broad spectrum empiric antibiotic use, because those are really the two challenges we face. So they really wanted to determine the epidemiology of antibiotic resistant pathogens and outcomes in both cultures. And what they found is both really provide poor outcome. If you give inadequate therapy, no surprise, 20 to 40% higher odds of death, but also adequate therapy that was excessive. So broad spectrum with a carbapenem, when you could have used stevia, you could have used a lot of stevia. But if you give inadequate therapy, you're going to end up with a lot of sepsis. And so what they found is both really provide poor outcome. Carbapenem, when you could have used ceftriaxone, was also associated with a 20% increased odds of death. So really honing in that correct therapy is important. So what occurred with that unnecessary broad spectrum were the adverse events. 26% had an increased risk of C. diff or acute kidney injury. So we know the disruption of gut microbiome is increasing in science of showing that there's a negative effect on the immune system and outcomes in sepsis. So we need a tool. We know all the consequences. If we only knew what we were treating, we could prescribe the right antibiotics. So what is the ideal tool? It produces the results rapidly. We can use readily available clinical specimens. It detects a broad range of pathogens. It's easily accessible, highly sensitive, highly specific, portable, integrates with our workflow into Epic. And then the last one, it's affordable. And that becomes the barrier in many hospitals. But if you know the value of the testing and show it, then you can cost justify it, just like we do with the newer expensive antibiotics. So with that, we're going to turn now to the need for new diagnostic methods. And Dr. Ed Septimus is gonna be up next and talking on bloodstream infections. Welcome Dr. Septimus. Well, thank Deb. It's, I feel like the middle child between two outstanding colleagues. So let's go on and we'll show you, I'm gonna start off with a case. A case from several years ago in February, 32 year old gentleman came to our ED with a four to five day history of a dry cough, fever, myalgias. And then one day prior to admission, he got a much more productive cough and he developed shortness of breath. Otherwise, he's a healthy gentleman. In the ED, he had minimal decrease in blood pressure. He was tachycardic, he was tachypneic. His O2 sat on room air was only 88% and he was febrile. His exam showed tubular breast sounds and crackles at the left base. I do teach the house staff to examine the patient. There was no murmur, the abdomen was soft and the extremities and neurological were normal. His Y count was 8,400, but he had over 15% bands. His creatinine was elevated at 1.7. His bicarb was 21. His lactate was elevated at 3.4. His procalcitonin level was 2.5 and his coag studies were normal. His chest texture, which I'll show you in a minute, showed as you might expect from the exam, a left lower lobe consolidation. In the ED, unfortunately, he became more hypoxic despite supplemental oxygen. And he required intubation and then a rapid test was performed. So there's the chest x-ray, which I think you can see does show a very nice left lower lobe consolidation. So in the ED, he was given two liters of Ringer's lactate in the first two hours. Per order sets, he was given ceftriaxone and azithromycin in less than three hours after presentation. His blood pressure improved actually after one hour, but as I mentioned, he was still hypoxic and required intubation. And then as I mentioned, a rapid diagnostic test was performed, which was very helpful in the management of this patient. So here's my first question, which I hope you'll all answer. What rapid diagnostic test was actually performed? A gram stain, which is over a hundred years old, a PCR for a specific pathogen, a multiplex panel, which we'll discuss further during this symposium, or perhaps you're unsure. So let's get your sense as to which one of these rapid diagnostic tests was performed in this patient, which I will tell you was impactful in terms of the patient's outcome. So it's again, as Debbie said, we wish we had some nice background music, but this is the real world. This actually happened on a Sunday afternoon in February when I was called in to see the patient. So I think we're getting towards the end of our question time. And here we go. The majority, or I guess plurality of you, chose a multiplex panel. And I know that Trevor's going to talk a lot about this. Some of you said a specific, and less than 20% said a gram stain. Well, I'm sort of an old fashioned doctor. The first thing I did before I even saw the patient was go to microbiology lab. They get very nervous when I show up. On a Sunday afternoon. And I said, I'm going to look at the gram stain of this patient's tracheal aspirate. And as you can see from this particular slide, there's gram positive cocci in clusters, which was very suggestive, of course, of staph aureus. And if you think about it, we started the patient on ceftriaxone and azithromycin. And the question, and Trevor is going to talk about this a little bit more in detail. What is the incidence of MRSA in your community? And should this patient be started or added another antimicrobial agent? So let's kind of look at the initial hospital course. The patient, of course, was transferred to the intensive care unit. We were able to oxygenate him quite easily. We stabilized his blood pressure without pressors. And then at 13 hours, literally from presentation, which is actually less than 12 hours since the blood cultures were drawn, our microbiology lab called us that there were gram positive cocci in two out of two blood cultures. So the old traditional scenario is that we'd get fluid or tissue. We did the gram stain, which by the way is a rapid diagnostic test, which is very old fashioned, but can be very useful as it was in this patient. You then have to put it on culture media. It takes 24 hours or 12 hours, depending upon how fast it grows in the microbiology lab. The microbiologist looks at the plate. This may look like hemolytic staph on the plate. They may have been able to do certain tests to see if it was coag negative or coag positive, in which case it would be positive, so it's staph aureus. And then they do susceptibility tests. And so within another 24 to 48 hours, susceptibilities are done and it would have turned out in this particular patient, the patient in fact had MRSA, which was not covered by the initial antibiotics, which were the typical order sets, which are frequently used in many of our emergency departments. Now, the other scenario is what technology are now available? We have PCRs for staph aureus and MRSA. We have multiplex PCRs, which Trevor will talk about in much more detail. We have multi, which is a wonderful test. We have, so multi there's, we'll talk about in a minute. Some of these are culture independent and some of these are culture dependent, meaning that the cultures have to be grown first before we can use the rapid technology. Many of the multiplex, especially the upper respiratory and lower respiratory panel are culture independent. In this case, we have multi in our lab, which we were using. So that first went on multi, we found that it was a staph aureus, but of course multi doesn't do susceptibility testing or look for resistance genes, but our multiplex PCR does, and it did detect within 16 hours of that presentation versus 48 hours, this is what we got back in our electronic health record. So this is, we use varigene, it identified it as staph aureus and it did detect the MEK-A gene. So we knew within 24 hours of this patient at MRSA would traditionally, it might've been 48 hours. So we were able to actually from the gram stain was actually initially able to add vancomycin. And then we knew within 16 hours that vancomycin needed to be continued. Now there's a whole variety of tests. Just in this slide, this briefly reviewed multi versus multiplex. Multi is fast, it's accurate. Per test, it's very, very inexpensive. The high cost of multi is actually the platform to run the multi. But once you've done that, the cost per test is extremely low and it's really not that technically complex to perform. However, as I mentioned, it is the platform is more expensive. It doesn't identify resistance genes and sometimes it may require a culture of the organism. It does require a culture of the organism to actually put it on the machine. Although some people are doing it directly from positive blood cultures without having to grow the organism first. Now the multiplex can be either culture independent or culture dependent. So in many of the blood cultures, again, that usually is based after the organism begins to grow in blood culture pool, we can use it. But many of the other syndromic panels such as the GI panel, the meningitis panel and the respiratory panel, which Trevor will talk about later, are culture independent. These are really cool tests. It's a closed system. It can identify organisms very quickly with multiple resistance genes. It does require trained personnel. The cost per test is higher than it is for multi because you're testing for multiple pathogens. So both of these really can work hand in glove with each other and are very difficult to identify organism. Multi is extremely helpful. And for those of you who've had experience with Candida auris, which is frequently misidentified, multi is very, very useful. Now, the ones we use for blood cultures, which was the case in this patient, there's two panels that are primarily used across the United States and the world. One is the