Hello, and welcome to today's Journal Club Critical Care Medicine webcast. The webcast, hosted and supported by the Society of Critical Care Medicine, is part of the Journal Club Critical Care Medicine series. In today's webcast, we feature two articles from the January issue of Critical Care Medicine. The webcast will be available to all registrants on demand within five business days. Simply log into mysccm.org and navigate to the My Learning tab. Hello, my name is Tony Gerlach, and I'm a clinical pharmacist at The Ohio State University Medical Center here in Columbus, Ohio, and I will be the moderator for today's webcast. Thank you for joining us. Just a few housekeeping items before we get started. First, during the presentation, you'll have the opportunity to participate in several interactive polls. When you see a poll, simply click on the bubble next to your choice. Second, there will be a question and answer session at the conclusion of the presentation. To submit questions throughout the presentation, type into the question box located on your control panel. Third, if you have a comment to share during the presentation, you may use the question box for that as well. And finally, everyone joining us for today's webcast will receive a follow-up email that will include an evaluation. Please take five minutes to complete the evaluation, as your feedback is greatly appreciated. Please note, this presentation is for education purposes only. The materials presented is intended to represent an approach, view, statement, or opinion of the presenter, which may be helpful to others. The views and opinions expressed herein are those of the presenter and do not necessarily reflect the opinions or views of SCCM. SCCM does not recommend or endorse any specific test, physician, product, procedure, opinion, or other information that may be mentioned. And now I would like to introduce today's two presenters. First, we have Dr. Prateek Sehgal, who is currently a Critical Care Medicine Fellow at the University of Toronto. He completed medical school as well as his residency specialization in internal medicine at the University of Toronto. His academic interests include undergraduate medical education, postgraduate teaching, infectious disease with critical care environment, and aiming to better understand long-term outcomes of critical care. Second, we have Dr. Diana Verboom, who received her medical training at the Erasmus University in Rotterdam. She worked as a PhD student at the Julius Centre of UMC in Utrecht and her doctorate research resulted in a thesis called Novel Approaches for the Diagnosing Sepsis in the Critically Ill. It was performed within the Molecular Diagnostics and Risk Stratification of Sepsis or MARS project, a collaboration that aims to assist physicians at the bedside in tailoring the treatment of an individual patient suffering from sepsis by generating rapid molecular information about the causative pathogen and the host response. And just this month in January 2021, she started a residency in internal medicine in Rotterdam. Thank you both for joining us today. Before we begin, do any of you have any disclosures tonight? Preetat? Nope. Diana? No, I don't. Perfect. Thank you both. And I will turn everything over to Preetat for our first presentation. Thank you so much, Tony. So as mentioned, I'm a critical care fellow currently working at the University of Toronto in my last year of fellowship. This is a project I did at Sunnybrook Health Sciences Centre with Dr. Nick Dannemann, a infectious disease consultant, along with the excellent antimicrobial stewardship team at Sunnybrook Hospital. So first I wanted to kind of talk about the different models that are available for antimicrobial stewardship within the ICU. And there's significant amount of variance, both within us, within my own city, and just worldwide on how stewardship is approached in the ICU. One of the common models is a physician-led model. And again, there's a couple of different ways that this is approached. One is the infectious disease physician rounds with the ICU team itself and provides real-time feedback on recommendations for antibiotics and critical ailments. It's a bit of a more cumbersome approach, but it is utilized in some centres. Another one that we have in one of our hospitals is, at the end of the week, the ID consultant who's on for the week reviews the ICU census and provides more of a holistic feedback on antibiotic use within the ICU. And this is not so much micromanaging what antibiotic is prescribed when, it's more kind of getting a general idea of trends that are observed over the course of the week for antibiotic prescriptions. Probably the most common model that I've read about and seen is a pharmacist-led model, where you can have either just a clinical pharmacist or a pharmacist specialized within antimicrobial stewardship reviewing the ICU census in real-time to decide on antibiotic suggestions. This can be either with review with an infectious disease physician or without review and direct feedback from the pharmacist. And like I said, I think this is probably the most common model that I've seen at least. This is not so much a model of stewardship as it is more of a philosophy on antibiotic use. Some centers do restrict which antibiotics are allowed to be prescribed without consultation from an infectious disease physician. And some centers will decide on only targeted antibiotics to review from the stewardship team, as opposed to every single antibiotic that's prescribed in the hospital or in the ICU, which can be very, very time-consuming, as you can imagine. And then the feedback frequency has a lot of variance as well, whether feedback is provided in real-time, such as at the end of the day or end of every third day, versus more of an intermittent type of feedback. So the question I wanted to get a sense of is, what type of model does your ICU have for antimicrobial