Well, it is 2.30, and so I think we'll go ahead and get started. I'm Trevor Van Schoeneveld. I'm an infectious disease physician here at a critical care conference. Thanks for attending this and not watching the end of probably what's an exciting playoff game. If you care. Some people probably don't care. I'm going to talk a lot about pneumonia panel. I'm going to talk some about other microbial diagnostics in pneumonia. I'm going to talk a little bit about procalcitonin and kind of how they all fit together in the management of pneumonia. So this is all material that I've generated from our use of pneumonia panel, but BioMaria is sponsoring this. These are my objectives. Talk about why we should care about what we do with antibiotics in the ICU. Talk about how we might implement these rapid panels, because I think they are things we really have to do thoughtfully. And then talk about how we can use them to better use our antibiotics. So I always start off these talks because I do stewardship. That's what I do. So we have to talk about antibiotic overuse. You're like, yeah, yeah, yeah. Everybody's always like, yeah, yeah. The ID guys always talk about antibiotic overuse. But we keep talking about it, but nobody's really doing anything about it. So this is some worldwide data looking at antibiotic use. Access antibiotics are like narrow spectrum antibiotics. Watch are like third generation cephalosporin, cefepime, things like that. And it doesn't matter what your income in your country was. They just keep going up. This is some data from the CDC. They did some of these point prevalence surveys. I've run a stewardship program since 2009. And the data in the US is, well, we've changed a little bit what we use. We haven't really changed how much we use. And then what was really discouraging was this paper that was just published a couple years ago looking at the appropriateness was depressingly low. Just depressing for me. All that work. It's a long slog. That's what I tell myself. In the ICU, very important. As you guys all know, lots of people get antibiotics. Retract infection is the most common indication. And depending on how you measure it, about 30% of that is just not needed. And the most common reasons it's not needed, well, we continued broad spectrum therapy when we could have narrowed. There wasn't an infection to start out with. And then we kept it around longer than we needed. And I actually really like this paper that was just published at the end of 2022. What they did in this paper was they tried to see what's the harm of each additional day of antibiotics. I want to keep it around. That's the discussion I often hear. When I call my ICU docs, they say, well, I want to keep it around for another day. You know what? Is that bad? Is that good? Well, so they looked at 71 randomized trials with 23,000 patients and they compared shorter versus longer courses. Seven versus five. Ten versus seven. And they looked at the harm of each additional day of antibiotics. And what they found is about 4% per day, each day, increase in adverse events. Now the question is, does that provide any harm? And I would say the consistent finding in all those shorter course studies is there's no benefit. It's all just harm. And so we as infectious disease doctors keep moving that course shorter and shorter and shorter. Now that's not what this talk is about. That'd be a separate talk. We're going to talk about pneumonia. And we have these two, I think, really useful guidelines that help us to know how should we manage pneumonia. Now one thing I would point out about these guidelines is they don't actually talk about how to diagnose pneumonia. They talk about what tests you should do when someone has pneumonia, what management strategies, what antibiotics. They have all kinds of great advice once you're really confident somebody has pneumonia. The question is, how often are we really confident somebody has pneumonia? I'd say that's uncommon. Often it's, well, I think it might be pneumonia, but it might be this or that or 14 other things. And then after a couple of days, you kind of figure it out. And the reason they didn't go into this is because pneumonia is a compatible clinical syndrome plus an infiltrate on imaging. That is the gold standard for pneumonia diagnosis. If you read the studies, that's what they do. Do they have symptoms? They have imaging findings. And there are symptoms like cough and chest pain and fevers and exam findings and biomarkers like procalcitonin and white count and imaging changes. And the problem is they're poorly predictive for pneumonia. If you had fevers, crackles, hypoxia, and tachycardia, which, boy, that sounds really concerning for pneumonia, only 57% positive predictive value for having an infiltrate on imaging. Now, not everybody with pneumonia has infiltrates on imaging. There was a study just published on that interesting entity of pneumonia without an infiltrate. I don't know. Is that really pneumonia? I don't know. And then we know that overdiagnosis of VAP is common. I mean, in this study, 231 cases, over half of them weren't actually pneumonia. And they got bucket loads of antibiotics that, frankly, aren't going to help. If it's not pneumonia, an antibiotic won't help. And so I give all of this background because we're going to talk about a highly sensitive molecular diagnostic test we're going to do on a non-sterile specimen, which is sputum. And so we have to put that result in the context of pneumonia diagnosis. Now, the other thing in the guidelines, they do have a lot of information on how to use microbial diagnostic testing, sputum culture and antigens and things like that. And they don't mention the multiplex panels because they weren't around, understandable. So we're going to talk about how we've integrated these into our guidelines, what our recommendations at our facility are, and how you might think about using them. Now, the problem and why do we have molecular diagnostics for pneumonia, well, we do a really poor job of diagnosing pneumonia, or at least we have for the last 100 years or however long we've been doing it. It's so poor that the CAP guidelines said, the overall poor yield of sputum evaluation for detecting organism causing CAP limits its impact in management and patient outcomes. And so they