FilmArray panel, which detects 19 bacterial targets, three resistance genes and five yeast targets. The Vera gene, which is the one that we use at our place can detect 12 bacterial gram positive pathogens, three resistance markers for gram negatives, nine bacterial targets and six resistance markers. And you can see it's highly accurate. Now, these are the FilmArray panels. I'm not going to go through this in detail. One of the big difference between that and Vera gene, the FilmArray only detects three resistance markers here, as you can see, but it also gives you five yeast topics. So depending on what kind of an institution you have and how big an immunosuppressed group of patients, this can become very useful. The Vera gene, as you can see, has pretty much the same number of bacterial targets, but they detect a much broader array of resistance genes, primarily against the gram negatives. Now, another panel, which you do directly from blood, is called the T2 bacteria Candida panels. You can see the panel, you set that directly from blood, and within three to six hours, if these pathogens, this is so-called escape pathogens, and here are the Candida pathogens, from whole blood, they can identify this within three to five hours. So this is culture independent. Now, for the rest of my time, let's talk about the sepsis bundle and how that, first of all, impacts this particular patient and how it may impact appropriate and potentially overuse of antibiotics. So this was proposed in 2018, this one editor panel, and they gave this, the bundle elements on the left, the grade of evidence on the right. You can see lactate was weak, obviously drawing blood cultures to identify a pathogen. We think it's very important, those of us involved in stewardship, to move from empiric therapy to more targeted therapy to do that. We have to be able to identify the pathogen, and that's why certain culture techniques and molecular techniques are going to really enhance our ability to do more targeted therapy. That's sort of a best practice statement. Administration of broad-spectrum antibiotics, it's a strong recommendation, but as you can see, and I'll show you how to divide this a little bit better, it's got a moderate quality of evidence, and Deb's already showed you one study by Kumar, which I'm going to come back to in a moment. Rapid administration, they put a fixed amount of crystalloid for hypotensive patients or patients whose lactates are greater than equal to four. Again, it's a strong recommendation, but it's got a low quality of evidence. And of course, adding vasopressor, if they mean persistently hypotensive, you know, again, that's a pretty strong recommendation. Again, it's got a moderate degree of evidence. So some of the criticisms of this one-hour bundle are listed here. Number one, a broad-spectrum antibiotic is necessary in all patients with presumed sepsis. Debbie showed you that very nice study from Chana Rhee's group to show that if you over-treat in individuals that don't have a potential for resistant organisms, there can be harm. What about dosing strategies for seriously ill patients? Do we optimize the therapy we give? And what's the appropriate duration? We're getting better at that, but there's still more to learn. Is the antibiotic regimen active against the offending pathogen? Obviously, that's very important. You've got to use your own local epidemiology to help target this in a more effective way. The difficulties from the ED physicians, having spoke to so many of them, is that, you know, to make this thing within one hour, sometimes people who prevent with vague, non-specific symptoms, it's pretty hard to arrive at a diagnosis of sepsis within one hour of arrival to their emergency department. That leads to unfortunate adverse consequences associated with overuse. And then the problem of equating sepsis without septic shock versus those that have septic shock are all potential criticisms of the one hour bundle. So let me go through some inconvenient truths. Number one, there's a high level of over-diagnosis. Even those individuals who you think you have sepsis who get into the ICU, only 60 plus percent of the time is the diagnosis correct after further study. We know that that one hour bundle has promoted excessive antibiotic use and the unintended consequences of C. diff, acute kidney injury. Other thing that we see, the rush to hang that first dose sometimes before appropriate cultures are obtained. And when that happens, our ability to do target antimicrobial therapy goes out the window. We certainly can select for MDROs. And then that microbiome dysbiosis is so very important. And I've already alluded to this one paper of the 40% admitted to the ICU with the admitting diagnosis of sepsis, probably didn't have a bacterial infection. And that cultures were only positive in a minority of cases. And I can tell you, and that includes me to be absolutely honest, when there's diagnostic uncertainty, sometimes you have a low threshold and a very, very ill patient to continue antibiotics. The time to antibiotic probably matters for septic shock, but it's not as obvious for those without shock. And I'll show you the evidence for that. Two recent studies have confirmed delay in antimicrobial administration and mortality in septic shock. It was literally low association of those without shock. And I'll show you those studies in a minute. There was a study from overseas looking at, well, if earlier is better, let's give it in the ambulance. They did that. Most of these patients, by the way, did not have septic shock. They had sepsis, but it showed that there was no difference in outcome between the group that got the antibiotic in the ambulance versus those that got it after arrival to the hospital. The other thing is these rigid mandates hinder clinical judgment. There's that rush to say, listen, I'm going to get judged by this. It's on my dashboard. I've got to look good on my dashboard. So those rigid mandates really don't give me enough time to use clinical judgment. And I think that's most unfortunate. And the studies have demonstrated that recent antibiotic exposure actually increases our risk of sepsis by altering the microbiome. So I think all this is very important. So this is the Kumar study that was shown by Debbie earlier. And it is true that every hour delay is increased with an increase or decreased survival rate. But I want you to know the definition here. These patients had septic shock. These were not all patients. The next study is a beautiful study from New York State Mandate that Seymour and Associates published. And they looked at the association of every hour delay of broad spectrum antibiotics and death in a fairly large cohort. And what they showed was that each hour delay in bundle completion was associated with a 4% increased odds of death driven primarily where they found delay in antimicrobial administration. Well, let's take a little deeper dive into that and look at the difference between those that required vasopressors, which is a marker for septic shock, versus those that did not. And it's quite clear that the ones who benefited were the ones that had septic shock, not the ones that had sepsis without shock. And then the Kaiser group did a very similar analysis looking at absolute mortality with delays in antimicrobial administration. Once again, the ones that clearly benefited were the ones that came in with septic shock, not those with the old definition of severe sepsis or sepsis. So as Debbie said, what's the harm of a few doses? Well, let's take a look at a couple of studies. First of all, we're not very good at deescalation, I'm sorry to say. So this was a retrospective group of almost 250 patients, given what I consider nothancomycin, vancomycin and piperacil and tasobactam, a very frequent administration in the ED. Only 63% had antibiotic regimens deescalated by 72 hours and only 72% at 96 hours. More importantly, there were no difference in patients with positive versus negative cultures and in patients who had a positive culture. So there we can do targeted, targeted deescalation. They were only deescalated in 63% of cases and antibiotics started in the ED are frequently continued after the patient gets in the hospital. Well, here's another one that looks at surgical prophylaxis that shows you that every day and every dose matters. Look at acute kidney injury and the incidence of C. diff as we continue antimicrobial prophylaxis beyond 48 hours. Think of that again, the patient being treated for infection. It's the same concept. Every day, every dose matters. What about antimicrobial resistance? This is a huge retrospective study of over 7,000 patients with severe sepsis and septic shock. On the left-hand side, we had 6% develop new resistance to the anti-pseudomonas beta-lactam within 180 days and the median was only 17 days and for each additional day of antimicrobial exposure, it was associated with a 4% increased risk of developing resistance and this was statistically significant. So I want you to think about that just for a moment. Oh, we'll just give the patient an extra day or two of antibiotics to be sure. It's probably not going to hurt. I think, I hope we've convinced you and Debbie has convinced you that it does and then we have the CEP1 core measure and I'll wrap things up and turn it over to Trevor. You all know it came into effect in 2015. It requires a three and six hour bundle of care that many hospitals around the country are now devoting substantial resources to meeting and abstracting this measure because it has to be done manually. Well, you all are very familiar with the CEP1 measure, lactate, blood