stewardship? Is it an ID physician-led model, an ID pharmacist-led model, an ICU pharmacist-led model, which doesn't necessarily have an infectious disease or a stewardship pharmacist associated with it? Or is there currently no antimicrobial stewardship at your ICU? I'll wait for a couple of votes to come in. Okay. So a bit of a split. Yeah. So as I kind of figured, I didn't think the ID physician-led model is too common. Glad to see the ICU pharmacists are very involved, which is, you know, of course, they're there every day and kind of get a better sense of what's going on. And it looks like a fair bit of people don't currently have a stewardship model, at least in the ICU. Thank you. Okay. So, you know, there are a number of barriers that are present to stewardship implementation that I think we've identified. One is just having the personnel available within the hospital. Now, this could be the ICU pharmacists, but more and more as medicine's getting specialized, there's an increasing number of stewardship pharmacists and infectious disease physicians who are focused towards stewardship. And you really need to set time aside to focus on stewardship, either in the ICU or on the ward. And, you know, there really needs to be a system and infrastructure in place at the hospital for this to be present. The next big thing is priority. Does a small community hospital that serves a couple of thousand patients really need a stewardship program? Maybe not. Maybe that's not their primary focus and maybe they don't have as much of a concern about antibiotic use. A large tertiary care academic center, though, where you have many more complex patients, more antibiotic use, more broad spectrum antibiotic use, maybe there there's a bit more of a focus towards implementing stewardship. And finally, cost is another consideration, at least upfront. You would have to pay an infectious disease physician to assess these patients outside of their already busy workdays. And hiring stewardship pharmacists upfront can be potentially expensive endeavor, at least in the short term. Despite this, though, there are, I feel like there is a very strong place for stewardship in the ICU. And there's a couple of outcomes that have been shown to be improved. The broad spectrum antibiotic use definitely does decline after the implementation of stewardship. This is actually a study that was done back in 2012 at Sunnybrook, our hospital, which showed that the implementation of stewardship did reduce broad spectrum antibiotic use, specifically carbapenem use, which is, of course, a big concern with resistance moving forward in the next several decades. Along with that, there was a decline in C. diff rates as well, associated with stewardship implementation in the ICU. And this is an important outcome for morbidity. It may not affect the mortality for patients, but it definitely can prolong hospital stays and lead to further complications down the line and longer ICU stays. And not just C. diff rates may fall, but also overall multidrug resistant rates within the hospital might fall as well with implementation of stewardship, though there's a lot of confounding factors that might be present, such as community spread of multidrug resistant organisms, which may not be directly impacted by stewardship in an intensive care environment. Cost I think actually will go down in the long run if you're able to show that you can reduce these morbidity associated outcomes and use a broad spectrum antibiotics. So while the upfront cost of stewardship might be a little steep, the long term outcomes, though, I believe it will pan out for institutions. Though it's hard to get financial information, particularly in Canada about this. And the major outcome that everyone, you know, that should be the biggest focus is, of course, mortality. And when you're implementing stewardship, what I feel is that you're implementing an intervention that is restricting use of broad spectrum antibiotics. So I wouldn't necessarily think that would improve mortality, but a systematic review that was done in 2019 showed that the implementation of stewardship in ICUs did not affect mortality, which I think is a important positive outcome. So there's a number of questions that remain, which we aimed to tackle in our study. What are the long term trends and suggestion and acceptance rates of these stewardship suggestions in the ICU? What predicts whether a stewardship team will make a suggestion to the ICU team? And then what predicts whether the ICU team will ultimately accept that suggestion made by the stewardship team? So this was a study taken at Sunnybrook Health Sciences Centre, which is a 627 bed capacity in midtown Toronto, Ontario, Canada. It's a tertiary care centre with a number of specialized care populations, including trauma, cardiovascular surgery, burns, neurosurgery, stroke, and oncology, in addition to general medicine. As of right now, there are three Level 3 ICUs, which are essentially intensive care units that can take mechanically ventilated patients. And this is a medical surgical slash trauma ICU, a cardiovascular ICU, and burns ICU, making a total of 48 beds at the time of this study. There's also two Level 2 ICU beds, which can take patients with vasopressors, but not mechanically ventilated. And this is 21 beds at the time of this study. Both of these have since expanded. In terms of the model at Sunnybrook, this was established in 2010 by one of our stewardship pharmacists, Miriam, along with our infectious disease team. And after a proof of concept, it was brought into the ward. As of right now, there are two full-time stewardship pharmacists that do dedicated stewardship only. There's one part-time stewardship pharmacist throughout the year, and a number of stewardship pharmacist trainees who rotate through as well. All cases are reviewed by the infectious disease consultant on call for the week. And there's generally unrestricted