say, don't bother in most patients. It's not going to change anything. Just give them the typical antibiotics. And this is the big New England Journal study that tried to define the etiology of CAP, in which they found bacterial causes in only 16 percent. This is some European literature where they consistently find a bit more bacterial causes. They had 25 percent bacterial only and 7 percent bacterial and viral. And so, you know, 15 to 30 percent caused by bacteria. The guidelines would say you give 100 percent of those patients antibiotics. That's what they would say. If you think they have a pneumonia, you give them all antibiotics, knowing that we only detect a bacterial pathogen 15 to 30 percent of the time. Hence the need for better diagnostic tools. So what is the role of microbiologic testing? And this is something that I talk a lot about when I'm on rounds, when I'm with residents, with fellows, is microbiologic tests do not diagnose pneumonia. If you remember my slide earlier, there was no culture on that how do you diagnose pneumonia slide. That is not part of pneumonia diagnosis. Pneumonia diagnosis, do they have symptoms and imaging findings? Now, the microbial diagnostic tests we're going to do help us to define the etiology, but they don't tell us, yes, they have pneumonia or they don't. Now, the guidelines do say for VAP patients, if you don't grow anything on your quantitative, you know, bronchoscopy, you can stop the antibiotics. Now, is that because they didn't have pneumonia? Maybe. Is it because those three days it took to get the culture back is all you needed? Maybe. So, but that's about it. So that's an important thing to understand and some, a conversation I have in the ICU frequently and the residents say, well, they grew this and I'd say, okay, that's great. Do they have pneumonia? That's a clinical diagnosis you and I have to come and make, not something the culture is going to tell us. The culture just tells us what's in the respiratory tract, what's in the ET2. Now, that may mean that's what's causing pneumonia, but it may not be any pneumonia at all. So, again, just trying to lay the foundation for interpreting these tests. What microbial diagnostics do we have? Well, we have blood cultures and we have sputum cultures and bronchoscopy and urine antigens, nasopharyngeal swabs for a variety of viruses. We have multiplex panels for pneumonia. We have the MRSA nasal swab, which I'm not a huge fan of, which I can tell you why later. These are all tools we can use to sort out what do we use antibiotic-wise for pneumonia. What do the guidelines say about these? Well, for CAP, for outpatient, don't bother. Give them antibiotics. If they don't get better, then you can think about it. Same thing for non-severe CAP. The typical antibiotics work most of the time. And unless people have specific risk factors, you really don't need to do any diagnostic testing. Now, if you have somebody with severe CAP, right, they're in the ICU for their CAP, very appropriate to get them respiratory tract cultures or diagnostic testing. If you've treated them with typical antibiotics and they're not getting better, well, we should try and figure out what's going on. If you're giving them broad-spectrum therapy, it's actually recommended you do diagnostic testing, mostly so you can narrow those antibiotics down. If they're immunocompromised, not actually because they have more resistant pathogens, but because they have more various pathogens, viruses, mycobacteria, fungi, more likely to find those in those patients. And then for our patients with risk factors for resistance, you know, they've been on antibiotics recently and you're thinking about starting broad-spectrum or you're worried, very appropriate to do it in those patients. That's for CAP. Not VAP, everybody with HAP and VAP, you should try to get a respiratory tract specimen on. You should do some sort of diagnostics. The guidelines say for VAP, just get an endotracheal aspirate, whatever you can get, just get it. If you want a bronc, that's great, but don't wait, get a respiratory tract diagnostic. Now, which respiratory tract diagnostics, and I have a table here that's going to talk through this. So I want to get some lower respiratory tract specimen. The key question is, can they produce one, right? Lots of patients can't produce one, particularly for HAP or CAP. And so if they can produce sputum, we're generally in places where we think we need respiratory tract diagnostics, say do a pneumonia panel with a culture. If they can't, that's going to move us down the chain of, well, we can't get lower respiratory tract, which would be preferred. So let's do some things. You know, we'll do a respiratory pathogen panel, nasopharyngeal swab, looking for viruses. We'll send some urine antigens, we'll send the blood cultures, you know, but ideally we'd like lower respiratory tract specimen. But most people with, you know, it's only about 40 to 50 percent with CAP who can produce. Most people with HAP don't ever seem to. People with VAP should be 100 percent, right, because you just put a catheter down there and suck some goo out. So this is, I think, where the respiratory pathogen panel for treating pneumonia can have use if you can't get a lower respiratory tract specimen. This is what's in the respiratory pathogen panel 2.1 that Biomir produces. This has COVID. One thing you'll note is in the pneumonia panel, it does not have COVID. And so that's an important differentiator. But what you'll find is when we look at the pneumonia panel, it has like all the same stuff except for the Bordetella, which it exchanges for Legionella. It has paraflu, but it doesn't have like four flavors of paraflu, and you'll see that. So you know, and I don't have time to go into the literature on this, but if you couple this, particularly with biomarkers like PCT and alerts, it turns out you can actually get people to use less antibiotics. You can say, oh, they have a virus and their biomarkers are low and you probably don't need antibiotics for that. And people listen to that some of the time and sometimes they don't. So this is what's in the pneumonia panel. And if you've seen any other biofire panels, you'll see it looks somewhat similar, has Staph aureus, has the marker for methicillin resistance in Staph