cultures, administration of antibiotics and fluids if they're hypotensive and then at six hours, if the lactate's abnormal, you've got to repeat it. A vasopressor if they're still hypotensive and then some reassessment of volume status and tissue perfusion. So let's kind of come back and talk about our patient. So when the blood cultures were reported, and even before that, when I saw the gram stain, we gave him a dose of ankylomycin, and then when we got the varigene back, we discontinued the azithromycin and the ceftriaxone. The next morning his Y count rebounded very nicely to 21,000, it was actually suppressed because he was so ill. His creatinine returned to normal, his lactate now is less than two, and there was an influenza A, because this patient was during influenza season, in fact, this was a post-influenza pneumonia, we confirmed that he did in fact have influenza A, and then again, we had the traditional susceptibilities at 48 hours, he remained on the ventilator for three days and was successfully extubated. I think it's clear that by getting this patient on vancomycin earlier, first by the gram stain, and then confirmation by varigene was extremely useful. Well, how did he do in terms of compliance with the CEP1 measure? Well, let's take a look. Did we get an initial lactate? Yes. Did we give the patient broad spectrum antibiotic? The answer is yes, we eventually narrowed that and more targeted actually within 16 hours, but initially he did get broad spectrum antibiotics that qualified by the CEP1 measure. Did we get blood cultures? Yes, we did. Now, here's the thing that we fell down on, that lactate level was done the next day, it wasn't done within six hours, it wasn't done within six hours. So you know why? We failed the CEP1 measure because we didn't get that lactate within six hours. The CEP1 measure is a binary, they get all or nothing. You cannot tell me that there's any evidence to show that getting a repeat lactate at six hours has any impact on patient's outcome. And finally, what's happened to broad spectrum antibiotics since the CEP1 measure went into effect? I think you can see from this particular slide, a huge jump in broad spectrum antibiotic to satisfy a rigid measure, a rigid measure. And then what Debbie's already said, we know that fungal and bacterial co-infections are very unusual. She mentioned the open form article just got published and yet 70% of patients are still getting antibiotics. So where do we go from here? I think we have to reassess that one hour bundle for sepsis without shock, that once an infection is determined, the patient should receive a target appropriate therapy. By the way, antibiotics should not be delayed if an infection is felt to be the cause of presentation, even if they're not in septic shock. Integrate biomarkers, rapid diagnostics. Hopefully we'll get to some e-surveillance measures, some stuff that our group has published. Develop a true predictive model that alert patients who's at risk of progression. Can we use precision medicine? I think we probably can. Research about the predictors of long-term morbidity and mortality. We know that many of these people have a shorter lifespan. They have cognitive issues. What kind of interventions can we do to prevent that? Risk adjust our outcomes. Better ideal endpoints for fluid resuscitation. What's the optimal fluid? Is it balanced crystalloid or normal saline? And what is the time zero? And with this slide, I'm going to end that antibiotics, believe me, were miracle drugs when penicillin was introduced in the 1940s. We thought there were no downside. The pharmaceutical companies were always a step ahead of us. You know, that is no longer true. The new paradigm is antibiotics are good when they're used well, but they're better when they're used thoughtfully. And with that, I'll turn it over to Trevor. All right, great. Thank you. That was a great part of the talk. I think it gets me fired up as a stewardship guy to hear talk about that. I'm going to talk about pneumonia in a particular pneumonia diagnostics. Pneumonia is the most common infection associated with sepsis and is a difficult diagnosis to make at times. And so a lot of empiric antibiotic use is provided for pneumonia. And so there are a variety of respiratory tract diagnostic panels out there. There are upper respiratory tract panels, which some of you probably have. There's a variety of manufacturers for these. They contain a variety of different viral and some include bacterial, particularly atypical bacterial targets. And then I'm going to focus more on the new pneumonia panels because I think we're relatively comfortable with interpretation of the upper respiratory tract panels, but interpretation of the lower respiratory tract panels because they're novel and a little bit complex is an important topic. There's one manufactured by BioFire and one by Univero. And in this next slide, I have the breakdown of the various targets. What you'll note is they have a long list of bacterial targets and they don't all cross over. And then there are also a variety of resistance genes here in the second panel that are captured more in the Univero than the Filmaray. And so the next thing I want to do is just ask, does your laboratory offer panel-based molecular detection of multiple respiratory pathogens? And you may say, well, we don't have any of these panels. Do you have just the lower respiratory tract panels, which would be the Univero or the BioFire? Do you have just the upper respiratory tract panels or do you have both upper and lower respiratory tract panels? Interested in seeing what the adoption is, particularly of the lower respiratory tract panels as they are relatively novel and new to the market. So we'll give it a couple seconds here to let people plug in their answers. And then after we see the answers, we'll spend some time talking about particularly how to interpret these lower respiratory tract panels and what they mean. All right, so it looks like about a third don't have either panels. 10% have just the lower, just lower, about 40% have the upper and then about 25% have both. And I think that's interesting. And so I'm going to start this talk by just talking about the pneumonia guidelines and their recommendations on the use of diagnostics because we want to couch any discussion of how to interpret a diagnostic tool into when we should use it. The CAP guidelines, which were published now about a year and a half ago, address the use of diagnostics, but interestingly not the use of how to diagnose pneumonia. They address antibiotic choices, when to cover for MDROs. They remove the HCAP category. Remember, HCAP is dead. May it rest in peace. And they address duration of therapy. And then the HAP-VAP guidelines, which were published in 2016, have a lot of information on the use of diagnostic cultures, but again, not much on diagnosing HAP or VAP. And again, antibiotic choices, combo therapy, but neither guidelines address pneumonia panels because they weren't available when they were written. So when we talk about this, I think it's important to think about how do we diagnose pneumonia? And that these microbial diagnostics don't diagnose pneumonia. Remember, pneumonia is a clinical diagnosis made by having a compatible clinical syndrome plus an infiltrate and imaging. And what's that compatible clinical syndrome? Well, it's what Ed laid out. It's cough and pleuritic chest pain and fever and tachypnea and crackles and agaphonia, all that stuff you learn in medical school coupled with biomarkers like CRP and procalcitonin and leukocytosis, and then changes on imaging. The problem is those tools are not that great at diagnosing pneumonia. The combination of fever, crackles, hypoxia, and tachycardia, well, I might say, wow, that sounds really concerning for pneumonia, had a positive predictive value of 57% for actually having a radiographically confirmed pneumonia. And for VAP, we do lots of over-diagnosis of VAP. Of 231 cases of VAP diagnosed at one institution, 60% of them weren't actually pneumonia and resulted in at least 1,100 unneeded antibiotic days. And so just really important, particularly we're gonna talk about molecular tests done on non-sterile specimens. These don't diagnose pneumonia. Pneumonia is based on clinical and imaging findings, and these microbiologic tests help us to define the etiology of pneumonia. Now, the one exception to that, if you read the guidelines, is in the VAP guidelines, they very specifically state that if you have quantitative cultures below the diagnostic threshold, you can stop antibiotics safely. And that's because that's been studied and it is safe to stop antibiotics in that situation. So when should we use microbial diagnostics in pneumonia? Well, in community-acquired pneumonia, for outpatients, it's not recommended you use them at all. For inpatients, if it's typical CAP, not an ICU, the typical antibiotics cover it really well. And the guidelines recommend not using any microbial diagnostics. I think some of us in ID have maybe slightly different opinions about the utility of some things. The exceptions are, when should we use diagnostics? Well, severe community-acquired pneumonia. If we're using broad-spectrum therapy, the diagnostic tools help us to deescalate immunocompromised patients because they may have more unusual organisms. And then if you've given them typical antibiotics and they're not getting better, you should probably try and sort out what's going on. For HAP and VAP, because we're almost always starting broad-spectrum therapy, it's recommended all patients have a respiratory tract culture. In particular, in VAP, they recommend