antibiotic use, except for these specific antibiotics, linizolid, daptomycin, ticlocycline, fourth and fifth generation cephalosporins, and colistin, which do require infectious disease consult. Anytime an infectious disease consult team is involved in the care of the patient, the patient is not assessed by the stewardship team. So in terms of how this would look like day-to-day, if the ICU team would go ahead and order an antibiotic, then the next question is whether this antibiotic is on the screening list of our stewardship team. And the ones that the stewardship team is targeting is third generation cephalosporins, diprocybin, tasovactam, carbapenem, fluoroquinolones, aminoglycosides, and IV vancomycin. So antibiotics that are either very broad spectrum or potentially associated with toxicities, such as aminoglycosides. If the antibiotic prescribed, such as if it was ampicillin or Ancep, was not on the screening list, this would not trigger a stewardship review, and the stewardship team would not even know that this patient exists. If the antibiotic is on the screening list, you get three days of basically free antibiotic use without intervention from the stewardship team. And the idea here is to allow appropriate blood, sputum, urine, culture sampling, growth of a potential bug, and determination of sensitivities, which is usually available by day three. By day three, this will automatically ping in the pharmacist system, and they will review the case directly through a chart review, as well as with the infectious disease consultant, and then provide either a face-to-face or phone suggestion to the ICU team. This happens, again, on day seven of the antibiotic use, and day 14, if still indicated, but the majority of the interventions are on day three. There's a number of suggestions that are possible. One is no suggestion whatsoever, so continue and change. Broaden coverage, change agent, antibiotic agent, change the dose of the antibiotic, change the formulation, which is almost always a suggestion to narrow from intravenous to oral, discontinue, and narrow coverage. So these are the potential recommendations that will be formally made. The study we did was a retrospective cohort study, looking at all ICU patients that were assessed by the stewardship team over a 10-year time period. There's two databases which we utilized. One is the SPIR database, which is a homemade database at Sunnybrook by Mary Ann, one of the pharmacists who was involved in the study, and this looks at all antibiotic prescriptions and stewardship recommendations, keeps track of them over time. The critical care information system is a provincial database that looks at more nuanced characteristics of patients and their illness severity, length of stay, and ICU level of care that's tracked over time. We wanted to look at basically two outcomes. One is the long-term trends, so how much is stewardship being utilized from a monthly proportion of ICU patients who were admitted and ultimately assessed, to get an idea of are we using more stewardship, less stewardship, or has it been the same over the 10 years? What proportion of suggestions are also being accepted by the ICU team within a 24-hour time period? For our second outcome, we wanted to look at what predicted whether the stewardship team would make a suggestion and what predicted whether it would be accepted. In this, we carried out a multivariable logistic regression with the help of our stats team, looking at a number of different variables which are patient-specific, illness severity-specific, and disease and antibiotic-specific, as listed here. This was input into a logistic regression model to determine, via odds ratio, which of these variables impacted the suggestion and acceptance rates. Looking at our data, what was the most common suggestion? There's a total of 7,749 assessments in 5,569 patient encounters, so if a patient is on multiple antibiotics, both of them would be assessed in one patient encounter, and that would count as two assessments. That's why there's a bit of a discrepancy. Stewardship team suggested a change in approximately 36% of cases. As you can imagine, in the ICU, the most common suggestion was to discontinue antibiotics, and this is almost two-thirds of the recommendations. The next one was, again, more of a restricted suggestion, which was to narrow antibiotics, so that was 14%, and change dose and change frequency were 8%. You can see, very rarely did the stewardship team suggest to broaden antibiotics, only 3% of the time. Most common indications for therapy, an unknown source of sepsis was the most common, just over 20%, community-acquired pneumonia, 15%, approximately, and intra-abdominal sepsis made up the top three. In terms of which antibiotics were assessed by the team, PIP-TASO was the most common, making up about a quarter of the assessments, and for us at least at Toronto, that's really the workhorse antibiotic that we use for patients that deteriorate and get admitted to the ICU. Of course, that can change depending on where the study is carried out. Ceftriaxone, the most common community antibiotic that's utilized, about 20%, and fluoroquinolones. In terms of our general results for Table 1, the mean age of the patient population was 63, predominantly male, which I think is likely attributed to the heavy trauma population that we have in Sunnybrook. Just over half the patients were mechanically ventilated, a third were on vasopressors. Three of the cases were assessed in the Level 3 ICU, and just about half of them were medical surgical versus burns and cardiovascular. In terms of long-term trends, so looking at the proportion of ICU patients that received a stewardship assessment, you can see over the 10 years there was a general increase in stewardship utilization. The mean percentage of patients that were assessed were 14% of ICU patients, and this increased by approximately 1.23% every year. Over the 10-year time