aureus, pneumococcus, Group A and Group B strep, a host of gram negatives, including some of the common CAP pathogens like H. flu and moricella. It has some gram negative resistance genes. So it has CTXM. CTXM is an extended spectrum beta-lactamase, an enzyme which chews up third generation cephalosporins. That's not the only ESBL gene out there, but it's the most common in the US. And so I think it's useful because it makes up the majority, but it's not all. It also has the carbapenems we encounter, which hopefully you don't encounter too often. But if you do, they really do change how you manage therapy. And then it has over here what's in the respiratory pathogen panel, except without all the details, right? There's just Flu A. There's not Flu A, you know, 2009 H1N1. There's not H3. It just says Flu A, Flu B. And then it has Chlamydia, Legionella and Mycoplasma. And so one of the things you might take away from this is if you do this, you really don't need to do respiratory pathogen panel. You've kind of done it in somebody who has pneumonia. Now, if you've done a respiratory pathogen panel, we've seen this sometimes where we do a nasopharyngeal swab, we don't detect something, we do a large tract specimen, we do detect a virus down there. And we have found that particularly with some of our immunocompromised patients. And so it may be reasonable to do this test, even if you've done a nasopharyngeal swab, if you get a better specimen and the patient has pneumonia. So we implemented this in May of 2020, and our goal was to develop, you know, bring on rapid, highly sensitive tests, improve the diagnosis of the etiology of pneumonia and hopefully deescalate people off of antibiotics faster. And you might say, well, like May of 2020, that was a crazy time. And it was a crazy time. But, you know, we had a lot of people with pneumonia then, right? And a lot of them got started in antibiotics. A lot of them didn't need it. We did, I think, better than some places with starting antibiotics. But once people have been in ICU on the ventilator for 10, 12, 14 days, and then that sort of fluctuating fevers and respiratory status will go up and down and they get put on antibiotics. And we thought a lot of it was probably just COVID. But it's hard to sort out because you couldn't really we really didn't know what the biomarkers were doing then. And so this is a tool to be like, oh, yeah, there's not you know, we don't need bank and cefepime, we can just do ceftraxone. We don't need zosyn and linazolid, you know, we just need linazolid. We sort of narrow that down much more rapidly rather than waiting 72 hours. And so we implemented this back in May of 2020. So I'm going to just show you a couple of cases that I just pulled off a couple of weeks ago. This is the first case just illustrating where I think this can be useful. It's a 66-year-old with CKD stage 4, O2-dependent COPD. He'd been hospitalized a month ago at a different facility, gotten antibiotics for pneumonia, had been sent to a long-term care facility. He came to our ED with confusion. He'd had hypoxia in a fall, was febrile, tachycardic, tachypneic, hypotensive, got unstable, got intubated. As you can see, he was really sick. And this is his chest x-ray. Yeah, he he's got pneumonia, pretty clear cut. And so what kind of pneumonia is this? Well, HCAP is gone, right? HCAP no longer exists and he would be community acquired pneumonia with risk factors resistant pathogens. And so starting him on something like vancomycin, zilocin, lenazolid, things like that would be appropriate because he's been recently hospitalized and gotten more than five days of antibiotics. And he got started on vancomycin. He got a BAL that same day he was admitted, lots of white cells, gram-positive flora. His pneumonia panel came back two days after his gram stain, two hours after his gram stain, 10 to the fifth staph aureus, MEK-A negative, so MSSA. The team stopped the vancomycin after that first dose. The cultures were growing MSSA the next day, and so we called, the stewardship team called and said, hey, why don't you narrate us a phaselin, which is exactly what they did. Patient recovered, got extubated, did well. And so I think this illustrates where we can more rapidly de-escalate and get people on the appropriate therapy. I think it also shows the benefit of stewardship and that the ICU team was very content to say, well, he's on cefepime, it's active, we'll just kind of leave. And we're like, well, you don't really have to. I mean, we are talking about C. diff rates at our hospital. And so, you know, one way to do that is get down to a more narrow spectrum antibiotic. Next case, 43-year-old with CML and a bladder sarcoma, recently had a long hospital stay, getting treated for his malignancy and got a bunch of antibiotics during that hospital stay. Came back to the ED a couple of days after discharge, shortness of breath, respiratory acidosis, got intubated, chest x-ray and consolidations. His Y-cal wasn't that high, but his pro-cal is over five. He got put on lenazolid and zosyn. His BAL showed moderate white cells, gram-negative diplococci, a few gram-positive rods. His pneumonia panel, which came back right after his BAL gram stain, showed 10 to the 6th acinetobacter. And so when the team saw that, we didn't actually have to call him for stewardship. They switched him over to meropenem alone, grew acinetobacter, got treated for an acinetobacter pneumonia. And so I think this shows you where you can also not only de-escalate more rapidly, but in that occasional case where you need to escalate, it can help you with that. So what I'm going to do now is talk about some key things to understand about the pneumonia panel if you're going to implement it, because these are really important things to understand and be thoughtful about how you implement. The first thing you're going to find out is you're going to detect way more organisms. So this is the big multicenter validation study over 1,800 BAL and sputum specimens. They detected a lot of pathogens. The most common pathogens were Staph aureus, Pseudomonas, H flu and rhinovirus. What I'd point out right away is that three quarters of the sputum specimens had something detected and half the BALs. The second thing I point out is over half of the sputums had more than one thing