a non-invasive sampling. And that means not a bronchoscopy, just get some sputum. And that's one of the things that, personally, on rounds drives me crazy, in that somebody gets intubated, antibiotics are adjusted or added, and nobody thinks to get a respiratory tract sample. There's just not an excuse. You can easily aspirate from the trachea once you have an ET tube in place, and you just have to wait 24 to 48 hours to get a directed bronchoscopy. There may be a role for that, but the guidelines recommend just get some cultures. And what tools do we have? Well, we have blood cultures and respiratory tract cultures and, excuse me, urine antigens, and now these multiplex panels. The problem with our traditional methodologies is they have poor yield, particularly for community-acquired pneumonia. Blood cultures have a low yield. They rarely result in changes in therapy. Sputum cultures often aren't available. Once you start antibiotics, particularly community-acquired pathogens are difficult to detect within 12 to 24 hours. And while the urine antigens improve our sensitivity for detection of things like Legionella and pneumococcus, they still have relatively low sensitivity. And so that's one of the reasons that CAP guidelines aren't very positive on using microbial diagnostics. And this is a quote from the CAP guidelines. The overall poor yield of sputum evaluation for detecting organism causing CAP limits its impact in management of patient outcomes. And I agree. And just to support their conclusions, here are two studies that looked at the etiology of CAP. The first one here is a large U.S. study published in the New England Journal a few years ago that found six out of the seven top pathogens detected were viruses. And the only bacteria that made it in there was strep pneumo and that 84% of people with community-acquired pneumonia didn't have a bacterial pathogen detected. Now, most of them got antibiotics because they had community-acquired pneumonia. And the guidelines say to give people with community-acquired pneumonia antibiotics, even if you detect the viral pathogen. Data from Europe, very similar from a pneumococcal vaccine trial showed about 25% had a bacterial pathogen detected. A little higher, but still a huge proportion of patients had no bacterial pathogen detected. Does that mean that they didn't have a bacteria causing their pneumonia? I don't think so. Some of them probably didn't, but we don't have great tools for detecting it. So we implemented the pneumonia panel back in May. And so why would we do that, particularly in sort of the heart of a COVID epidemic? Well, our thought as a stewardship program was rapid detection of pathogens helps us to one, improve the etiologic diagnosis of pneumonia and I'll show you some data that we actually can sort out what the etiology of pneumonia is better and we can do it more rapidly, which allows us to adjust our antibiotics more rapidly. And why would you do it during COVID? Well, everybody with COVID has pneumonia and a lot of them get started in antibiotics. A lot of them have prolonged intubations, they have fluctuating respiratory status, biomarkers kind of went out the window as we didn't know what they meant. And there was just a lot of unknowns. And so rather than just escalating every time somebody got slightly worse, we wanted a tool to help adjust antibiotics more rapidly. I'm gonna focus on the film array. That's a panel we implemented. And these are the targets that are in the film array, gram positives, gram negatives. These atypical and viral pathogens are the same thing that's in the respiratory pathogen panel that's produced by BioFire. Although it's a little less detail, it doesn't have all the different subtypes of influenza A, just has influenza A. It then has some resistance markers, five carbapenemase genes, and then CTXM, which is a common, probably the most common ESBL or extended spectrum beta-lactamase in the US. So do these pneumonia panels allow us to identify more organisms? The answer is yes, they definitely do. This is a large multi-center validation study of the BioFire pneumonia panel, 900 bronchoscopies and 900 sputum specimens. They identified Staph aureus, Pseudomonas, H flu and rhinovirus enterovirus, the most common pathogens. And 50% of the BAL specimens are positive. And 70% of the sputum specimens are positive by the pneumonia panel. Now, one of the things you will see with these multiplex panels is the detection of multiple organisms. It's more common in sputum and a quarter of the sputum panels had three or more organisms. And so when you detect three or more organisms, that is gonna be a little more complicated to sort out. This was a subset of that group of about 260 bronchoscopy specimens where they collected clinical data because in the large validation study, they only had microbiologic specimens. Here are the increased yield of bronchoscopy by 63%, by using the pneumonia panel over standard methods. What's the sensitivity? Now, in the validation study, the sensitivity was compared to a quantitative culture of not 103, but 10 to the 3.5 colony forming units per ML. That was a cutoff that they used for what was considered positive, the gold standard. The sensitivity varied a little bit, but overall was very high. Only 16 false negatives out of 1,682 tests. Now you have to couch that and it only detects what it's designed to. And so if you had a pathogen that's not on the panel, it's not going to be detected. Second, the specificity. And so the overall take for my take on the sensitivity is it's very sensitive. And of the 16 misses actually only 10 of them they actually couldn't find the organism. They found the organism, but it wasn't there. And so there were some, but very few misses. The specificity is more of an issue. There were a lot of detections where they didn't meet this gold standard culture diagnosis. Most of them were Staph aureus, H. flu, Marek cell, and Pseudomonas. And in fact, they had what were considered 875 false positive results. And were they really false positives? Well, a quarter of them, the organism was there on culture, but it was below the level of what they considered positive. So it was below that 10 to the 3.5. Basically the other three quarters were there when they went back and looked with other molecular methods. And there were actually only three truly false positive results where they couldn't find the organism by any other method. And so my take on this is they are detecting organisms that are there, but they're in very low levels, generally below what we would detect by traditional microbiologic culture method. And so how these are reported are an important thing to understand. The manufacturers have tried to report them in a way that helps us to interpret them and understand that topic. Now, the bacterial pathogens are reported semi-quantitatively into these bins. The viral atypical and resistance markers are reported as detected or not detected. And so they group the DNA copies per ml. And that's the report you get, DNA copies per ml. It's not a column form units per ml, it's DNA copies per ml, and they group them into bins. And if it's below 10 to the 4th, it's considered not detected. And then it's grouped into these 10 to the 4th, 10 to the 5th, 10 to the 6th, and 10 to the 7th or greater bins. And so what do these bins mean? Well, the first thing you have to remember is that the DNA levels are much higher than the culture levels. And the correlation with culture is not really fantastic. And then if it says 10 to the 4th DNA copies, that doesn't really correlate with 10 to the 4th colony form units per ml. And that's what this table is showing you. This is the pneumonia panel here on the left here, and this is the quantitative culture. What you can see is if it's not detected by culture, it's really good for not detecting it by the pneumonia panel and vice versa. But if you have 10 to the 4th DNA copies, that may be 10 to the, maybe not detected by culture mostly, but maybe in the 10 to the 3rd or 10 to the 4th colony forming units per ml. As the DNA copies go higher, you can see you end up mostly, excuse me, as the culture results go higher, you end up in the higher results. And so there's not just great correlation here. But the negative correlates really well. And when you have greater than 10 to the 5th organisms, that correlates quite well with having at least 10 to the 5th and usually 10 to the 7th or greater DNA copies per ml. The one thing that is pretty good is concordance and that the most common pathogen, the most prevalent pathogen on the pneumonia panel is usually the most prevalent pathogen on the culture. And so you have to sort of change your thresholds here and that 10 to the 5th DNA copies per ml is often not going to be detected by traditional culture methodologies. But 10 to the 7th or greater usually will be. This is another study that took over 200 bronchoscopy specimens and then went back and did pneumonia panels on them. What you can see is in the purple here, it increased the detection significantly of most pathogens, particularly of Staph aureus and H flu. Now there were some pathogens that weren't on the panel and so weren't detected by the panel. And that's what this donut is highlighting. And then over here on the right side, you can see these are the DNA copies per ml and whether they were detected by culture or panel. And for both Staph aureus and H flu, when the DNA copies were low, particularly 10 to the 4th, 10 to the 5th, they were really rarely detected by culture and mostly were detected only molecularly. But as you got up to 10 to the 7th, they were detected by both culture and