period, there has been significant expansion of the stewardship program in the ICU. In terms of the acceptance rates that we looked at, there was a 67% acceptance rate on average. This was actually pretty stagnant throughout the year, a couple of ups and downs, but overall there was no significant change in the acceptance rate over the 10-year period. So then we looked at what predicted whether a stewardship team would make a suggestion. So what first increased the odds of suggestion? So patients that were admitted to cardiovascular or burn surgery ICUs were more likely to receive a stewardship suggestion compared to those admitted to the medical, surgical, and trauma ICUs. Their genital indication of therapy was about an odds ratio of 1.61 compared to bacteremia. And people that received aminoglycosides or vancomycin were significantly more likely to receive a suggestion from the stewardship team to alter the course of the antibiotic compared to a more generic drug like ceftriaxone. A couple of notable variables there. In terms of what made it less likely for the stewardship team to make a suggestion, there was a couple more variables that we identified. So patients that were a bit more critically ill, so ones that were admitted to a level three ICU or were already on vasopressors were less likely to have an intervention proposed by the stewardship team. And what we found was people with primary CNS infectious sources, intra-abdominal infectious sources, and skin and soft tissue infectious sources were much less likely to receive a suggestion compared to patients with bacteremia. There weren't many things that would increase the odds of the ICU team accepting a suggestion. The only one that we did identify was if the patient was admitted to a level three ICU compared to a level two ICU, they were more likely to go ahead and accept whatever suggestion was provided by the stewardship team. And finally, there's a number of variables that actually decreased the likelihood that the ICU team accepted a suggestion. So patients that were admitted to the specialized burn ICU were at an odds ratio of 0.64 compared to medical surgical, so much less likely. People with any kind of pneumonia, community-acquired, hospital-acquired, or ventilator-acquired pneumonia were significantly less likely to accept a suggestion compared to if the patient had bacteremia. An unknown source of infection, again, less likely. And pretty much any suggestion by the stewardship team that may kind of limit the scope of the antibiotic was less likely to be accepted. So a suggestion of narrow, suggestion of change the formulation was usually meant going from IB to oral, and a suggestion to change the agent were all less likely to be accepted by the ICU team. So a couple of key points in our conclusions. So we were able to show that there were actually relatively stable acceptance rates despite an increasing number of assessments over the 10-year time period. And I think this is a positive. Any kind of quality improvement initiative that's implemented generally has some sort of diminishing returns. And I think over a 10-year time period to show a stable acceptance rate, I think, shows really good buy-in by the infectious disease team, the stewardship team, and the ICU team on the program itself. I don't think we should necessarily be aiming for 100% acceptance rates. I don't think that would be necessarily productive. But I think showing that there's general stability and uptake of the program is important. And that's one of the main outcomes that we wanted to identify. And I think part of this was the more face-to-face personal model that we utilized in our stewardship team. And looking at it from when I'm on as an ICU fellow, I really do appreciate being able to talk to the pharmacist and getting the specialized pharmacists that are working with the infectious disease team to really look at our antibiotic use of the lens. Whereas that might just be issue number three in seven issues for a trauma patient. In our eyes, I think really allows for much more clear communication and allows the stewardship team and the infectious disease team to highlight issues that we may overlook at times in the ICU. And I think that contributes to the more stable acceptance rates. We did find that patients that were in specialized critical care units were less likely, or sorry, are more likely to receive suggestions, but maybe less likely to accept them. The burn surgery population is a very, very unique one, I think, in ICU. And this is one of the very few burn centers that are present in North America. And these are very sick patients that require very frequent debridements. So I find that the surgeons and the intensivists that work there, it's a very high risk environment. And narrowing the scope of antibiotics may not be as well received right now. But I think that just means that we need to better understand stewardship and antibiotic use for these patients and study them a bit more to better understand how we can deliver appropriate antibiotic care in the long term, especially because they are at high risk for multidrug resistance. We did find that increased markers of critical illness meant the stewardship team was less likely to intervene. And I think this is, again, just cooperative. If you have a patient who's on higher doses of vasopressors or now intubated, yeah, you'll probably be less likely to go ahead and suggest that they discontinue their Fiptazol and Banco that they were just switched to. Similarly, I think if the diagnosis is unclear, everyone's a little bit hesitant to intervene. A classic example of this is something like ventriculitis, which usually doesn't get an organism, and everyone is not quite sure whether it's ventriculitis and it's a high risk situation. So you're probably less likely to go ahead and intervene on those cases at least. Similarly, if the diagnosis is unclear, classic example there being, for