detected and over a quarter had more than two. And so you're going to find more than one organism if you implement this. The study over there basically looked at a subset of those BALs where they had clinical data and they found a 63 percent increase in detection of organisms on their BALs, a number of which weren't detected on the standard of care cultures. Now, how sensitive and specific is it, which is always a question when you talk about diagnostic assays? Well, the bottom line is it's very sensitive for what it's designed to detect. Now, in the study, their gold standard was if it doesn't grow at least 10 to the 3.5 coniform units per ml, that's a negative culture. And so that was their gold standard. The sensitivity, because they looked at each individual assay, ranged a bit. But overall, they only had 16 false negatives out of 1,682 tests, 1,682 panels. So they don't miss a lot except for things that aren't on the panel. Right. And I have a slide on that here in a minute. Now, the specificity or the false positives is, I think, a little more of an issue. They have 875 results that were positive that they didn't detect using that gold standard of 10 to the 3.5. Most of those are Staph aureus, H. flu, Moraxella and Pseudomonas. Some of those are colonizers, could be from nasopharynx. Some of them are hard to grow like H. flu. When they went and did more digging, a quarter of those bacteria, those 875, they were there, but they were below 10 to the 3.5. Basically, the other three quarters, other than three, there were only three they couldn't explain. The other three quarters were detected by other molecular methods. And so we have as a test that is significantly more sensitive than what our gold standard is. Right. Our gold standard is culture and it detects bacteria down to a certain level. But now we have an amplified DNA test that detects a whole bunch more. And so it's a highly sensitive molecular test being done on a non-sterile specimen. And so it's going to detect more pathogens. And some of those pathogens are going to be like, what do I do with that? I'm not sure what that means. So this is, I think, one of the better comparative studies. So this was a study where they looked at cat patients are coming to the ED and they compared pneumonia panel plus sputum culture and blood cultures to this very extensive panel that they were doing for their patients. And so they were doing blood cultures and they were doing sputum cultures and they're doing restored pathogen panels and they're doing nasopharyngeal swabs for Staph aureus and pneumococcus and urine antigens. And then they just compared that and they did pneumonia panels and they said, well, what would have been the difference? What they found is with pneumonia panel, better detection of all the pathogens. And by the way, it was cheaper and easier. So this is probably the, you know, this is way more than the guidelines would ever say you should do. So it's better than the most extensive sort of baseline, I think, diagnostic testing we can do. Now, the next thing to understand is you're not going to detect everything on culture that you found on the panel. And so you're going to have to think about what does that mean? So the way this is reported is the viral atypical resistance are detected, not detected. Flu detected, not detected. Legionella detected, not detected. But the bacteria Staph aureus, Strep pneumo, E. coli are reported as DNA copies per ml. And so they get grouped into what they call bins, 10 to the 4th and the 5th and the 6th or into the 7th or greater. So that's DNA copies that does not equate to colonies per ml in any way, shape or form. The number of DNA copies is significantly greater than the colonies per ml. And what I would point out here. So this is about this was from this. This is part of that validation study. This is some of our data. Anything below 10 to the 6th, you're probably not going to detect it because there's very low levels of those pathogens. But you might once you get over 10 to the 6th, 10 to the 7th, you're more likely to detect it. But still, even 10 to the 7th, you know, we're about 60 percent. The validation study is about 75 percent. You're not going to find some of those bugs. Why is that? Is that because they already got on antibiotics and all the bacteria are dead? Is that because even though there's a lot of DNA copies, there's just not that many bacteria. Is it because of the bacteria? It's not easy to grow like H. Flu? The answer to that is probably yes. All of those different factors in there. So it's important to understand you're going to find things on the panel you aren't going to grow. And so you need to think about what am I going to do about that? Next, it only design it detects what it's designed to detect, so it's not on the panel. It's obviously not going to find it. This was a study that looked at a bunch of B.A.L. and speed of specimens from a facility in the pink are organisms detected by the panel only in the green are both panel and culture detected. And what you'll see here is for an organism like H. Flu, mostly was just detected by the panel. And for H. Flu over here, you can see this is by DNA copies. It wasn't until you got to 10 to the seventh that you're even able to pick up much of that H. Flu on culture. Now, Staph aureus here, you can see, again, a lot more detected by molecular panel alone, about as many molecular panels as panel and culture. And you can see once you get up 10 to the sixth, it's a little easier to find. And if you talk to the micro people, they say, yes, pretty easy to find Staph aureus in a culture. And but what I like about this, they showed what wasn't detected. So it's like steno and citrobacter and a few other things. We out of fifteen hundred panels had fifty five organisms that weren't detected. The most common things we didn't detect by the panel that grew were crinobacterium striatum. We can debate about whether that's significant. Some people probably is, others it isn't. Steno, kind of the same story. And then citrobacter. And so the others were just a mishmash of individual things. And so those were our top three not detected. And so I think taking it into the context of the patient, what grew before, boy, if I have just total gram-positive rods in my gram stain, maybe I should