panel. So there are issues with implementing these pneumonia panels in that they really do need to be coupled with cultures. You would never want to do a molecular panel without a culture to provide you phenotypic susceptibility data. Another issue is that poor quality sputum is still poor quality sputum. And so I don't want to culture spit and I don't want to run pneumonia panels on spit. And so if you have sputum quality rules, they should apply to both. Another thing is redundant testing. These tests aren't cheap. And so because this contains all the same things as a respiratory pathogen panel, you don't want to run this when you run a respiratory pathogen panel. You probably don't need to run your urine antigens when you do this. And so this replaces a lot of your other microbiologic diagnostic techniques. And then these are highly sensitive molecular panels. You don't really need to repeat them. There is some issue with specimen type there and that if you got a sputum and then somebody got intubated, you may want a bronchoscopy specimen because it may be a little more sensitive for detecting things. We've certainly seen that with some of our respiratory viral testing and that their initial nasopharyngeal swab may be negative. And then a bronchoscopy, particularly in somebody with pneumonia is positive. And so there are some issues, particularly building clinical decision support around that. And so when we implemented this, we also only wanted it used in people where we really thought it would change therapy. And that's, we thought was mostly in our ICU patients where we have our severe cap, where we're going to be more aggressive diagnostically, our VAP, where people are quite ill and we're using broad spectrum therapy. And that ICU, usually you're using broad spectrum antibiotic therapy. And we wanted it used by people who understood it if it was going to be outside the ICU. And so we restricted it to our ID and pulmonary physician. We built some things in our record so you couldn't repeat it within three days if it was a sputum or seven days as a BAL. And then we strongly suggested avoiding the respiratory pathogen panel while used. We wanted to build some really fancy clinical decision support, but then we ended up sort of, because it was complicated, doing it on sort of a really simple method where it basically put some things in here that said only for ICU patients are ID or pulmonary. And you have to click the button saying which one you are. This patient's in the ICU or maybe in the ED going to the ICU. I'm an ID or pulmonary physician or I don't meet the criteria. And if you click the button that says you can do it, it's coupled where both the pneumonia panel and the culture are there. And so that's sort of been our stance in building things is we want to make the right thing easy and the wrong thing, not perhaps impossible, but difficult. Like you have to purposefully screw this up to not get a culture. And then if you don't meet the criteria, then you can call our friendly micro director or get an ID or pulmonary consult. This is how the results actually come off the BioFire machine. You can see it's really nice. You can easily interpret this where the how much pneumococcus and H. flu are in this specimen. Unfortunately, we don't live in worlds where a lot of our informatics systems talk to each other in a super friendly way. And so our lab information system couldn't actually handle exponents. And so we have to show people a bunch of zeros, which I think impacts people more than exponents. There's just something about seeing 100,000, which means something different than 10 to the fifth. And so this is an actual result. You can see it's a polymicrobial specimen with a lot of pseudomonas, some CLAB, a skosh of marxella and MSSA. And the culture results showed only pseudomonas and CLABSIELLA, illustrating that these low levels, 10 to the fifth, 10 to the fourth, aren't generally detected by traditional microbiology techniques. What is the impact in antibiotics? What I'll tell you is there hasn't been a lot of study on this. In fact, there's only been one very small real world study that I have seen as yet published. This is a theoretical study where they looked at the theoretical impact of the pneumonia panel on that 260 bronchoscopies where they had clinical data. And so they looked at the results and they said, what would have happened if people acted on the results? Well, in about half the patients, there would have been an appropriate de-escalation. What that means is the pneumonia panel said there's no MRSA here, vancomycin could theoretically have been stopped and the culture didn't grow any MRSA. And so I think that's where we see a lot of promise from a stewardship standpoint is that about half the patients were gonna be able to stop at least some of the antibiotics. Also important is appropriate escalation. That means the panel said their pseudomonas present, anti-pseudomonotherapy wasn't being provided. And so anti-pseudomonotherapy could be added and the culture then grew pseudomonas. And so that, I think from a safety standpoint is very important. And that's gonna be a small, but I think important proportion. Inappropriate de-escalation. That means the panel says there's no MRSA, but we grew MRSA in the culture. 1%, very rare that's gonna occur. Theoretically could occur. And then inappropriate escalation. I think from an ID standpoint, this is our concern is that there'll be a lot of things detected and people throw antibiotics at a lot of things that aren't needed. It says MRSA, it never grows in the culture. We started vanco, we didn't need. And so that's about 15% of the time. And then about a third, it was no change. Now the overall gestalt of that in this study would have been six fewer antibiotic days per patient. And so I think the overall, I think this is going to be a win for antibiotic stewardship. This is the one real world study. It was a abstract published at ID week. That's now out in open forum infectious disease. Small case matched, case control matched. 26 pneumonia panel, 13 usual care. No difference in clinical outcome, but higher rates of de-escalation and antibiotic discontinuation. And that occurred about 24 hours earlier. And I think that's the real advantage here is the ability to stop or de-escalate antibiotics earlier than waiting 24 to 48 hours for traditional microbiology. Now we can do it at four to six hours. So what do we do with the results? We actually developed a guideline on this because we felt our clinicians needed some help. And frankly, we as a stewardship team wanted to make sure we were all on the same page of how to interpret this. And so for example, for Staph aureus, if the PCR test from a lower three track specimen says there's no Staph aureus, well, there's probably no Staph aureus and there's definitely no MRSA. And so we would say, boy, you don't need vancomycin in that case. You just don't need it. Don't continue it, just get rid of it. If it says there's MSSA, the typical cap antibiotics and most beta-lactams are gonna treat that really well. And so again, you don't need vancomycin. If it says MRSA, well, it's hard for me to argue in an ICU patient where a test says there's MRSA, particularly if it's a high level to cover MRSA. Now you could argue, what if it's just 10 to the fourth and that's the nuance that's really hard to capture in a guideline like this. I have lots of Klebsiella and just a little bit of MRSA. What does that mean? I don't know. It's difficult to say there. For Strep, it's a little more easy to know what to do with your antibiotics because susceptibility can inferred by identifying the organism. One of the things we did add because some of these organisms won't be grown in culture, particularly things like H. flu is some oral de-escalation options thinking that in a couple of days, people may want to de-escalate and without susceptibility data, it'd be nice for them to know what they should use. Gram-negative organisms is a little more complicated animal because not all E. coli's are the same. And if you've just come from the community with E. coli, that looks very different than say an E. coli from a patient who spent a month in the surgical ICU. And so we tried to couple that data with our antibiogram data, highlighting that say for enterobacterial cloacae, cefepime would be a reasonable choice. And we generated both a respiratory tract and a traditional antibiogram and tried to provide some of that data. But you know, for E. cloacae, boy, ceftrax and or piptazo would not be very good choices. And then for E. coli, we integrated the CTXM gene, this marker for ESBL. And if it's not present, well, then a third generation that was born or piptazo is reasonable. And boy, if it is present, that might be a patient you wanna use a carbapenem. And then we do have some recommendations on these molecular markers. The problem with the molecular markers is just because they're present doesn't necessarily mean they're gonna be resistant. And just because they're absent doesn't necessarily mean they're gonna be susceptible. And so you really have to couple them with your antibiogram data. And I have a case later that addresses it. And so I think there are some challenges with this. What is the clinical significance of low levels of organisms or of any of these? Because we even have that struggle with culture-based methodologies. And that, yeah, we detected this by culture, but we don't see clinical evidence of pneumonia. And that's why I think it's really important that we couple all this data with the clinical diagnosis of pneumonia. What's the significance of mixed detections? Is the