example, a ventilator-acquired pneumonia. Nobody really knows how to diagnose a ventilator-acquired pneumonia, unfortunately. And you're probably less likely to stop antibiotics in someone who's got a neopasity, whether it be from atelectasis, a pneumonia, or edema, and their oxygenation is much worse. You'll probably be less likely to go ahead and narrow their antibiotics. There was evidence that there was a lot of reluctance to narrow or change the therapy. I think this is multifactorial. There's obviously concern about safety of oral antibiotics and critical illness, which has been shown, it's been shown to be safe in a ward situation for pneumonia and UTI to narrow antibiotics relatively quickly, hasn't really been shown in intensive care situations just yet. And a recent, a very recent study that just came out by Brad Langford in the Clinical Microbiology Disease Journal did look at an eight-year period of stewardship assessments on the ward, and again found very lower rate, lower odds of acceptance for recommendations to decrease antibiotic exposure. So everybody was okay to start antibiotics, but there's a bit of a hesitancy to go ahead and decrease exposure ultimately. I think some future studies can look at longer-term outcomes in different stewardship models. I think it would be interesting to see, are some models more successful than others? I'm not sure how you define success necessarily, but it would be interesting to see a bit more of a variation. And how would this work in areas with different rates of antimicrobial resistance? USA probably has a bit more antimicrobial resistance compared to Canada, and so their outcomes might be, and predictors might be a little bit different. India has very high rates of resistance, so I'm sure they're potentially going to have different outcomes and predictors there as well compared to Canada. And this is our stewardship team I wanted to highlight. So Marion is one of our primary stewardship pharmacists, and she has been managing the SPIRIT database for a number of years now. She's been at Sunnybrook for almost a decade. Jennifer Lowe, another stewardship pharmacist who also works tirelessly in the ICU and on the ward. Nick Dannemann, who's the PI of our study, and he's the division head of the infectious disease team. And Phil Lam, who's now the director of the Antimicrobial Stewardship Program. And I wanted to give a big thanks to everyone here. Okay, so now I'll hand it over to my colleague across the pond, Diana. Oh, thank you, Pratik. So now I will proceed. Good afternoon, everybody, or good evening for if anyone is listening from Europe. I am very glad that I get the opportunity to discuss the study with you today. I performed the study during my time as a PhD student on the ICU of the University Medical Centre of Utrecht, and it's an observational study reporting the diagnostic yield of routine blood cultures upon ICU admission. And first, I would like to start with two polling questions. Is there an antibiotic stewardship team member present at your centre daily? And then A could be yes, B is no, not daily, but present at least weekly, C is no, but we do have a daily alternative, for example, consultation of an infectiologist, or D, no, we do not have an antibiotic stewardship team. Okay, so let's see if the listeners are answering. Oh, I did not see the answers, yes. Okay, so there are still, there's actually no one who does not have anything like an antibiotic stewardship team. Okay, so we can proceed to the next polling question. And that's a question that relates to the study I'm discussing particularly. The question is, can a blood culture be taken without a clinically suspected infection have any diagnostic relevance? The answer could be yes or no. Okay, so still 60% thinks a blood culture taken without a clinically suspected infection could have a diagnostic relevance. Okay, so I will proceed. When we talk about blood cultures in septic patients, we know that only 40 to 70% of septic patients will have a positive blood culture and several variables have shown to be predictive for blood culture positivity. A very important one, namely the presence of a suspected sepsis by the clinician, is all very known to us, but we also know other clinical variables, so like fever, chills, like a cytosis, or presence of an endocarditis. And some decision rules have been able to select a high-risk population, but these decision rules were never specific or sensitive enough to rule out a bloodstream infection beforehand. So this means it is still very difficult to identify all patients that will need a blood culture. On the other hand, you don't want to culture everybody. So blood cultures play an important role in the management of sepsis. Why? Well, they help us narrowing down the differential diagnosis, they assist in risk stratification of critically ill patients, as blood culture positivity is strongly associated with a higher mortality, and finally, they help to optimize and to guide antibiotic treatment. This is very important as septic patients are treated initially with broad-spectrum antibiotics and narrowing down antibiotic treatment is desirable to decrease antibiotic exposure and potential antimicrobial resistance. So the surviving sepsis campaign bundle is a set of important treatment and monitoring steps that have to be initiated immediately in case sepsis is suspected. So they include treatment with broad-spectrum antibiotics, early recitation, but also the obtainment of a blood culture, and forms an essential part of this bundle, and full and early completion of this surviving sepsis campaign bundle is associated with better outcomes and improved survival, and actually, delayed completion of the sepsis bundle, to be specific, after three hours, is associated with higher mortality. However, timely initiation of the sepsis bundle is hampered by several factors, which I will now mention. First, early recognition of sepsis is