think about crinobacterium even if the panel's negative, if I really think they have pneumonia. All right, next thing. You want to make sure you have a good sample and you always want a couple with culture. So bad sputum is bad sputum, right? We don't want to, like, run molecular panels on this. This is spit. This gets rejected in our micro lab. So if you send something down that looks like this, you get a gram stain, you don't get a culture, you don't get a pneumonia panel. If you're going to do this, you should use the same standards you do for sputum quality. The other thing is you never want to do a molecular panel without a culture because you want to know the phenotypic susceptibility. And so we built it so you literally can't. Like, you can't order a pneumonia panel without a culture. Now you could do sputum or if, you know, it's one of our CF patients or lung transplant patients which you do a little different culture, you can do that. But you just can't order it without a culture because we think you need a culture. And so you should try to build some of these same quality things into how you use pneumonia panels. Next, you should use it where it provides value and don't like where it doesn't have value, which is sort of a no-duh. And so that's why this is how we built it into our guidelines, right? So we said for non-severe CAP, yeah, you don't need a pneumonia panel. You don't need anything. You need COVID testing because that's what everybody needs. But other than that, you just give them the antibiotics. And if they don't get better, then you can think about it. In severe CAP, like we talked about, we do recommend pneumonia panel. If you can't get it, that's where the urine antigens and the respiratory pathogen panel can be useful. Same thing for people with, you know, an MDRO therapy. Same thing, you know, compromise. And then for HAP-VAP, yeah, you should really just do pneumonia panel. And so we built this right into our pneumonia guidelines on what to do. The other thing is because it's kind of a complicated test, we did put some restrictions on it. We restricted it to the ICU. So if you're in the ICU, you can go and order it because that's where all the VAPs are going to be. That's where all the severe CAPs should be. That's where a lot of the HAPs end up. And so that seems like an appropriate place for using a pneumonia panel. Outside the ICU, we did restrict it to our ID physicians and pulmonary critical care docs. We also said don't repeat it. You know, because once you've done a highly sensitive molecular test, you know, you kind of know it's there. You don't need to do it again. Now, we do have a little difference there between sputum and BAL, and that's because, you know, sometimes they'll get a sputum, and then somebody will get intubated, and they'll want to do bronchoscopy. And, you know, this seems reasonable. I'm okay with doing the panel again there. Might get a little better specimen. We could argue about that, but, you know, I'm not going to argue a lot. So we built a little different. But, you know, if you've gotten a highly sensitive molecular test, you don't need another one. Next, I think it's important to present the results in a way that helps interpretation. I think all microbiologic testing should be this way. And people in my field in clinical micro are increasingly saying that. This is how it comes off the actual machine. Very nice pictorial representation. You know, how much flu, how much pneumo is there. You know, these other things aren't detected. I wish ours looked like this. When I show this slide, I'm just embarrassed at how it's very sad how we present it. It makes me want to cry. It's just gross. And this is because our lab information system and the way that it integrates with our electronic record is that I literally can't take MEKA and put it by Staph aureus. You'd say, boy, that makes a lot of sense to put them. Nope, can't do it. Sorry. Computer wins. So this is a pretty typical result. Well, maybe not super typical, but, you know, greater than 10 million pseudomonas, a million CLEB. This is a patient of ours who's indebated for, like, two weeks with COVID. This is about two weeks into hospital stay. 100,000 MSSA and 10,000 moraxella. What grew on the culture, pseudomonas and CLEB. Didn't even find the Staph aureus, didn't find the moraxella. Not surprising, the levels they were detected. And I think it's just grotesque we have to report all the zeros. I would love to run a study where you report zeros and report just exponents. I think people are more impressed by zeros. That's just my opinion. I think there's a lot of interesting things you could do with how we report results and how physicians interpret them. So, next. I think it's super important you help people know what to do with the results. This could be a two-edged sword. This could be your worst enemy if you do stewardship. Right? But if you do it right, I think it can be really helpful. You want to help people know what to do with the results. And so we published a whole guideline. It's like eight pages long. You're welcome to go to our Nebraska Medicine stewardship page. Read it. It's available to everybody. In there, I wrote a bunch on how to use it, how to interpret it. And then what we did was write guidance based on each marker. Now, I couldn't write a guideline for what to do with that panel I showed you with four different things of different levels. I couldn't even tell you what to do if you have 10 to the 7th Staph aureus versus 10 to the 4th. Because, again, you have to put it into context. But if you don't have Staph aureus detected, I can really help you with what to do with vanco. I can tell you to stop it. Because if you don't have Staph aureus, you don't have MRSA, and we don't need the extra doses of vancomycin. If you don't have the MEK-A gene and Staph aureus is detected, again, we don't need the vancomycin. We can just get rid of it. We don't have the mark of resistance. Now, if we do have the MEK-A gene detected and you do have MRSA, we do say, well, it's reasonable to give some anti-MRSA therapy. Now, again, I didn't say you have 10 to the 4th MRSA and 10 to the 7th Pseudomonas and its Gram-negatives on the Gram stain and no Staph on the Gram stain. But, you know, I'm not going to argue. But if you don't grow