organism alive or is it dead? What's the impact of antibiotic therapy? They've been on antibiotics for two days. Does that impact this? What about colonization? These interpretations, I already talked about resistance. And then frankly, I think all of this requires stewardship. And so I think the key things are you should utilize this where it's gonna help the most. Is it gonna help that much in typical cat patients who aren't that sick that you're giving typical therapy to? No, no, I really don't. I mean, it might help us to make a diagnosis, but usually it's gonna show pneumococcus or H flu or something we would treat with ceftraxone and that'll be just fine. But we should use it in those patients, particularly those who are ill, severely ill, or we're using broad spectrum to rapidly deescalate. It's very sensitive. It's being used in a non-sterile site. You're going to see multiple positives and we're not going to grow some of those positives, particularly with staph aureus, H flu, and even pseudomonas. It's reported semi-quantitatively and not everything you detect has to be treated. You can use that relative quantity to help determine what needs to be treated and what doesn't. And I think this is where stewardship is super important to helping teach people how to interpret this and to get comfortable with the results. And with that, I'm going to turn it back over to Debbie to talk about sort of how to put all this together. Thanks so much, Trevor. What a great presentation. So when we look at the challenges to the timely diagnosis of bloodstream infections and pneumonia, you know, lack of availability, awareness, and access are three of the challenges. So looking at some of the common questions, are these syndromic panels accurate enough to be confident in the results? What if there's multiple organisms identified? Can a non-microbiologist understand these test results? And can clinicians understand the micro report? These are just some of the questions we're going to address in the Q&A. So Trevor, I'm going to turn it back to you to go through one of your cases to help apply this to the audience in patient care. Great, thanks, Debbie. So this is a case that we had earlier this year, a 77-year-old woman with dementia, O2-dependent COPD, coming from a long-term care facility with confusion and increased oxygen need. Like many of our long-term care facilities, COVID was present in that facility and circulating. On exam, she did have crackles. She was febrile. Her white count was normal. Her procalcitonin was 0.08. She had bilateral hazy opacities concerning for developing consolidations, and she ended up going to the ICU due to a high O2 need. And so a case like this generates a lot of questions when you evaluate it. First, what kind of pneumonia is this? How would you classify this? Is this community-acquired pneumonia? Is this HCAP? Is this community-acquired pneumonia with resistance risk factors? Is this just COVID-19 pneumonia? What am I going to call this? Depending on what you call it, am I going to start antibiotics? If so, what am I going to start? I'm going to say, well, it's probably all just COVID and not give antibiotics. Am I going to start traditional CAP antibiotics? Am I going to say, well, I'm still operating in sort of a HCAP or CAP resistance risk factor model and give broad spectrum therapy? Or am I going to just sort of hope the ED guy started something and just do whatever he did? And so how are we going to refine our choices? And that's, I think, a really important question, something we're trying to address as clinicians. So is this viral pneumonia due to COVID? Well, we need a COVID test. The problem is a lot of our COVID tests take 24, 36 hours to come back. And so in that interim, does she need antibiotics? Well, the CAP guidelines would suggest even if you diagnose pneumonia from influenza, you give antibiotics. Not sure I'm completely in agreement with that. I think we have some other tools to help sort that out, but that's for a different talk. If you do decide to give antibiotics, the question is, does she have risk factors? The new CAP guidelines have eliminated HCAP, in particular have eliminated long-term care facility as a risk factor for needing broad spectrum antibiotics. And they would highlight looking back to the history of respiratory tract isolation and does she have a history of MRSA from the sputum or pseudomonas from the sputum? If so, empiric therapy in an ICU-bound patient for those pathogens should be provided. Another factor you would look back at, does she have a recent hospitalization in IV antibiotic use? And the CAP guidelines emphasize the presence of needing both. Or maybe you're operating facility where you have locally validated risk factors for MRSA or pseudomonas. But really, these are the factors we should be looking back at now. Is there a history of respiratory tract colonization with resistant pathogens? Have they had recent IV antibiotic use in the hospital? And if not, we don't need to provide those therapies. If those risk factors aren't present, the typical CAP antibiotics are what's recommended. They have a history of MRSA, we would add vancomycin, history of pseudomonas, we would add an anti-pseudomotile agent instead of ceftriaxone. If they have history of IV antibiotics in hospital stay, we would give both vancomycin and an anti-pseudomotile agent, plus something to cover atypicals because the patient is coming from the community. Now, the other thing that I get asked so many times is, what's the role of procalcitonin or biomarkers in COVID pneumonia? Can we use it for deciding on antibiotics? And I have this disclaimer here because as you are all aware, the literature on procalcitonin on everything with COVID is rapidly expanding. And there could be data presented at this conference even that goes against what I'm saying. And so it's all sort of caveated by I could change all these opinions next week. But right now, this is what I found from the literature. So the procalcitonin levels in COVID are generally low upon admission. So it's a large database from New York. The median procalcitonin level is 0.2. Now, CRP seems to be significantly elevated in COVID pneumonia for reasons that I don't think we totally understand. The procalcitonin levels certainly do vary by severity of illness, and the levels may be prognostic. This is some data from New York looking at a small subset of patients who survived or died. And patients who died, the procalcitonin levels skewed higher. Now, they weren't super high like we see in bacterial sepsis or bacterial pneumonia, but they certainly did trend higher. This is another study looking at patients with moderate, severe, and critically ill COVID pneumonia. What you can see, those who were severely ill, critically ill, had higher procalcitonin levels, although not that high. I mean, median was around 0.5. Those who got better and were discharged, their procalcitonin levels declined. Those who didn't, who went on to die, their procalcitonin levels rose and did not decline. And that's actually a trend we see from numerous procalcitonin studies in the past, showing that a rising procalcitonin level is a bad prognostic sign in many conditions. And then I'm amazed at the proliferation of meta-analyses around COVID. Within a few months, there were already numerous meta-analyses looking at numerous factors with COVID. This is just one looking at the prognostic value of procalcitonin in COVID. And elevated, and I put elevated in quotes because it was defined by individual sites as to what was elevated, was associated with a poor outcome, mortality, severe disease, or ICU stay. And a procalcitonin of 0.5 or greater had an AUC of 0.88 for a poor outcome. And actually, in this study, was the best marker for poor outcome. And so in my, this is again, just my opinion, it's a good marker for people who are going to have a bad course for their COVID. Why? Is it bacterial infection, just bad COVID, cytokine release? I don't really have the answers to those questions. What about its role in helping us with the antibiotics? This was a large multi-center study from Michigan. 1,700 patients hospitalized with COVID. They looked at early, early antibiotic use only in bacterial co-infection. Empiric antibiotic use, like has already been mentioned, about 56% of the patients. And bacterial infection, very low, 3.5%. And the predictors of bacterial infection were kind of the predictors you'd expect, being older, other comorbidities, severe illness, high white count. Procalcitonin was measured in 900 of the patients, and a level of 0.5 or greater was actually the best predictor for bacterial infection. But 20% of people who had a level over 0.5 didn't have a bacterial infection defined. Is that because they didn't, just couldn't detect the bacterial infection that was there, or they didn't have one? We don't know. And a procalcitonin level of less than 0.1 had a negative value of 98% for bacterial infection. So what I generally tell our clinicians is a low level of less than 0.1, yeah, I don't think we need antibiotics, particularly once we define that this is COVID. Over 0.5, I'm not going to argue with antibiotics. Those usually portend a bad outcome, and so we should evaluate them for bacterial infection. If we don't find one, you know, I think we could probably stop the antibiotics. In between 0.1 to 0.5, that's where I think we trend and sort of make decisions over time. And so in this patient, that patient had a pneumonia panel that was negative, which was positive for SARS-CoV-2, and all the antibiotics were discontinued. I want to move on to a second case, looking at some of these