SCCM Resource Library-Sepsis: The Source Matters!

Video Transcription

Hello, and thank you for the opportunity to speak to you today. My name is Dr. Ryan Mavis. I'm an infectious disease and critical care physician at the Naval Medical Center in San  Diego, California. I have the pleasure of moderating today's session entitled, Sepsis, The Source Matters.  As we all appreciate, sepsis is not a single disease, but rather a multitude of different disorders with certain common features, such as organ failure, but for which a homogeneous  approach to care may be unsuccessful. Although hemodynamic and respiratory support, empiric antimicrobial therapy, and the like  are the key to the care of the septic patient, individualized approaches are needed to improve the outcomes for many of the patients under our care.  We are fortunate to have two expert panelists today to discuss common pitfalls and approaches to improve care. First, we'll be speaking with Dr. Eric Mueller.  Dr. Mueller is the Assistant Director of Clinical Services and Research and an Adjunct Associate Professor at the University of Cincinnati Medical Center.  He'll be speaking to us on common pitfalls in antimicrobial therapy selection in the septic patient. He'll be followed by Dr. Matthew Tadaloch.  Captain Tadaloch is a trauma and critical care surgeon at the Naval Medical Center in San Diego, California.  He is also an Assistant Professor of Surgery at the Uniformed Services University and is the Trauma Advisor to the Navy Surgeon General.  He'll be speaking on challenges related to source control in the care of the septic patient entitled Drugs Won't Save You, the Impact of Source Control, including a discussion  of the recognition of early source control in sepsis and septic shock and understanding the need for early expert consultation.  Thank you very much in advance to Dr. Mueller and Dr. Tadaloch for your talks, and I hope you enjoy our session. Thank you very much. Thank you for being here today.  My name is Eric Mueller. I'm the Assistant Director of Clinical Services and Research at the University of Cincinnati  Medical Center and Adjunct Associate Professor of Pharmacy Practice and Administration at the University of Cincinnati College of Pharmacy.  Today under the broad topic of sepsis for source matters, I'm going to be presenting on common pitfalls in antimicrobial therapy selection.  I have no financial disclosures to make today. The learning objectives for my part of the presentation today is first to review differences  in antimicrobial spectra, tissue penetration, and resistance that may influence therapy selection in a patient with sepsis or septic shock.  I'll also very briefly introduce the objective to identify pathogen-specific factors, including  virulence factors, and immunologic responses that impact the course of disease in the patient with sepsis.  As we know, antimicrobial therapy is a cornerstone for managing patients with sepsis and septic shock.  The 2016 Surviving Sepsis Campaign Guidelines recommend, if able, giving intravenous agents administered within one hour of sepsis or septic shock onset with an empirically broad  spectrum, often requiring one or more agents, focused on hopefully risk-factor-driven or patient-specific-based disease elements that may help us identify likely pathogens.  The regimen should be optimally dosed, being mindful of the suspected or known site of infection, being mindful of principles of pharmacokinetic and pharmacodynamic dosing  that ideally are sepsis-specific, derived from research in sepsis or septic shock patients in particular.  From there, therapy should be de-escalated, focused on the causative pathogens that are  found or discovered, minding clinical improvement as well in our decisions to de-escalate patient's therapy.  From there, ultimately, we want to provide evidence-based durations of definitive therapy to manage the patient's known or suspected infection to help treat fully the courts,  as well as to be mindful of the long view of avoiding unnecessary prolonged use of antimicrobial therapy in these patients.  The first four elements of these recommendations really depict the cornerstone principle of providing timely, appropriate, empiric antimicrobial therapy.  So we're going to first dig down into this definition of timely, appropriate, empiric antimicrobial therapy.  We know that delays in appropriate antimicrobial therapy may often be perilous for our patients. As compelled by Kumar and colleagues and many others since, we know that every hour  delay in appropriate empiric antimicrobial therapy decreases the likelihood of survival in our patients. I'm not going to focus too much today on the pitfalls around delays.  We know that using standardized processes of care and workflow in an interprofessional  collaborative way, and ultimately with performance improvement principles in play, centers have  been able to demonstrate improvements in timely application of appropriate antimicrobial therapy.  Again, focused on closing the time gap from the onset of sepsis, the onset of septic shock, to the administration of antimicrobial therapy.  What I want to focus on today is mainly the pitfalls around inappropriate therapy. Again, antimicrobial spectrum, tissue penetration, and or antimicrobial resistance as really  being some of the major disruptors of our best laid plans of providing timely, appropriate antimicrobial therapy.  So in focusing in on inappropriate therapy, I think there's two characterizations. The first is a spectrum of our regimen mismatch to the causative pathogens.  The second characterization is really emerging in an exciting field in my opinion as a clinical  pharmacist, and that's inadequate source presence of our antimicrobial agents and regimens. So if we focus on the first characterization, we'll review today, you know, really where  the empiric spectrum may have missed. So I chose a spectrum of agents, antimicrobial agents, and the causative pathogen fell outside that spectrum.  I missed the organism, if you will. Another characterization sub within that is I hit the pathogen, but the pathogen is eliciting resistance to the regimen that I chose.  Normally the regimen would cover it, but this pathogen has resistance. So we'll talk a little bit about the implications therein.  Some have been also defining inappropriate therapy maybe with an immunosuppressed population  where perhaps a bacteriostatic or a fungistatic regimen was chosen when maybe a bacteria or fungal static sinusoidal regimen may have been more preferred.  The second characterization of inappropriate therapy, going back to source presence, I'll  try to focus on site penetration or availability of a specific antimicrobial regimen to different sites of infection.  We'll talk about some pharmacokinetic and pharmacodynamic targets and or implications of disease with sepsis, and then ultimately how that relates to our dosing strategies.  These are all lumped together as pitfalls in our attempts to provide appropriate and timely antimicrobial therapy in our patients with sepsis and septic shock.  