complicated as not all symptoms are as specific in the early stages of sepsis, especially not in patients that are immunocompromised or older. So up to 10% of the patients with sepsis admitted through emergency wards remain initially undetected, and we all know that a sepsis suspicion can arise over time. Second, when the patient is transferred from a ward or emergency department to the ICU, important patient care information can get lost during transfer and after office hours. And third, available patient charge, even when they are electronic, may not all support real-life registration of vital signs, administration of antibiotics, or registration of blood cultures that are taken. So this is another reason why information may not be up-to-date or accurate by the time a patient arrives to the ICU. Of course, there are targeted quality improvement programs that are introduced to solve these problems. However, we know that up-to-date, these factors I just mentioned, still delay the recognition of sepsis. So in our center, we noticed that there seemed to be a suboptimal availability of blood cultures, and we therefore decided to change our blood culture practices and to observe what would happen. So we implemented a new strategy in order to increase blood culture rate, and our aim was to improve this rate in patients with a suspected infection during admission. However, the practice was implemented in all patients. This was a pragmatic decision because clinical suspicion of infection could arise over time and was not always present in the first hours of admission. So we made use of the MARS study and its database. The MARS study prospectively collects comprehensive clinical data, daily blood samples, and detailed description of infectious episodes. And this data was collected in all patients with an expected length of stay over 24 hours. Study occurred in the 32-bed mixed ICU of the University Medical Center of Utrecht, and this is a tertiary care hospital in the Netherlands. And clinical microbiologists attended daily multidisciplinary meetings in the ICU. So our change in practice meant that we initiated an automatic order requesting a single blood culture in every acutely admitted patient. So our study was a before-after analysis, and we collected data in a two-years-before routine blood cultures per common practice, and this period functioned as a control period. We also collected data in the two years after ordering routine blood cultures, and we looked at the following outcomes. The presence of bloodstream infections, contaminants, results were stratified by clinically suspected infection upon admission, and we also report potential adverse effects and expected effects, such as repeat blood cultures, vancomycin use, and treated infections in both periods. So all blood cultures consisted of one aerobic and one anaerobic vial, and were processed following routine clinical protocols in our local laboratory. And bottles were incubated in our blood culture system with automatic signaling of microbial growth, after which Malditov and Gram-Sanning were used for pathogen identification. Positive culture results were immediately communicated with the clinicians, and for the study, positive results were adjudicated afterwards as being a contaminant or pathogen. And for this adjudication, we used commonly used characteristics that you might know, such as type of pathogen, number of positive cultures, and time to positivity to distinguish between skin contaminants and true pathogens. So during our study, there were about 1,800 patients included in both periods. Baseline characteristics distribution, such as age, sex, disease severity, and comorbidities seemed very similar. Also 30-day mortality seemed very similar. However, there were still some differences, such as a higher number of patients that had a prior ICU admission, and a higher number of patients that was transferred. So in the before period, about 998 blood cultures were taken, against 2,646 in the after period. The percentage of growth seemed to be largely similar between the two periods, and as we will see in the next slide, it was largely driven by a high number of contaminated blood cultures in the after period. The number of cultures expressed in median cultures per patient was 0.5, and 1.4 cultures per individual in the before and the after period. The percentage of patients in whom a blood culture was taken increased, of course. It was 32% at first, and 85% in the after period. We identified 5.4 patients with a bloodstream infection, compared to 8.2% in the after period. And the percentage of contaminated cultures increased to, and was actually four times as high, increasing from 2.3% to 9.6% in the after period. We actually estimated that around 1,009 additional patients were cultured in the after period, and this is 1,582 minus 573, and this yielded 59 additional bloodstream infections, which is the difference between 95 and 154. This would correspond to a number needed to culture of 17 to detect one additional patient with a bloodstream infection. So when we stratified a population by suspected infection, we saw that this increase in bloodstream infections largely occurred in a group of patients that had a suspected infection upon admission. So this definition was based on the prescription of therapeutic antimicrobial therapy in the first two days of ICU admission, and it was part of the prospective data collection of the MOSS study. The red boxes show the proportion of patients with a bloodstream infection, and we also see that about 40% of the included patients were suspected of infection. And in this group, 93 patients had a bloodstream infection versus 142 in the after period, and this was 2 and 12 for the non-infectious population. So this is the proportion of bloodstream infection and contaminations over time. So bloodstream infections were more common in the after period, but we also saw that contaminated blood cultures