it, then I'm going to say, hey, you didn't grow it. You wouldn't have treated it in the first place. You got all this other stuff. Why don't you treat that and forget that little tiny skosh of MRSA that probably doesn't have any role? And so that's how we've kind of built the guideline for Staph aureus. This is one of the places where it's really useful, and I'll show you some of our data that supports that, that it really does help people to get off of vancomycin faster. So they don't get two, three, four doses. They get one dose, and then we're done. Things like pneumococcus. Boy, if you know they have pneumococcus, you don't need more. You give them penicillin, high-dose penicillin. You're great. Same thing with their beta-hemolytic strep, haemophilus influenza. You can rapidly de-escalate these people off of their vancocephalopene. Now, the gram-negatives, it's more complicated, right? Because my E. coli's may not look like your E. coli's. Even my E. coli's for patients who have recently been hospitalized look very different than the E. coli's after a month in our surgical ICU. They just, they look different. And so we tried to put some antibiogram data in here. You know, some of this is overall antibiogram data. Some of it is restorative tract antibiogram data, which looks terrible because everybody who gets a restorative tract culture has usually been in the hospital for a week or two. And then we gave some guidance here, and it differs by pathogen. And so, you know, if you have Enterobacter, cefepime's probably the, you know, the agent you're going to reach for. And maybe you don't grow it, but it's the only thing. And so we have some de-escalation options. This is how we integrated CTXM. So it's not the only ESBL, but it's the most common. And so knowing that, you know, ceftriaxone is active about 80% to 90% of the time, if we knock out 75% of the ceftriaxone-resistant isolates by saying there's no CTXM, that means it's like 88%, 95% active. And so you can feel pretty comfortable using a ceftriaxone or piptazo. Now, if you have CTXM positive, you're probably going to need to reach for a carbapenem and wait for the phenotypic susceptibilities. So we've tried to build this to give clinicians advice, and this is actually what we as stewards use when we call teams and give them advice. Because I wanted to know what to tell people. So, you know, we built this so we would know what to tell people. All right. I think it's useful to couple it with other tools for pneumonia. And I just want to talk a little bit about procalcitonin. So this is a case of a patient I saw in service, 30-year-old guy, diabetic, came in a fever, cough, chills, nausea, vomiting. He literally said he felt like a truck ran over him. His Y count was normal. He had this maybe something on his chest x-ray, and he was rapid flu B positive. And so when I show this case, I ask people, would you give antibiotics? And I'd say about three-quarters of the time people say no, at least at my institution. But about a quarter of the time they say yes. And it depends on how you would interpret his chest x-ray. Does he have pneumonia or not? If he has pneumonia, the guidelines would say 100% need antibiotics. Now, what if I told you his procal is 0.75? You're like, huh, something a little funny going on here. And something funny was going on here because he had group A strep that grew in his blood about 10 hours later. And so he had group A strep pneumonia, and his chest x-ray looked a bit worse the next day. But actually our medicine team started him on antibiotics because his procal was high. Now, if his procal had been 0.1, I'd have said, boy, you really don't need antibiotics. And let's just watch him. If you're worried, we'll repeat one. But I think this illustrates in cases of, well, I'm not sure of this pneumonia. Do we need to give antibiotics or not? And I think procalstonid is a nice complement to the pneumonia panel. And so to really briefly summarize, you guys probably all know about procalstonid. It's a biomarker. It's produced ubiquitously by cytokines stimulated by bacterial infection, particularly apopolysaccharide. It is the most specific biomarker we have for bacterial infection. It's not perfect. There is no perfect biomarker out there. The levels rise rapidly, generally decline predictably, as long as you've controlled the infection or treated the infection. It's not impaired by neutropenia or other immunosuppressive conditions. But there are a bunch of things that happen, particularly in the ICU, that make it go up, that you do have to be aware of. Major surgery, trauma, cardiogenic shock, any sort of shock. Doing CPR on somebody makes your procalstonid go up. And then if you remove that, they're not being traumatized anymore. It should decline. If they're in persistent shock, their procalstonid is going to stay up. Renal dysfunction will give you some elevation in procalstonid. Something I had to add to my sort of list of things, MISC and A will make this go up because they make all your biomarkers go up. And then certain agents that stimulate cytokines, particularly some of the induction agents we use for transplant, can make this go up. And so that's sort of the short version of procalstonid. There's tons of literature. I've been reading procalstonid literature since I was a fellow, which was like 15 years ago. And there were a bunch of randomized trials when I was a fellow. And so most of the literature is in respiratory tract infection and sepsis. We're all relatively familiar with this. But I just want to highlight three studies that I think help to put the context around procalstonid. Because when I go talk to my colleagues, they don't like procalstonid. They're like, no, no, no, that's useless. I don't like procalstonid. It's really not built for the ID doctor. I don't order a lot of procalstonids. But I think they're really useful for a hospitalist. And I think they're useful in the ICU. And I think this literature shows why. So this is a big patient-level meta-analysis. 