So we focus first on just hitting the bullseye of the causative pathogen, really not even resistance being major here, but just in the broadest of sense, did I hit the target of  the causative pathogen? And so a number of groups have published local rates, multicenter rates, and most recently  systematic reviews or large big data sets of sepsis and septic shock patients, particularly  around blood cultures and whether or not and how often we've hit the spectrum-based bullseye. And as you can see, there's a wide range here. On the x-axis is the respective study.  On the y-axis here is the rate of inappropriate empiric therapy. What we can see is most of these studies suggest that we are not making an A grade, rarely  a B grade, and most often we're C, even D grades in our rates of providing appropriate therapy.  So we focus on the largest three, again, systematic reviews as well as national data sets. We have rates of inappropriate empiric therapy somewhere between 20% and 30%, and that's  pretty staggering that we're missing just the simple spectrum bullseye between 20% and 30% of the time.  That's not even minding more deeply the considerations of resistance and or the considerations of source presence in our antimicrobial dosing strategy.  So as we dig a little bit further, let's talk about rates of resistance and the implications of resistance on providing inappropriate therapy, and these are just three studies.  One is a systematic review, about 73,000 mixed infections. The other two are single center evaluations, but all three of them help earmark the influence  of the presence of multidrug-resistant pathogens. So pathogens, yes, that were known to be present and likely in the prediction, but  unfortunately had antimicrobial resistance variance within that specific patient. In addition, we see that antimicrobial use, even if a patient did not have a cold-resistant  pathogen, antimicrobial use, particularly broad spectrum, have been implicated as increasing the odds of having inappropriate empiric therapy present in a patient's subsequent  sepsis bout or subsequent septic shock bout. So what we see here is really a paradox that we all practice within, and that's we have  this emerging or solidified concern with inappropriate empiric antimicrobial therapy that then leads us to engage in a broader spectrum regimen.  These broader spectrum regimens, as history will tell us clearly in our experience as  well, will then increase the risk of antimicrobial resistance or even selection of other resistant or unpredicted pathogens.  That then fuels our need to, and our discovery of increased rates of inappropriate antimicrobial therapy.  That then fuels the increased interest, therefore the increased spectrum, more exposure, so this vicious circle between the presence of resistance, increasing spectrum, increasing  resistance to increasing spectrum, is really where we sometimes find ourselves. Maybe not so much in the day-to-day because the short view isn't as apparent, but as we  look over years of time, whether it's locally or nationally or internationally, we start to see the influence, and these studies help quantify the impact of that usage on the rates  of inappropriate empiric therapy later. It's really important to understand the bullseye missing by pathogen, here implications of missing because of resistance.  We have to think about, are we missing because we have outlier or unexpected pathogens? For example, we have healthcare-associated pathogens in community-based infections, or  we have non-bacterial pathogens, or we've underappreciated a source, perhaps maybe a GI source with anaerobic pathogens, even though we have positive blood cultures otherwise,  or when we talk about and the guidelines emphasize understanding our local ecology and local antibiogram, are we really being led in a specific and sensitive way?  For some patients, maybe, but for many others, perhaps the sensitivity and specificity of our local learning doesn't hit high enough to really get that patient-specific pathogen.  And then, as I've mentioned, multidrug-resistant variants of the predicted pathogens that we are expecting to see in the context of our patient's diseases.  Thankfully, we have guidelines, usually organ-specific guidelines or national guidelines that help  earmark some risk factors that we might see and consider when determining which agents to use.  So, for example, considerations for antifungal therapy from the IDSA-Kinda-Diasis Guidelines. MBR or risk factors for ventilator-associated pneumonia and hospital-acquired pneumonia  in the bottom box. And then even so far, to get into specific resistant variants, such as Janelle and colleagues  that help depict a scoring system to increase the predictability of carbapenem-resistant  Klebsiella pneumoniae, we can see here colonization plus a number of features that our patients in the ICU commonly present with.  So together, understanding risk factors may help us curtail our spectrum. Do we need to be all broad? And as broad as we think we do, or can we be narrow?  All in the efforts to address the near-term severity of illness, but be mindful of the long-term implications of antimicrobial use.  I think one of the first things here to understand with hitting the bull's-eyes from a spectrum perspective is our local ecology.  I'm borrowing this large data set from me and colleagues that included 95,000-plus Shepsis cases and looking at the prevalence of organisms across positive blood cultures.  This is a seven-year series, including 136 U.S. hospitals, and as we can see on the far left, about 15 pathogens were identified. You can see a decrease in prevalence.  Those pathogens go from E. coli to Staph aureus, all the way down to anaerobic bacteroides, including Candida species kind of in the middle there.  They broke down the pathogens then into community and hospital-related pathogens, and so this is pretty much for the ages we know that pathogens in the community are different often  from pathogens in the hospital. We're starting here in Cincinnati to see blending of that, so we are seeing de novo resistant  pathogens beyond MRSA, particularly in enterogram-negative rods that are in community-based infections, but obviously having no succomial flavor in those pathogens.  So if we look what happens as we move from community to hospital, almost characteristically,  we see a decrease in Streptococcus and or E. coli, if you will, but then we see an increase in a myriad number of other pathogens. MRSA prevalence goes up.  We start to see emergence of enterogram-negative rods that are resistant, Enterobacter or Citrobacter.  We also see Pseudomonas and Acinetobacter increasing in their prevalence, as well as Candida species evolving to join as well anaerobic pathogens, particularly for intra-abdominal  infections. So I think this is really the first, broadest tier of responsibility we have in understanding  how we design our empiric regimens is the very high level upstream, what are the causative pathogens locally, whether it's institution level, ICU level, or even disease state level  within mysepsis and septic shock patients that one may see at their institution. Talking about antimicrobial spectra, this is a pretty simple bugs and drugs.  