were more common than real bloodstream infections, especially in the first half of the after period. We found a crude relative risk of 1.5 for detected bloodstream infections in the after period. We corrected for potential confounders such as age, gender, comorbidity, and disease severity, and this yielded a similar adjusted relative risk of 1.6. When we look at the potential adverse effects, we see that indeed blood cultures are repeated more often, also when we only consider patients that had a previous contaminated blood culture. Pancomycin, however, was not given more often in the after period, and the percentage of patients receiving pancomycin for both prophylactic and therapeutic reasons was only 1% or even lower. As for the patients that were suspected at any point of a catheter-related bloodstream infection and other treated infections, the number remained similar throughout the two periods. So if we consider our results, we have to take into account our hampered study design, and a very important point of critique is the suboptimal adherence to the protocol. So we did show a 1.6 increased relative risk of finding a bloodstream infection, which results in an improved documentation and detection of bloodstream infection. However, we can question whether this intervention would indeed guide treatment decisions and would substantially aid in the decision of narrowing down antibiotics, as we did not look at this variable. Furthermore, I think it's important to consider the setting of this report. It was performed in a center with an involved antimicrobial stewardship team, and this may have prevented an evident overuse of antibiotic therapy in the after periods. So I would like to conclude with the following main messages. Many studies have focused on the optimal target population for blood cultures, and this has resulted to be very difficult. Furthermore, complex logistics may delay sepsis recognition and decrease the blood culture rate. And if you culture all patients acutely admitted to the ICU, we found an increased number of detected bloodstream rates, bloodstream infection rates, and calculated the number needed to culture of 17. This leads us to the conclusion that routine blood cultures could offer good value for health and improve detection of bloodstream infections at a, for us, acceptable increased rate of blood cultures taken. This benefit, however, should be weighed against the increase in the number of contaminated blood cultures. So it remains important to focus on early recognition of sepsis and protocolized collection of cultures to improve the management of sepsis. I would like to acknowledge all the co-authors that made this project possible. Also I like to thank the always actively involved MARS investigators and all the other research members of the epidemiology of sepsis group. And now it's time for the Q&A that Tony will moderate. First of all, before I start, feel free to ask any questions in the question box and I'll get to them as time allows. And I would like to thank Prateek and Diana for excellent presentations on our antibiotic stewardship webcast today. And I have a quick and easy question for Diana. It was striking to me as a critical care pharmacist that you had such low rates of vancomycin use in your hospital. And I'm guessing that you guys have a low rate of MRSA. Here in the States, it's about 50% at my institution. Do you know what the rate of MRSA is of all Staph aureus species at your hospital? I am not aware of the current rate, but it must be very low, maybe 2% even lower. I was guessing that being in Europe, you're going to see some differences there because definitely that's what's in the literature and it's just striking to me. But I do have a question, especially early on, it looked like contaminations were a big part when they implemented these routine blood cultures. Do you know what are some of the lessons learned if you did a project like this and what we can learn from that? Yes, so an important one is we noticed, of course, you can never, as it was not a controlled study, you can never relate it completely to the education we did, but there was a moment of education and after this, the contamination rates decreased. So yes, I would at first start with a full education program on contamination and the education should be given to nurses that are taking the blood cultures, yes. Very good. It sounds like you need to get all your ducks in a row and make sure that you have the education on the appropriate way to do this if you're going to make any changes. Now I'm going to go on a little bit to critique and what was interesting, at least to me, is that the higher acute ICUs, the level three ICUs, had a higher acceptance rate than the level two. Do you have any idea why that might be? Yes, so I thought that was interesting as well. My, of course, this is all hypothesis generating, but my theory would be that to tell that the most common suggestions were to, were on the more narrow end of the spectrum from the stewardship team, whether it was narrow, discontinue antibiotics or change the type of formulation. I actually wonder if a lot of that odds of acceptance or lower odds of acceptance at level two ICUs came from the stewardship team being more likely to suggest reducing antibiotic exposure once you're in a level two ICU setting. So I think a common example that we see is a recommendation to, you know, as the vasopressors start coming down to switch from IV antibiotics to oral antibiotics, there's always a bit of hesitancy on there. So I wonder if that caused lower acceptance rates in the level two ICUs, which kind of made the level three ICUs look like they had relatively higher acceptance rates, because you're not going to suggest narrowing antibiotics as much in a level three ICU, I think. That was my theory, at least, behind that. Yeah, very good. I think it also probably has to do with familiarity with people. And I see this, you