6,700 patients, 26 trials, 12 countries. HAP, VAP, CAP, bronchitis, COPD mostly, but a little bit of upper respiratory tract infection. Mostly in the ED and the ICU. What they found is procalstonid use decreases antibiotics. That's not terribly surprising if you'd read those randomized trials. Be like, yeah, they all showed less antibiotics. What was interesting about this study was the mortality benefit. Procalstonid-based management strategies of lower respiratory tract infections decreased mortality. And almost decreased treatment failure. Not quite. And they decreased antibiotic-related side effects. And so I personally think this suggests that procalstonid-based management of lower respiratory tract infections, which that's what pneumonia is, is not just safe, but it's probably better for our patients. That's what this literature would suggest. I think there's some similar literature in sepsis. Two recent randomized trials in sepsis. The SAP study. It's hard for people when they come in with sepsis not to give antibiotics, right? The guidelines say you should give antibiotics. They're a little more moderate now. And they give you a couple hours of time window if you're not 100% sure that they have sepsis. And so they say, well, you can think about it, but you should probably give anybody who's really sick antibiotics. And I don't disagree with that. What this study looked at was early discontinuation. And so they randomized people to early discontinuation based on procalcitonin or just do what you normally do. And they enrolled ICU patients with presumed infection. And they did it in Dutch ICUs, which is important to understand because they're relatively conservative with their antibiotics in Dutch ICUs. As pointed out by the median duration of therapy and standard of care being seven days. So they really do give seven days of antibiotics for sepsis in Dutch ICUs. What they had was a PCT-guided stopping rule. If it was less than 0.5 or 80% decline from peak, they said stop. They had about 60% compliance at 24 hours, about 90% compliance at 48 hours. What they found is less antibiotic use. Median duration of therapy dropped to five days, less defined daily doses. But what was really interesting out of this study was a 5.4% absolute reduction in mortality in sepsis patients. So this study would suggest that procalcitonin-based early antibiotic discontinuation not only uses less antibiotics, but actually decreases mortality in sepsis. Which is really interesting because there's not a ton of therapies out there that decrease mortality in sepsis. You might say, well, that's just a fluke, one study. Where's the second study? A progress study. Similar style. Here they enrolled sepsis patients, lower serotract infections, pylo and bloodstream infection. Again, PCT-guided early discontinuation. Very similar rule, 0.5 or 80% decline versus usual care. They had 77% compliance with the stopping rule, and so the median duration of therapy dropped from 10 to 5. Now, in this study, they wanted a more long-term look, and so they looked at C. diff and multidrug-resistant infections and isolation of multidrug-resistant organisms out to 180 days. And the PCT-guided group had a lower infection-associated adverse outcomes. Now, most of the adverse outcomes are lower, although a lot of them are rare and so didn't have significant p-values. But the one that did was 28-day survival, which was lower in the procalstonid-guided group. And so now we have a second randomized trial suggesting that procalstonid-guided early discontinuation of antibiotics and sepsis doesn't just use less antibiotics, it actually improves patient outcomes. So again, this is just data. I think procalstonid is a tool to use when using the pneumonia panel. The pneumonia panel says this. I'm really skeptical they have pneumonia. The procal is low. I did another one. It's low again. Boy, they don't need antibiotics. I don't care if they have 10 to the gabillionth E. coli. If I don't think they have pneumonia, I'm not going to give them antibiotics. All right. Last question. Does it actually change antibiotic use? And this is where there's not a ton of literature out there. I'm going to show you one published study. This was an early report at ID Week from the U.K. guys on their large randomized trial. And I'm going to show you we did a retrospective study, which we've not published yet. And so you can take it with a grain of salt because we haven't done all the analysis yet either. So this is one randomized trial. This is from the U.K. single center. They randomized them to usual care for pneumonia. And this is ICU patients with pneumonia, sputum culture versus pneumonia panel with culture. And they did ProCal in both groups. Everybody had already been on antibiotics. And then with the pneumonia panel group, they actually gave some advice on what to do because it was a new test. This study got stopped early, not because they'd reached their end point, but because COVID was happening and they're like, we don't have the resources. The groups were relatively well matched. Sofas were five to six. About half the people were on the ventilator. About 40% cap. They had a little more VAP in the pneumonia panel group. COVID infection was about 12 to 15%. ProCals were generally in the we should give some antibiotics for pneumonia range. Time to micro result, not surprisingly, way faster, 1.7 hours. Increased detection of pathogens went from 50% to 70%. And the outcome they looked at was time to results directed antimicrobial therapy, i.e. I switched the antibiotics based on what we found. And so in the control group, I think it's actually surprisingly low, only about 20%. In the pneumonia panel group, it was more like 80%. And it happened much faster. Results directed therapy went from 8% to 42% and happened about two days faster. Escalation actually happened more often with pneumonia panel and was much faster. There was no difference in how long people were treated. And there was no difference in clinical outcomes, although they certainly weren't powered to detect any of those clinical outcomes. So I think this suggests what we'd expect. Detected more pathogens, made adjustments faster, and people got more results that helped them make those adjustments. All right. The question we are trying to answer is, does this impact antibiotic use in the ICU? And then the second question is, do we need to like call everybody whenever these come back and tell them what to do? Because that takes a bunch of time and a bunch of work. And so is it worth our time to do that? Or once physicians are comfortable with