I focused here on antibiotics. The first thing we see up top are our common empiric regimens, you know, Cefepine, Peptazole,  maybe Carbapenems, being our beta-lactam workhorses, perhaps Ceftriaxone for community-associated  infections, Aminoglycoside, Ciprofloxacin, and then Vancomycin and Azolib, at least here, and I think from a lot of national data sets in the United States, still being the two  most common empiric MRSA agents. I, you know, even if you broaden this to Ceftaroline with anti-MRSA activity but no  pseudomonal activity, or even Cephidericol, the newest kid on the block, having very good gram-negative coverage but nearly void of gram-positive coverage, excuse me, it's still,  we don't have a regimen or an agent or class that has coverage for all of these suspected  organisms on the far left, and so we are left with clearly multi-drug regimens, gram-negative,  gram-positive, and even within our gram-negative spectrum, maybe dual agents, you know, a beta-lactam  plus a fluoroquinolone or aminoglycoside that we're seeing very, very commonly these days in recommended empiric regimens across myriad disease states.  So if we look at our national antibiogram, if you will, you know, I'm using Sentry data over the previous five years, I ran a report from this publicly available database looking  at all U.S. adult ICUs that report to the Sentry Surveillance Program, it was just over 2,600 isolates, and what we see is on the bottom left here, looking at a number  of our gram-negative pathogens and the percent susceptible to respective agents on the far right, and I also have them noted in this 3D graph by the bars.  So I think what we notice first is even among E. coli, which might be the most garden variety of enteric gram-negative rods, we start to see, you know, 70 and 80 percent activity  among cefepime and piptezo. Carbapenem still have great activity against E. coli, but we start to see that degrade when we get into Klebsiella enterobacter.  Surely among Pseudomonas, where these three agents, cefepime, piptezo, and meropenem are basically a tie. No advantage among Pseudomonas between these three agents.  Ashmetobacter, we see a gap in nearly every opportunity, particularly among these first  three agents. You know, our aminoglycosides tend to have the best activity across these pathogens that  are depicted, but again, with an asthenobacter, we still have a major gap of activity. The newer agents are beta-lactam, beta-lactamase inhibitors, ceftolazine, ceftaz, with their  respective beta-lactamase inhibitors, we do see above 90 percent, nearly 100 percent activity  across the enteric and native rods, Pseudomonas, thankfully, but again, asthenobacter, we really have a gap of activity.  There are emerging reports of using ceftaz or beta-lactam in asthenobacter, particularly carbapenemase producers.  We know cefvericol has activity, but really just not included in the Sentry Database, but an emerging role, mainly in definitive therapy for those agents rather than empiric  so that we can maybe sustain their utility over the next number of years. If we look at aggregated activity from this gram-negative spectrum, maybe across this  U.S. dataset, we see that some of our more traditional agents, cefepime, peptazo, are about a 70 percent at best agent. Maybe meropenem is getting close to 80 percent.  We know our aminoglycosides are going to stay in the 90 percent, so together maybe in the 90 percent as combined regimens.  Ciprofloxacin across the nation dataset in a problem that we've had here, really has limited utility as an empiric regimen, it appears, and again, even though our newer  beta-lactam, beta-lactamase inhibitor combinations get us around 90 to 100 percent activity, we should be cautious, I believe, with what history has demonstrated to us in that if  we have unbridled enthusiastic empiric use from an all-comer perspective, the utility of these agents is probably short-lived to come.  Now going beyond spectrum, thinking now about the emergence of resistant pathogens, we really  see, at least here in the United States, the Centers of Disease Control helping to remind us of these emerging threats, and the urgent threats are really upon us.  Coveropenem-resistant antibacter has been a culprit for decades now. Coveropenem-resistant antibacteriaceae, perhaps even more concerning because of the prevalence  of E. coli, Klebsiella, Enterobacter, is higher than Ashenetobacter in most centers, but now we're seeing these resistant variants living among normal human hosts and patients  that present with sepsis and septic shock. We see emergence of different fungal pathogens, different candida species, drug-resistant candida, as well as Aspergillus.  And then what has always been serious has been drug-resistant Pseudomonas. We know Pseudomonas has an attribute of mortality in sepsis that outweighs other pathogens,  rivals fungal etiologies as well, and so mild drug-resistant Pseudomonas really continues  to emerge and continues to be different nearly every year or so in its resistance mechanisms that we're seeing.  So very quickly, just to review, and again, I'm focusing here on antibiotic resistance because it's most well-characterized.  Bacteria are matching the breadth of their spectrum in how they come up with mechanisms of resistance.  They're matching the breadth of our antimicrobial and antibiotic spectra, and so it's really... We do, they respond.  We then respond, they respond sort of biology, which is challenging for sure, impressive nonetheless, but intricate.  Each time we see a new pathogen emerge in the multidimensional and the multi-mechanistic approach to resistance that they employ.  So the major categories of resistance we see are antimicrobial destruction, beta-lactamases  and carbapenemases, altered binding sites where the pharmacologic target has been altered. An extension of this is impaired access to the target.  So surely the antimicrobial agent can't even bind to the target if it can't access it. What's really impressive is these pathogens are bypassing their normal biochemistry and  creating alternate pathways, so that even if the antimicrobial does sit on its target site, it doesn't matter.  The bacteria, the fungus has determined ways to no longer need that biochemical mechanism, so it just effectively bypasses that pharmacologic interaction.  And ultimately, not letting the agent in the shell membrane or through the shell membrane at all, actively kicking it out, et cetera.  So really complicated aspects, genetic to phenotypic variants, but really important to understand. Thankfully, we have molecular diagnostic tools available to us now.  I imagine a myriad of the sites that you're at do have some of these tools in place that may help us in between the culture being obtained and sent to our micro lab to the  final culture, which may take three days, grandstanding between, to now molecular diagnostics  suggesting with at least important information of genetic variants, maybe even pathogen specificity  early on, that we can alter our empiric regimen and avoid some of the pitfalls in that delay  of timely appropriate therapy. Jumping back into the Sentry database, I ran a report of antimicrobial activity, antibiotic  activity for pathogens, E. coli, Klebsiella, and Enterobacter that elicit extended spectrum beta-lactamases, carbapenemases included, and we can see here these are daunting to  our empiric regimens. E. cefepime has nearly sub 40% to 30% activity, peptasia is also effective in a substantial way among Klebsiella and Enterobacter.  