utility of these markers for antibiotic changes. This is a 56-year-old male with bicuspid aortic valve, had to undergo a repair. Seven days post-op, developed leukocytosis and fever with some increased cough. He had a progressive left lower infiltrate from three days ago when he just had a little atelectasis and effusion. He developed hypoxia and progressive respiratory failure, required intubation. Blood cultures tracheal aspirate and pneumonia panel were sent, and he was started empirically on vancomycin and cefepime for hospital-acquired pneumonia. The sputum grand stain showed many white blood cells and a few mixed gram-negative and gram- positive flora. The pneumonia panel was positive for greater than 10 to the seventh or 10 million E. coli and 100,000 haemophilus influenzae. And so in this patient, on vancomycin and cefepime, what should we do with the antibiotics? And I think the first question is, what should we do with the vancomycin? And so as we talk about this through this next little section, I just want to remind you about genotypic versus phenotypic resistance. Remember, genotypic resistance is detection of genetic markers for resistance. That's the MEK-A or C gene, which stands for meth, which predicts methicillin resistance in staph. VNAB, which predicts vancomycin resistance in enterococci. KPC, which predicts carbapenem resistance. They predict it, but it's not 100% that they will be resistant. It just detects the genetic marker. Phenotypic resistance is that report you get from the micro lab where they do susceptibility testing and tell you specifically what is it actually susceptible for. The panels provide genotypic resistance markers. Our traditional cultures provide phenotypic susceptibility results. You've probably all seen studies on using nasal screening for MRSA to decide if we should use antibiotics or not for treating for MRSA pneumonia. This is a large VA database looking at many years worth of nasal MRSA screening. If you had a negative MRSA screening of the NARES, that had a negative predictive value of 96% for MRSA positive lorister tract infection. That varied a little depending on where you were, depending on the prevalence of MRSA in your population. Important to always keep in mind with negative predictive values, but a negative PCR test from the NARES is a really good way to decide whether I should continue vancomycin or not. This was a study done at six ICUs over a couple of years, looking again, negative predictive value of nasal MRSA screening. In this site, it was 99% negative predictive value. Again, the prevalence was relatively low at this site, which means most tests would have a really good negative predictive value. Interestingly, it was largely ignored. Antibiotics continued more than three days after negative screening in many patients resulting in 2,600 unneeded days. Then vancomycin was started after a negative screening, giving you nearly 7,000 additional days. About 10,000 totally unneeded antibiotic days. If you have a negative lower respiratory tract specimen for MRSA, I think that's even more useful than a nasal swab. You could argue that it's just a NARES swab and there could be lorister tract colonization we haven't detected, but if I have an adequate lorister tract specimen that I have a molecular test that's negative for MRSA, I think that's adequate data to discontinue vancomycin. We don't have data showing it works for lower respiratory tract specimens, but we do have it for a less adequate specimen where it's really good for ruling out. I think it's a strong argument that if it's negative for MRSA, we stop the vancomycin. The second question is, what do you do with the E. coli? The answer to that question is, well, it depends on where you live and what your prevalence of resistance is in E. coli. I've just highlighted a variety of prevalences around the world and in Europe and how these vary significantly from in the Scandinavian countries, eight to nine percent, to Southern Europe, 20 to almost 40 percent. That varies by your institution. That's where I think using antibiograms and some of these genotypic markers for resistance can be useful. We have this CTXM gene. CTXM is the most common plasmid mediated ESBL gene in both Europe and the U.S. It predicts for resistance to ceftriaxin and cephotaxin. Can we use it to decide on what to do with therapy? This was a study done at two sites where they developed what's called a genotypic antibiogram. They basically said, if you have E. coli that's CTXM positive, what's that going to be susceptible to? Versus E. coli that's CTXM negative. Now, they did this on blood culture data, not respiratory culture data, but I think it's relatively, it's useful and applicable. If you had CTXM positive E. coli, boy, you really shouldn't use ceftriaxin or cefepime or probably peptazo or a quinolone, but a carbapenem would work really well. If you had an E. coli and it was CTXM negative, because CTXM was a predominant ESBL gene, 98% of those were susceptible to ceftriaxin, and so you could de-escalate down to a third generation cephalosporin. Now, you still might be reluctant to use a quinolone. The data was similar for Klebnumo, you can see, but a little bit more variable in that quinolones became more of an option and ceftriaxin eroded a little bit more because there are a few other CTXM, a few other ESBL genes out there like SHV and some other ones, and so that's why when we developed our, in our recommendations, we basically said if you have E. coli and it's CTXM negative, ceftriaxin or peptazo are very reasonable choices. If it's CTXM positive, I'd probably reach for a carbapenem until we have phenotypic susceptibility, and so in this patient, they were de-escalated to ceftriaxin, it was susceptible to ceftriaxin, and vancomycin was also discontinued, and I think this illustrates how we can rapidly make these de-escalation decisions in patients within, you know, 6 to 12 hours of starting initial empiric therapy, and so we can move that de-escalation window back much earlier. I'm going to turn it back over to Debbie here. Thank you, Trevor, and I want to remind our audience if you have questions, you can type them into the chat section for questions, and we'll be happy to address them. We have several coming in now, but just want to make sure you know to do that. So how do we, let's look at collaboration across the interprofessional team. How do we get people to react to the data? You know, that to me is the role of the stewardship program, providing education and guidance, integrating these new tests into the stewardship program. Trevor so beautifully showed how he integrated the pneumonia panels into the electronic health care records, and, you know, nudging the ICU clinicians with our stewardship team. You know, I look at working with my colleagues, my SICU PharmD is Tony Gerlach, just a great example of a surgical steward helping in the ICU. So diagnostic stewardship is as important as antibiotic stewardship. We need to select the right test for the right patient at the right time, and then figuring out actually how to use the results. So working hand-in-hand with our microbiologist is such a key component, and this slide really just shows that diagnostic stewardship with antimicrobial stewardship must be integrated together to get it right, and that's the whole key of a stewardship program. It's comprehensive care, and collaborating amongst all of our health care professionals. I work in a 1400-bed hospital. When we started stewardship, it was myself and a physician. Now it's a much more elaborate team as we continue to grow, and the need for stewardship is clearly identified. But I remember starting, and I would never think to impose guidelines without having my ICU clinicians who are caring for the patients not be at the table with me. So I think you saw how we looked at these new diagnostic tests. We need the microbiologist at the table. Trevor so beautifully showed how he worked with the intensivist to streamline this just in the ICU, and it's a collaborative approach. It can't be done alone, and that's what care is all about. So we have about 15 minutes now to move to Q&A, and I'm going to start with some of the questions that have already come in, and some of them that we had proposed earlier. So Trevor, I'm going to have the first question go to you. Since you implemented the pneumonia panel, can you explain kind of your first steps you took with your ICU clinicians to add this pneumonia panel to what I would call the ICU toolkit for pneumonia? Because you provided them with another valuable tool. What were your first steps in really identifying the need for this? And then I'll follow up with a couple other questions that have come in. Yeah, so that's a great question. So I think it really is important that you work with your ICU clinicians, that they understand the test. And so the first thing we did on our end was develop some guidelines for interpretation. And we have a multidisciplinary stewardship committee, and then we sent those out to our ICU physicians and discussed with them, and I actually met with three of sort of our key clinical leaders in our ICU and talked about the test and what it means and how to interpret it, and made sure that we had their buy-in and their understanding. Because with the tests like this, they need to understand what it means and how to interpret it. Now, we then, I also did a grand rounds for our pulmonary critical care group, and I've tried to go around and educate a variety of the very, because we have a variety of clinical care providers, surgeons, anesthesiologists, et cetera. And so just trying to help people to understand. But