this, can we just kind of turn them loose and they do their thing? And so we designed sort of a pre-post study. So the pre is pre-pneumonia panel and unfortunately pre-COVID, but there was just no way to do this without having COVID be part of it. And so pneumonia panel with aggressive stewardship intervention is our second six-month period. And then we actually stopped doing stewardship. We just quit. We're like, you guys do your pneumonia panels. We're not doing anything. You can adjust if you want. We'll call you if we were going to call you because of bacteremia or something else, but you do your thing. And so I'm going to present the first two parts of that. We're still analyzing and sort of validating that third part, and so I can't present that data to you yet. So this is just our data. I'm going to show it real quick. So median age is relatively similar. Percent male is about the same. Which ICU went to is different, but that's because we added a dedicated COVID ICU. So we're going to have different. We also had a bunch of COVID patients in this. We had about 16% COVID patients. People were in the hospital longer, and they had higher mortality. And so I think one of the problems we had, these patient groups are kind of different. But I didn't really have a choice because the pre-period was before COVID, and we implemented it as COVID. And so this is where we're going to have to do some matching and some more statistical analysis here to try to match this. The second thing I'd point out is these patients didn't necessarily have pneumonia. We did not validate that they had pneumonia. We just looked at somebody ordered a respiratory tract culture or somebody ordered a pneumonia panel. And it's interesting what happens sometimes. The respiratory tract culture shows no MRSA. There's not a respiratory tract infection, but the team says, oh, the culture doesn't have MRSA. We'll stop the vancomycin. Okay, well, that's fine with me. I'm doing stewardship. I'm content with that outcome. But it's interesting how sometimes cultures are utilized. ID console was about the same. We had more VAPs here, less CAPs, PIT scores. We measured PIT scores, and they were about the same. So this is what we found. People used less respiratory pathogens, not surprising. People used less Legionell antigens, less pneumococcal antigens. We did detect more pathogens, interestingly. And so, again, suggesting there's a little bit of difference in the patient groups. These are the pathogens we detected. A lot of staph, H flu, pseudomonas. We had 57% of the pneumonia panels positive. We had a median of one organism detected, a mean of about 1.6. The max was actually six things detected on the panel. We only had one Legionella. We don't see a lot of Legionella. We didn't have a lot of viruses. This was during COVID. We didn't have a lot of viruses, other than COVID. No carbametamases, two CTXMs. We did review a lot of these. We didn't review them 24-7. We got to about 80% of them. Our reviews, we did do some stewardship reviews in the control patients, not just because we happened to. The reviews did happen earlier. We only intervened where we thought there needed to be an intervention, which was about a third of the time and about 90% acceptance. Mostly we're recommending, hey, why don't you de-escalate, or why don't you stop? We don't even think they have pneumonia. And we're relatively successful. We did have some escalate. These are the antibiotic outcomes. Time to antibiotic de-escalation, first de-escalation, 18 hours earlier. Time to stopping MTMRSA therapy, 11 hours earlier. Time to stopping any Pseudomona was, I think, about 16 hours earlier, although not quite statistically significant. Time to effective therapy didn't change. People got one day less of antibiotics. And then what did they stop? Interestingly, vanco, azithro, and metronidazole, which why metronidazole? I think a lot of people think we need it for aspiration pneumonias. And a bunch of those calls were like, hey, you think they have pneumonia. You don't need metronidazole. You have moncephamine. That's adequate for oral flora. You don't need it. And so I think the question is really going to be how much of this was us doing stewardship, and how much of this is the pneumonia panel. That's what part three of this is going to tell us. My guess is things like vanco and azithro stopping, probably going to see that continue. I don't know if we'll see some of this. So our conclusion was, seems to be a useful tool, but we'll see what the stewardship intervention does. So final thoughts. I think when you're going to use any microbial diagnostic, it's important to be thoughtful in how you use it. The pneumonia panel is a highly sensitive test. It's going to detect a bunch of pathogens. And you're going to have to be thoughtful about how you implement it. You need to couple with culture. If it's a bad sputum, it's a bad sputum. Don't do a panel on it. You're going to think about what do we do with detections that aren't found in the culture. You don't want to do redundant tests. Don't do this, antibreast period pathogen panel and urine antigens. Not really useful. Only use it where you're going to change therapy, right? What does a pneumonia panel use if they have an intra-abdominal infection? I ask my surgeons that sometimes. They're like, well, we didn't really know. And so we just kind of did everything. I'm like, OK, all right. All right. I think you should present the way that improves interpretation. Don't present it the way we do. That's gross and unseemly. Present it in a useful, informative way. That's what the guidelines would recommend. I think it's super important to help people understand, right? You don't want to introduce a complex test like this and just be like, yep, go for it. Do whatever you want. You want to have some guidance. You want to help people. You want to nudge them. I think it's very important to test like this. And you want to couple it with other tools, right? Use biomarkers, things like that. I think they can be useful with this. And with that, I will stop and take any questions. Thanks a lot.

pneumonia panels

molecular diagnostic tests

antibiotic stewardship

etiology of pneumonia

de-escalation

escalation of antibiotic therapy

limitations of pneumonia panels

specimen quality