We see carbapenems nearly now 70% or less, aminoglycosides by way of amnication still have relevant activity, but again as we've mentioned there's a therapeutic index concerns  often with aminoglycosides as well as tissue penetration concerns, which we'll briefly review.  We do see ciprofloxacin really being a non-contributor here, and so beyond the spectrum of usual  pathogens, risk factor driven, beyond now antimicrobial resistance variants, our traditional  bacterial regimens, particularly on the ground negative side, really start to be daunting  in the breadth of their coverage and the ability of being able to provide appropriate empiric therapy.  What this takes is understanding of the local spectrum, understanding of even ICU level, even disease state level, and perhaps most importantly the state, the patient specific  assessment of what is the likely causative pathogens, the history of their resistant organisms in their medical care that will hopefully help us pinpoint the likely pathogens  causing their infection. So we get now to the second characterization of how antimicrobial therapy is defined as appropriate.  So we just discussed spectrum, resistance, maybe some strategies to understand and depict and to predict what the causative pathogens are and their resistance patterns.  The second characterization that I think is important to understanding appropriateness of our regimens is whether we used a regimen that penetrated the site of infection and  whether we dosed it among sound pharmacokinetic and pharmacodynamic principles. So just a quick review, we know that there is pharmacokinetics or the way the body, in  this case the patient, handles the drug that determines the plasma or for an infection the tissue or site drug concentration.  We also know that there's bacterial kinetics, if you will, at liberty to think about how the bacterium or the fungus handles the antibiotic or the antifungal agent.  Does it let it get into its site? So is there absorption, if you will? Does it get to the binding site for pharmacologic activity or is it destroyed by enzymatic breakdown  or is it eliminated by pore and channel efflux, et cetera? All of this then relates into does the patient have side effects of the agent or the host  pharmacodynamics or does the agent lead to antibacterial effect or antifungal effect,  whether it's decreased growth or stasis or whether it's killing or bacteria or fungicidal activity.  So this is a pretty simple characterization and depiction but I think it's important to  understand we have patient kinetics and dynamics but we also have bacterial or fungal or pathogen  kinetics, which is really our mechanisms of resistance as well as the dynamics within there and whether or not the agent has an antimicrobial effect.  So if we go further into this, how do we know if this is appropriate? Well I think we have to understand the alterations at the patient level in drug fungal kinetics,  particularly in patients with septic shock. And so what I'll focus on here is volume of distribution, so which compartment or where does the drug go?  Whether there's an effective drug concentration at the site of infection is key and crucial. And then clearance. Is the drug being cleared rapidly or slowly?  Is there a safety concern or are we devoid of drug in an effective manner over an effective period of time? So this is just a simple depiction. Many have done  studies to characterize these pharmacokinetic alterations. Some are good, longer drug present, but some are bad because that may also have more adverse effects. But I will say that  all of these are challenges mainly because we are not privy to the drug concentrations  at sites. We may not have serum drug concentrations for all of our agents where we practice. And  so all of these patient variants that we see in the lighter blue will have either increased  or decreased implications on these pharmacokinetic parameters, which we know will ultimately  drive then the concentration of antimicrobial on the site of infection and the interaction  they may have with the causative pathogen. So real quick, from a site perspective, understanding the common sites of sepsis, we see across some recent sepsis studies or pooled data  sets like we and colleagues as we readdress here, we see that the four most common sites of infection are pulmonary, abdominal, urinary, and blood sources of infection. So it seems  to me that it would be beneficial in this 90th percentile of infection sources that we target these and think about our regimens in the context of, does my patient have a  likely pulmonary source? Will the regimen penetrate therein? Do they have an abdominal source? What's the likely presence there? Genitive urinary, and then blood.  So what we see here is a qualified depiction, again, of source presence, particularly with antibiotics. If we look head to toe, what I've done here is depict kind of dark green  are agents that have very good presence. Light green is good enough to maybe use clinically.  Orange is guarded, perhaps maybe not a primary therapy. And then red, likely not any utility in the respective sources. So CNS, we see really only light green agents. Pulmonary,  we have a mix of dark green and light green agents. Thankfully, most of our workhorse regimens for ventilator-associated pneumonia and general sepsis do overlap for pulmonary  sources. Most of our broad spectrum agents will address genitive urinary sources of infection. Bone is often challenging. We do have some light green agents, but really no  dark green agents. From a GI perspective, including pancreatic and biliary presence, we see there are carbapenems and fluoroquinolones along with metronidazole are workhorses and  appropriate types of cycline could be considered. We know that there's concerns with overall  patient outcome perhaps in a black box warning here in the U.S., but nonetheless kinetically has very good intra-abdominal penetration. We do have some light green agents. We don't  know that many glycosides are listed in multiple places, but as I said, as single agents they're  guarded with a narrow therapeutic index as well as maybe again here guarded tissue penetration.  I did want to draw your attention, this dot here is supposed to be on the blood vessel depiction noting blood presence. Most of the agents that are green in these deep-seated  spaces are also green in blood. I will say though aminoglycosides, maybe lenazulate empirically  and then tied to cycline as being red would really cause precaution in use for patients  with bloodstream infections. So just a qualitative depiction of maybe general sources of penetration.  Here more quantified, I'm not going to go through this given the interest of time, but nonetheless you can see the agent on the left and then the pulmonary epithelial lining fluid,  some estimates from studies whether they're averages or ranges, and just to then depict CSF which is maybe thought of as the deepest seat for our infections that do happen in  sepsis and septic shock patients on a less common occasion, but nonetheless often thought of as the  most challenging site to penetrate. So we can see in comparison pulmonary presence versus CNS with CSF presence. And then finally pharmacodynamics. I do think as a clinical  pharmacist and I think a lot of evidence does support the idea that the way we dose our antimicrobials, not just to be mindful of the site penetration, but also be mindful  of the interaction that the antibiotic concentration is having with the pathogen at the site are  relevant. And so what I see here, and this is a pretty common depiction, we often see  the interactions of the drug concentration with the pathogen minimum inhibitory concentration.  