know, in my practice where I've worked really in the same surgical ICU for over 20 years. And as you get more familiar with people and see the same people, I think it also helps out a lot, as well as you having the appropriate rapid diagnostics and education for it. Because it's very well documented in the literature. You can add rapid diagnostics, but with people don't know how to interpret the results, it makes a big difference. So I think another question that I had coming up is, I'm not sure everyone's familiar with this concept called handshake stewardship. And it's the idea of bringing the ID and ICU teams together, where they actually round concurrently, which is very time consuming on the ID hand, because rounds often are longer. But I think anything that we do as a follow up question to promote that collaboration and learn from each other is actually going to help. And maybe that's what we were seeing in the level three and the level two, as well. And what are some further ways do you guys think we can promote both infectious disease and ICU collaboration? Yeah, I think one way that I think we can improve that collaboration is, so I like the face-to-face model. I think that's the most important. Being able to speak directly to the pharmacist or the infectious disease consultant. I think that's critical to, you know, forget the acceptance rates being up or down. I think that's important for collegiality and just getting everybody on the same page and get buy-in from the ICU, because I think you need to get the buy-in from the ICU team that, you know, I can trust these pharmacists, they understand our critically ill patient population. And I think to build that understanding, you need to speak directly to the teams. So I think that's one of the important things. What I like to know whenever, because in Toronto, there's a number of hospitals that we rotate to as fellows. I think one important piece of information that should be readily available to the critical care team would be antibiograms and being aware of what the local sensitivities are. Our sensitivities are different from one hospital on one side of the street to the other side of the street. And when you're changing rotations a lot, you don't necessarily get that same understanding. So I think having openly available antibiograms is important, because then you might realize that, you know, you don't need to add the example you gave us, intravenous vancomycin empirically, because the MRSA community rates are so low. So maybe you don't need to do Cortezo-Vanco when somebody comes in necessarily. Yeah, I think that's always something I struggle with. Would a face-to-face model be the same as a daily meeting, or how would this face-to-face model work? It's my understanding with the hands-to-steak stewardship, and it's mostly been described in pediatric ICU, so the hospitals are a little bit smaller, is that they actually round together both teams simultaneously on the patients. So I'm not sure they follow all the patients, but the ID physicians and pharmacists are actually on rounds with the ICU team to promote that collaboration. But, you know, I think the other thing, which you bring up a good point, is, you know, is there a way at least once a week that you guys can, everyone can get together? And I agree 100%, you know, in my hospital, between the medical ICU and the surgical ICU, there are a little bit differences, and especially the gram-negative antibiograms, I mean, in the States, our rates of MRSA are way higher than what you see in Canada, and obviously what's in the Netherlands out there. But I also think, you know, you need a lot of education, and I think that's one of the biggest pieces that, why we need to have face-to-face promotion, and if you really look at one of the tenets of antibiotic stewardship, it's really education of what's going on. And I think the more that you can educate team, whether it be me as the ICU pharmacist, the ICU teams, as well as the ICU teams, the ID doctors, especially, you know, surgery, doing more procedures that are a little bit different, it's just going to increase everyone's knowledge and help with some of these issues. Plus, I think it's nice, especially with electronic medical records, to really see what their history is and what their history of pathogens are. Although you might have a low rate of ESBL-producing strains, you have a patient that just had an ESBL UTI a month and a half ago, and now coming back in with a kidney stone, if you don't put them on a carbapenem, that's probably a bad thing. I almost wonder if there should be, like, we get a notification for someone's MRSA-positive that shows up as a flag on our system, I'm not sure if there should be one for something like that as well, if somebody has a multidrug-resistant strain, because then that can guide to come in septic, you know, ten months from now. And especially, you know, especially as you rotate and you might see them, you know, and it might be in, you know, hospital A, but you're at hospital B and recognize the patient. I completely agree with that, so the more the information would be better. Yes. Does the audience have any other questions? Well, once again, thank you very much to both Prateek and Diana, you guys gave amazing presentations and I learned a lot today. And again, everyone who joined us today, there will be a follow-up email that will include an evaluation. Please take about five minutes or so to complete the evaluation, because we use your feedback and it's always greatly appreciated. And on a final note, join us for our next journal webcast on March 25th, and this concludes our presentation today, and thank you yet again, everyone. Have a great day. Bye. Thank you.

antimicrobial stewardship

ICU models

C. diff infection

barriers to implementation

stewardship utilization

routine blood cultures

detection of bloodstream infections

contaminated blood cultures

sepsis management