also we're, the one thing I didn't mention is we're doing stewardship and that we review every one of these when they come back and call the team with recommendations as to what to do. So I think it's important to get them in early when you're developing the guidelines of how to use them. We wanted to talk with them and make sure that they were comfortable with our restrictions that we were putting in place. And then to make sure that we have them as sort of, we have some content experts in their group that even their peers could go to, to ask, how do I use and interpret this? So Trevor, what I just heard was the definition of collaborative care. You worked collaboratively with your ICU clinicians to ask them how they would use this new tool. Would they find it helpful? Because there's nothing worse than implementing a rapid diagnostic test that nobody uses or doesn't understand how to use. And then really working how it's going to report out the education you had to do. There are many steps it takes to get these things into hospitals working properly. So I think you really drove home a great example. And the question I have for you is kind of on the same realm for the blood culture rapid diagnostic test. Can you walk us through when you first implemented these in your hospitals? What kind of was your first step in getting them successfully implemented? Well, these are great questions. I kind of want to bring us down a little bit. I work in a 13 hospital system where there are two teaching hospitals. And so when we rolled out any of our molecular tests, in particular blood cultures, which was one of the first ones we did after clostridioides difficile, the question is how can we educate our clinicians how to interpret the microbiology report? And can it be understood by non-microbiologists? These were key questions I think that we needed to answer in terms of understanding. So we did have an education program. We rolled this out to our clinicians. We try to help them understand the gram negative and the gram positive panel. The other thing linked into stewardship, which is so critical that both of you have highlighted, is the way our microbiology lab worked is they would call the initial blood culture report to the floor or to the ICU, gram positive cocci and clusters as an example. Then the blood, we have a separate molecular lab. We have a central lab. And so the blood culture stuff gets sent over to the molecular lab. And then two to three hours later, the molecular results are put into the electronic health record, but there isn't another call to the clinician, one of my big pet peeves. And now we've educated our clinician, listen, two to three hours after that blood culture has been reported, go back into the electronic health record and see what the results are for the molecular to see whether or not you need to escalate or deescalate based on the preliminary findings. And we've been able to work that out pretty well, but we had to over and over again, explain it. Plus the fact, at least in my institutions, we have a closed intensivist model and we have an employed hospitalist model. So we sort of have them as a captive audience and we have multiple education opportunities for both of those. But at some of our community hospitals, it's not quite like that. You have independent hospitalists. You don't have an intensivist model. You have pulmonary care people doing intensivist stuff, but that's not full time for them. And so we have to work with different models and understand that much of the care that we have in the United States is not in academic tertiary centers, but in smaller community hospitals. So we've had to translate the educational material so that it would be applicable, not just in the academic setting, but also in the community hospital setting. So I think it's very important as we roll out these new molecular studies. Boy, you just segued to the next question, which was, do we need 24-7 staffing of stewardship to optimize antimicrobials and de-escalation? And I can just share, at our own institution, one of our stewardship initiatives was the management of Staph aureus bacteremia. And we know if it's mismanaged, there are poor outcomes for patients. And we actually succeeded at having the operation, we operationalized the management with our physicians collaboratively coming up with these guidelines by the pharmacy department. Pharmacists work 24-7 and we're here at all hours. Our goal as stewardship, my colleague, Kerry Bauer and Eric Wensler really led this project in teaching our entire pharmacy department, every pharmacist, and that does not mean all ID trained pharmacists. Working within our EPIC system, we developed algorithms. If a patient's blood culture turned positive for Staph aureus bacteremia, they had to look if they were on appropriate therapy defined by our guidelines. So we had rapid diagnostic tests. So were they to be on vanco or not? Was it Staph aureus or was it Staph epi? Anyway, we operationalized it through the pharmacy department and did provide optimal management 24-7. So I think there are examples of stewardship and you have a few people doing stewardship, but I believe every physician who prescribes an antibiotic is an antibiotic steward. We just have to guide them along the way. And yes, it is a 24-7 operation, but it doesn't mean the stewardship team has to be in the hospital 24-7, in my opinion. You empower others to do best of care. Trevor, I don't know, maybe you could explain how your program works. Is it 24-7 or how does it work? That's a great, I totally agree with you on that, Debbie, and that, yeah, our stewardship program is not available 24-7, but we have done the same thing where our team-based pharmacists are acting on results and training of our pharmacists, I think, is as important as training of our clinicians because our clinicians listen to our pharmacists. They look to them as experts for interpretation. And so I agree 100%. We don't staff this. I think, you know, that question is an interesting question. Do you get more bang for your buck if you staff these 24-7? You might. The problem is there is a reluctance to make many changes other than escalations, particularly in the off hours when it's covering physicians. And so your biggest bang is probably during the, you know, 6 a.m. to, you know, 8 p.m. And hours after that, it's more of filling in the safety issues. But at least that's my experience in the ICU is people are reluctant to make major changes in the overnight hours where they're covering patients. Okay. We have two questions we want to get to before we have to end our program. Have either of you experienced any issues currently getting reagents for the molecular pathology and impacting on stewardship? I'll answer. We have occasionally had some difficulty with getting the panels and with reagents. There was actually an earthquake out in Salt Lake City, which interrupted the supply of biofire panels for a while. And so I think we have had some, but they've not been major. Okay. And Ed, I have a question for you. You know, we've really, the sepsis management has focused on bacterial sepsis, but how do we address viral sepsis since we don't always have effective therapy? Can you provide any comments on that? Well, that's a great question, especially in the era of COVID, where clearly some of these people who come in with fairly significant COVID pneumonia certainly meet all the criteria as looking like they have sepsis and severe sepsis. Some of them, by the way, as you know, requiring pressers. And I think, again, you have to use, you know, there's no one set answer. There's looking at all the diagnostic and clinical tools that are available to you and actually customizing your treatment plan based on the best available information. And I think that Trevor mentioned the procalcitonin, which we use heavily in my institution because we've unfortunately had a lot of experience with COVID patients. And we found that the low procalcitonin level is given our clinicians, it took us a couple of months, comfort that these patients probably didn't have a co-bacterial pneumonia and that this was all viral. But there's no question that viral processes, severe viral processes can look like severe sepsis and even septic shock. The, I think multiplex panels are very helpful in trying to differentiate some of these disease processes. And the only viral pneumonia that we've really had, true treatment may be influenza, which right now is everybody knows is at extremely low levels across North America. But I think you have to understand that certain severe viral infections can look like bacterial sepsis. Thank you, Ed. So we're at the end of our program. I just want to share with you to earn your credit. You go to my SCCM.org and select my learning on the green bar. Click on the 50th Critical Care Congress, and you have up till March 31st to claim your CME credit. I want to thank both of our speakers. I think you did an amazing job providing the examples of how we apply this. And then if you want to claim your CME credit, you can also download a PDF that will just describe the steps I just told you. So with that, I thank both of you. Debbie, I think we should thank you for doing such a great job of moderating and providing also additional content. So kudos to you, Deb. We made a great team. I appreciate it. Us Buckeyes have to stick together. You got it. Thank you very much.

sepsis

pneumonia

diagnostic tools

antibiotic stewardship

rapid diagnostic tests

multiplex panels

pathogen identification

antibiotic therapy

local epidemiology

limitations

microbiology lab

collaboration

stewardship programs