We see is it a drug peak to MIC relationship, is it the time above the MIC, particularly with our  cephalosporin, beta-lactam, and carbapenem agents, is it area under the curve of these drug  concentration curves relative to the MIC like we'll see with vancomycin, nasal lid. And so I  think these are crucial elements and they're probably still emerging, but there is some good  evidence now that suggests in individual studies as well as systematic reviews that this may have  implications on our patient's outcome. Last red dotted line here is that there is also  consideration for a concentration of antimicrobial that may prevent resistance. It's a bit beyond  this discussion, but still something to be mindful of, and to be frank, quite concerning and scary if this is another target we need to be mindful of in the future.  So quick reminder, we have time-dependent antimicrobial agents and we have concentration-dependent. In the time-dependent feature, is it time above MIC like our beta-lactams  and carbapenems, or is it AUC over MIC like vancomycin and nasal lid in our azole antifungals?  For concentration-dependent agents, is it a peak to MIC relationship like our aminoglycosides, amphotericin B and echinocandins, or is it an area under the curve to MIC relationship  in a concentration-dependent manner like our fluoroquinolones? So a really brief review.  Why I think this is important is joined by some emerging evidence. In 2014, excuse me,  Roberts and colleagues conducted a cross-sectional point prevalence study taking two real measured  serum concentrations across myriad number of beta-lactams across 68 ICUs in an international  study of 361 patients. What they were able to demonstrate is looking at the blue bars here across some commonly used carbapenems, penicillins, pipericillin, and cephalosporins,  we see that still just about 80%, some 100% of patients will have minimal target attainment  of time above MIC. However, optimal will not. And again, these are blood samples. These are not even site-specific samples, which are very difficult to achieve and attain.  So what they were also able to demonstrate is there's a decreased odds of clinical positive  response if a patient did not achieve the minimal pharmacodynamic target. So there's a clinical  implication as well, on top of the implication that we're already missing the target with our  empiric regimens 20 to 30% of the time. Now our dosing may also be missing even an arguable  optimal target 30 to 20% of the time. It really calls for concern of this interaction between the target being a definition of appropriate therapy, but also is the dose appropriate?  And then finally, Bardakis and colleagues have conducted a systematic review and meta-analysis  of over 1,800 sepsis patients in 22 randomized trials, depicting that lack of a regimen that  would likely achieve a pharmacodynamic endpoint. In this case, prolonged infusion of beta-lactam  antibiotics is associated with a decreased probability of favorable outcome. In this case, favorable outcome or decreased all-cause mortality was favored with prolonged infusion  of these beta-lactam agents. You can see with all agents, it improved all-cause mortality  by decreasing mortality by 30%. We can see in Shephalosporins there was a trend, but in beta-lactam, beta-lactamase inhibitor, piperacil and tazobactam, for example,  and carbapenems, we do see significant decreases in all-cause mortality with a prolonged infusion  versus short infusion of these agents, likely owning to better pharmacodynamic target attainment.  And briefly, provided a figure here depicting maybe some of the commonality among bacterial, viral, and fungal sepsis variants, we see differences in response, but we also see  overlapping commonalities in response. Really, I think, are going to set a stage for future research  in how we battle different sepsis variants, whether it's virulence patterns, whether it's interaction with our immune system, and our dysregulated response for sepsis infection  that are going to be important frontiers of research. So in conclusion, I think the best way to avoid some of these pitfalls, or I grew up with pitfall as an Atari video game, if we  swing over the pit, we need coordinated, interprofessional clinical and operational pathways to make sure antimicrobial regimens that we want for our patients are available.  We need to have patient-specific assessment of risk factors for likely pathogens, including resistant variants. We likely have to employ empiric combination regimens, hopefully,  that are dosed at sound pharmacokinetic and pharmacodynamic principles, focused on the likely or known source of infection. We need to use our molecular diagnostics to maybe have  early heads up of causative pathogens, resistance variants, so we can modify, before the final culture returns, our empiric regimens. And then ultimately, really a plea  just to continue to be mindful, yes, to emphasize the short view of the patient in the now,  but also be mindful of the long view of that patient as they recover from critical illness.  Their risk for infections is likely not gone after we manage this current infection, and the implications of the way we use our regimens, the duration of their exposure,  on the subsequent infections that our patients are likely at risk for. So focusing on the  near term, the short view, but being mindful of the long view, I think are crucial, important aspects to breaking the cycle of inappropriate empiric antimicrobial therapy.  So I thank you for your attention very much, and everyone have a great day. Thank you.  My name is Matthew Tadlock. I'm a trauma critical care surgeon at Naval Medical Center San Diego,  and today I'll be talking about a topic near and dear to my heart, the impact of source control on sepsis. I have no disclosures to make other than to say  source control is only part of the solution. I'm a firm believer in adequate resuscitation  and early inappropriate antibiotics as well, but today we'll be focusing on source control.  First I'll define source control. I'll talk a little bit about timing, and then I'll switch gears and focus on principles in the management of intra-abdominal sepsis specifically.  Now published just about four years ago, the surviving sepsis 2016 guidelines make the  following best practice statements in regards to source control. First, infected foci suspected to  cause septic shock should be controlled as soon as possible following successful initial  resuscitation, and then they recommend a target of no more than 6 to 12 hours after diagnosis, and they say that this seems to be sufficient in most cases.  Now despite these strong recommendations, and in my experience, often usually with inexperienced  clinicians or residents, they spend a lot of time focusing on bundles of care, the one-hour  bundle, the three-hour bundle, the six-hour bundle, and the importance of identifying the source and  timely source control can often be lost in the whirlwind of the patient's care during those  acute moments. However, source control is important, and this is a recent review article from The  Lancet, and they state that observational data showed that inadequate early source control was associated with an increase in 28 mortality from 27 to 43 percent. However, while they  emphasize the importance, they only talk about source control for this paragraph.  So with that, let's talk a little bit about it more specifically. What exactly is source  control? And a definition that has been given that is commonly used is any physical intervention  that is undertaken to remove or eliminate a focus of invasive infection and to restore optimal anatomic function. And so typically we'll talk about the four Ds, drainage,  debridement, device removal, and then ultimately definitive reconstructive measures.  So first, drainage, and that's simple. It's the evacuation of a discrete focus of infected fluid through its conversion into a controlled sinus or fistula, and this is to reduce the  bacterial load. Generally in intra-abdominal sepsis, this is usually a percutaneous drainage  of an abscess, but this could be open surgery. It could be chest tube drainage. It could be a  thoracotomy for an empyema. When we talk about debridement, this is the physical removal of  devitalized infected tissue before the patient descends into irreversible shock. This dead  tissue is a potent bacterial growth medium, and it can trigger a significant host inflammatory response. What I think most of us are most familiar with are necrotizing soft tissue  infections, and most people understand the importance of timing and surgical debridement.  Third, we have device removal, and this is probably what most of us are familiar with,  but this is the removal of devices or foreign bodies that have become colonized with microorganisms,  and then they end up serving as a locus for ongoing microbiology or microbe seeding.  Various types of devices that we use, urinary catheters. One of the more common in the ICU is central venous lines, but also this can be prosthetic joints, vascular grafts,  prosthetic heart valves, and sometimes device removal means a big surgery for the patient.  And then finally, we need definitive reconstructive measures to reconstruct the damage resulting from  both the infection and the management to restore function and well-being. The biggest example in  an intra-abdominal sepsis is either repairing the bowel or putting it back together. For necrotizing soft tissue infections, either wound closure or skin graft or replacement  of those prosthetic devices that are more permanent.  So again, the 2016 surviving sepsis guidelines recommend a target of no more than 6 to 12 hours  for source control. Since these were published, there's been two interesting articles that talk  about source control and timing in sepsis and septic shock. This first study was actually came  out in 2017 and it was a secondary analysis of prospective data collected from 99 Spanish ICUs,  about 3,600 patients in total, about 1,200 of which underwent source control.  Now the patients who underwent source control were actually more sick and they had a poor  compliance to six-hour bundles. However, they had an improved mortality compared to those patients not requiring source control and this was significantly so, about 21% versus 25%.  They then looked at timing and they used the 12-hour mark as the cutoff and of those patients  that received source control, they looked at less than 12 hours or greater than 12 hours and they were not able to find an association with timing in either ICU or overall hospital  mortality. And so they weren't able to make any conclusions about the timing of source control in this study. They looked at a heterogeneous group of septic processes,  including as you can see here on the right of the screen, abdominal, respiratory, urologic, skin and soft tissue and others. This is the second study that came out of Korea in 2019  and it was a prospective multi-center registry-based observational study of 524 septic patients requiring source control. And they looked at the timing of source control from  presentation in the emergency department. Now they had again about 520 patients and the median time  to source control in this study was 13 hours, which is pretty good. And then the vast majority  of patients, 74%, actually received their source control between 6 and 12 hours. At the bottom of  the screen, you can see here the type of source control. 20% received some kind of surgery, but the vast majority, nearly 60%, received percutaneous catheter drainage.  So they found in this study that source control reduced 28-day mortality significantly,  but they could not find an association between timing, 6 versus 12 hours, and 28-day mortality.  But again, this is a heterogeneous group of patients and the vast majority of patients not only received timely source control in this study within 6 to 12 hours and 60%  required percutaneous drainage. And I would argue that there's a little bit different type  of septic process. So how do we reconcile these two studies with the surviving sepsis guidelines?  I think, you know, the Spanish study was a retrospective study. There's probably,  and it's a secondary analysis, there's probably some unintended bias. And in the Korean study,  as I've said, they were pretty good about getting source control in their patients in a timely  manner. And both of them look at a heterogeneous group of processes that require different types  of source control. So both demonstrate the importance of source control and their effect on mortality. But I really don't think any conclusions about timing can be made.  And I still think the surviving sepsis guideline recommendations stand.  So my take-home message in terms of timing is that prompt source control is required.  That target of 6 to 12 hours you should continue to shoot for. That said, timing of source control  does depend on multiple factors. It may depend on the primary septic process. So necrotizing soft  tissue infection is going to be very different than postoperative intra-abdominal abscess from  an asthmatic leak. And then patient factors are important too. Their comorbidities, their  functional status, and then their physiology in front of you at the time you're taking care of  them. So now that we've talked a little bit about source control and timing of source control, going to switch our focus and talk a little bit for the second half of this talk about  intra-abdominal sepsis. As you can see here from this list, there are many, many causes of sepsis within the abdomen. Various types of perforations from large or small bowel.  You know, some surgeons will say for appendicitis it's either the easiest or most difficult case  you've ever taken care of. The biliary system, and then also mesenteric ischemia, pancreatitis,  and other types of intra-abdominal sepsis. I'm going to just briefly touch on damage control  surgery. Over the last 20 years in trauma we have found that at least in abdominal trauma  that  the principles are to get rapid control of hemorrhage, restore perfusion, and get the patients out of the operating room as soon as possible so we can get them to the ICU  and restore their physiology. And then once they're stabilized, we take them back to the  operating room for definitive repair. And as we've developed these techniques, people have asked,  well, is that an appropriate strategy for our septic patients with intra-abdominal sepsis? And  generally three strategies have been described. First is that damage control approach and routine  open abdomen, get them into the operating room and back to the ICU as soon as possible, leave  the bowel in discontinuity. Or you take care of that septic focus, you do everything you need to  do, and then do close the abdomen permanently and take them back only if needed or relaparotomy  on demand. And then finally do a planned second look for every case of intra-abdominal sepsis. So those are the three general strategies that have been described, and admittedly  two and three are very similar, but there are some differences. So as we were doing damage control surgery for our trauma patients, there's two randomized controlled  trials that kind of looked at planned versus on-demand laparotomy and abdominal sepsis. The first study here was a randomized controlled trial of 232 patients, and they found no  difference in mortality. On-demand was 29% mortality, and then planned relaparotomy was 36%, but this difference was not significantly significant. Importantly, two-thirds of these  planned relaparotomies were negative. They didn't need to happen. And then on-demand laparotomy was associated with lower ICU and hospital stays or decreased days  and decreased direct medical costs. Another study the same year  published their results of 40 patients, which they stopped after an interim analysis, and they found that in the planned relaparotomy group, there was an increased relative risk of  death and odds ratio, and so they concluded that this was the wrong strategy.  And with that, there are several negative consequences to leaving the abdomen open.  Similar to an intubated patient where our goal in the ICU is to get that patient extubated as soon as possible, the same principle applies with those with an open abdomen.  We really want to close that abdomen as soon as possible. And some of the negative consequences,  first is an entero-atmospheric fistula, and this can happen in up to 9% of open abdomens, and this risk increases every day that belly remains open.  Second, open abdomens are associated with large ventral hernias, and once they recover from this  episode of intra-abdominal sepsis, several months to a year or two down the road, they're going to  need a major abdominal wall reconstruction to repair that hernia. And then finally, and most importantly, there's an increased mortality associated with delayed fascial closure,  and so for these three reasons, we really need to get those bellies closed as soon as possible.  And based on this, the Surgical Infection Society revised their guidelines, and they currently  recommend in terms of intra-abdominal infection, they recommend routine source control procedure,  unless a non-interventional approach is associated with a good clinical outcome. And I think the  best example that I can think of is intra-abdominal abscess less than three centimeters, usually those  can be managed with antibiotics alone. They recommend that you undertake source control in the abdomen within 24 hours of diagnosis, however, if they're in septic or septic shock,  that the source control needs to be done in an emergent manner. And they recommend sort of a step-up approach. You want to use the least invasive approach possible to achieve source  control. Temporizing approaches are okay with patients with major physiologic instability.  This might be a case where I leave the belly open if they're very sick in the operating room,  and I need to get them back to the ICU and resuscitated. And then they recommend against the use of planned relaparotomy as we've discussed. That said, there are some times  when you need to leave the belly open and you can't close. One of the most common examples is  in cases of mesenteric ischemia. Whatever the cause, you'll restore blood flow to the bowel,  you'll remove definitively dead bowel, but you can't tell sometimes whether all the bowel is going to survive or not. And these cases typically mandate a second look. Sometimes you  may have an abdominal wall necrotizing soft tissue infection with retroperitoneal involvement,  and you're unable to achieve source control in that initial operation, and that's going to mandate  another look or more, or even more. And as I've alluded to before, the patient physiologic status  in the operating room may dictate. And then finally, in patients with intraabdominal sepsis,  sometimes they can develop significant bowel wall edema, and you're unable to close the abdomen.  And there's things that we can do to check and see if we think there's going to be a large risk  of intraabdominal hypertension in the operating room. And sometimes we'll leave the belly open  in those cases as well. Finally, I'll touch on just percutaneous drainage. As I said before,  most abscesses in the belly less than three centimeters can be treated with antibiotics, but generally it's recommended for greater than three centimeters that those are drained  percutaneously, particularly in postoperative abdomens. It's really a huge risk for that patient to go back to the OR and operate on them again, and so it's ideal to drain those  infections. And the antibiotics resuscitation and percutaneous drainage will often very quickly  resolve their ileus and restore their normal function and bowel function. And then one of the pet peeves that I have are percutaneous cholecystosomy tubes. So I'm just going to  briefly touch on that. Often you'll have a patient in the ICU who's quite sick and septic  shock or pretty sick and they have cholecystitis and they're not a good candidate for operative  intervention. And there are some, you know, some observational studies that suggest timing of source control or timing of percutaneous cholecystosomy tube does matter. And I'll just  talk about that. This first study here looked at about 200 patients and patients with cholecystitis  and septic shock, and they found that a delay in source control, either operative or percutaneous  tube drainage of greater than 16 hours was associated with an increased mortality.  And then this next study, a smaller study, about 50 patients, they looked at patients with severe cholecystitis and they looked at the impact of delay in timing of cholecystosomy  tube placement to what happened at their ultimate definitive surgery. So every day that there was  a delay in placement of cholecystosomy tubes, there was a higher conversion rate from laparoscopic  to open cholecystectomy in operation. So if it's, if it's 2 p.m. and your patient needs a cholecystosomy tube, it's better if it happens that afternoon or evening as opposed to the  next day. So advocate for your patients in that regard. So in conclusion, we've talked about the  impact of source control. We've defined what that means. And, and really that target of 6 to 12 hours  after diagnosis, you really need to strive for that goal. And because adequate and timely source  control is associated with improved survival for your patients. That said, timing of source control  does depend on multiple factors. It depends on the primary process, patient comorbidities, their immune functional status, and the physiology right in front of you.  I'd like to thank STCM for having me talk today and thank you for your time.

Video Summary

Dr. Ryan Mavis, an infectious disease and critical care physician, presented a talk entitled "Sepsis: The Source Matters" at STCM. He emphasized that sepsis is not a single disease, but rather a collection of different disorders with common features such as organ failure. Dr. Mavis discussed the importance of individualized approaches to care and the need for timely and appropriate antimicrobial therapy. He highlighted common pitfalls in antimicrobial therapy selection, including missing the causative pathogen or using a regimen that does not cover the pathogen due to resistance. He also discussed the importance of site penetration and pharmacokinetic and pharmacodynamic dosing strategies in achieving appropriate antimicrobial therapy. Dr. Mavis stressed the significance of source control in the care of septic patients, recommending a goal of source control within 6 to 12 hours of diagnosis. He cited studies showing a correlation between inadequate early source control and increased mortality. While timing of source control may vary depending on the specific patient and disease process, the overall goal is to remove or eliminate the focus of infection as soon as possible. In the second half of the talk, Dr. Matthew Tadlock, a trauma and critical care surgeon, focused on source control in intra-abdominal sepsis. He discussed the principles of drainage, debridement, device removal, and definitive reconstructive measures. Dr. Tadlock explained that prompt source control is crucial in intra-abdominal sepsis and highlighted the negative consequences of leaving the abdomen open for too long, such as entero-atmospheric fistulas and large ventral hernias. He referenced the Surgical Infection Society's guidelines, which recommend routine source control procedures within 24 hours of diagnosis, unless a non-interventional approach is associated with a good clinical outcome. In cases where immediate closure is not possible, such as with mesenteric ischemia or abdominal wall necrotizing soft tissue infections, a planned second look or ongoing monitoring may be necessary. Dr. Tadlock ended by discussing percutaneous drainage in intra-abdominal abscesses and the importance of timely placement of percutaneous cholocystitis tubes. In conclusion, both Dr. Mavis and Dr. Tadlock emphasized the critical role of source control in the management of sepsis and the need for individualized approaches based on the specific patient and disease process.

Asset Subtitle

Sepsis, 2021

Asset Caption

The Society of Critical Care Medicine's Critical Care Congress features internationally renowned faculty and content sessions highlighting the most up-to-date, evidence-based developments in critical care medicine. This is a presentation from the 2021 Critical Care Congress held virtually from January 31-February 12, 2021.

Learning Objectives

Review differences in antimicrobial spectra, tissue penetration, and resistance that may influence therapy selection in the patient with sepsis
Identify pathogen-specific factors, including virulence factors and immunologic responses, that impact the course of disease in the patient with sepsis
Recognize the role of early source control in sepsis and septic shock, including an understanding of the need for early expert consultation

Common Pitfalls in Antimicrobial Therapy Selection
Eric W. Mueller

Drugs Won't Save You: The Impact of Source Control
Matthew D. Tadlock

Meta Tag

Content Type Presentation

Knowledge Area Sepsis

Knowledge Level Foundational

Knowledge Level Intermediate

Knowledge Level Advanced

Membership Level Select

Tag Infection

Tag Guidelines

Year 2021

Keywords

Dr. Ryan Mavis

infectious disease

critical care physician

sepsis

source matters

organ failure

individualized approaches

antimicrobial therapy

source control

intra-abdominal sepsis

	Society of Critical Care Medicine 500 Midway Drive Mount Prospect, IL 60056 USA Phone: +1 847 827-6888 Fax: +1 847 439-7226 Email: support@sccm.org	Contact Us About SCCM Newsroom Advertising & Sponsorship DONATE	MySCCM LearnICU Patients & Families Surviving Sepsis Campaign Critical Care Societies Collaborative	GET OUR NEWSLETTER
© Society of Critical Care Medicine. All rights reserved. \| Privacy Statement \| Terms & Conditions The Society of Critical Care Medicine, SCCM, and Critical Care Congress are registered trademarks of the Society of Critical Care Medicine.