Extraordinary might be a strong word, but I'll try to live up to it. I feel like we should make Dr. Taylor an honorary pharmacist based on all of the judicious antibiotic use commentary, so very much appreciate that. So, for some context, I have been a MICU and trauma ICU pharmacist at the Mayo Clinic in Rochester, Minnesota for about eight years. I also do some work in the emergency department. But more recently, they have made the decision to make me our ICU stewardship pharmacist. So now I have to answer to critical care and ID, and it, you know, getting stuck in the middle there gets challenging sometimes. I have no disclosures, and my objective today is to identify strategies to improve time to antibiotics in sepsis. And one of the things that I wanted to be very clear about here that Dr. Taylor did a great job of setting me up for is I am not telling you to take everything I'm going to tell in this presentation and give antibiotics to everyone within the first hour. I think I would lose my ID cred if anybody found out from the ID world that I did that. So, we've seen this slide a bunch of times already, but that guideline recommendation of septic shock or a high likelihood for sepsis, getting antibiotics within the first hour of recognition. As we all know, this data is all observational. It's technically a very low quality of evidence. But I also wanted to point out, this recommendation is almost entirely studies of shock and isolated shock, not the whole spectrum of patients presenting with sepsis. And so, when this came out a few years ago, IDSA and emergency medicine organizations did not endorse this guideline. And it was because of a variety of factors, but I want to focus on the timing piece specifically. IDSA said, look, if you're telling everyone you have an hour to give antibiotics, everyone's going to get antibiotics. There's going to be an excessive broad use in a huge proportion of patients. It's also important to note that at that time, and even now, in patients that don't have shock, three hours is seen as an acceptable target for antibiotic administration by the Senators for Medicare and Medicaid, I can't talk today, Centers for Medicare and Medicaid, as well as national quality organizations. The emergency medicine societies added in that the challenges of using QSOFA to find sepsis in the first hour in their patients is very difficult, as well as time zero being patient triage. It would be nearly impossible in many of our very busy emergency medicine settings. And it's also unclear that a one-hour bundle target for all patients with sepsis is either achievable or without harm. And when we think about what are the sources of bias in sepsis timing, and again, here I'm talking about the spectrum of sepsis, obviously anybody with septic shock, antibiotics within an hour is very well established and should be our goal target. And in 2020, Weinberg and colleagues published this really excellent overall review of where do the problems come from by trying to blanket make statements about timing of antibiotics and sepsis. And they came up with four major issues. One, lack of consistency in illness severity in the studies. So again, you've got studies that were only in shock versus studies that are across the whole continuum, and putting those patients on one scale isn't necessarily the most accurate thing to do. What is time zero in these studies? Is it ED arrival? Is it time of illness in the outpatient setting? Was it time in the ICU? Was it when someone on the floor noticed the patient is now septic? There's a lot of other confounding variables that I could spend probably an hour talking about all the potential confounders, but, you know, things like what is your physician's suspicion for sepsis? What are the other conditions the patient has and the patients have in the study? What is the setting of the study? Is it the ED? Is it the ICU? Is it the general hospital? And finally, they bring up this point that we apply a linear outcome to a non-linear process. And I've got a graphic displaying this. So this study is from Ferrer and colleagues in 2014. And they looked at about 17,700 patients across their spectrum of sepsis care and looked at with the impact of a delay in antibiotic to overall mortality. And what you have here is the representation of every hour delay and the associated mortality along the y-axis. And you can see that this is a curve, right? Clearly hour one, hour one to hour two, and hour two to hour three, you know, it doesn't really change across the whole spectrum of sepsis presentations. Once you get beyond three hours, and especially beyond four or five hours, the mortality increases significantly. And so it's a little bit disingenuous to take data like this and say every hour increase in delayed antibiotics is 10% mortality increase for all patients. It's just not the way that it actually plays out in the overall literature. And so, again, if someone is in shock, give them antibiotics. Give them antibiotics as fast as you possibly can. If somebody is in that more gray area, we have some time to evaluate the patient more before we rush to the antibiotic step. Here I've just got the numerical display of that data, really focusing, again, on that hour two to three and hour three to four as far as the overall odds of increased mortality. And again, you can see that it really picks up once you hit that greater than six-hour mark. And this is all-comer patients in sepsis, not just septic shock. So let me come to that. The other piece of the IDSA recommendations were possible sepsis without shock, time-limited course of investigation, and then antibiotics within three hours. So if you want to take everything else I'm going to say today and apply it to three hours, awesome. Go for it. Just don't tell anybody that I told you to do it all in one hour. So for my purposes, what is time zero in our population? You know, the CMS follows a three-hour timing from sepsis recognition. But time zero, again, it could be the ED presentation. It could be the clinical decompensation in the hospitalized patient. It could be the actual ICU admission. So for purposes of the rest of the talk, time zero for us is going to be one of you hit enter on your order to get the antibiotic in. You decided the patient was septic, you decided what antibiotic you wanted, and you put the order in. So we have talked a little bit about these patient locations for sepsis onset, and I'm going to spend a very brief amount of time focusing on what the ED literature has done to improve time-to-antibiotics in their patients. But as you'll see, this really is hard to apply across the spectrum of care in the entire hospital day. The study by Hitty and colleagues is a great example of this. They looked at all of the stages of a patient care in the ED, and here's their Pareto chart for how to optimize the time-to-antibiotic. So if you look over here, I should have a, there we go, all right. So the blue bars are the actual time in minutes that it took for something to happen. The red line is the cumulative percentage of time, and the green line is the average time. So from patient arrival to the ED to being triaged was about eight minutes. From getting triaged to getting roomed was about 25 minutes. From getting roomed to seeing a physician was 10 minutes. It took about an hour from the physician to see the patient to put the antibiotic order in, and then it took 54 minutes from the order to get in to when antibiotics went into the patient. And so they said, okay, this is the piece that we saw as a problem that we're going to try and fix. And their overall solution was to put vancomycin and purposylantazobactam in the automated dispensing cabinet in the ED setting. As you can imagine, when the antibiotic is literally in a cabinet next to you, you can give them much faster. So there were about 55 to 60 patients in each pre- and post-group when they did this study, and they found that pre-intervention, again, 55 minutes from the order to the antibiotic, but only 26 minutes from the order to the antibiotic infusing if it was in the automated dispensing cabinet. They also had an improvement in their patient arrival to the ED to antibiotic time from two and a half hours to about an hour and a half. We'll go over shortly why antibiotics in every dispensing cabinet is not a very viable solution. But then we get to the hospital, right? So much less data exists regarding the hospital onset of infection and proven antibiotic timing. Code sepsis really revolves around appropriate recognition of sepsis, and the key indicators of success in these bundled interventions involve multidisciplinary team members. So I'm going to have to disagree a little bit with Dr. Simpson in that I think you need a pharmacist on your sepsis teams because of all of the things that I'm going to talk about next. And it may not need to be an ICU pharmacist, but certainly someone at an operations level who can help you find the problems in getting the drug to the patient. It's also very important that you understand your site-specific barriers to getting antibiotics into the patients. And this takes a quality-focused approach, again, with a very multidisciplinary team so you can really understand all the facets. So I have the easier part of, you know, we already know what sepsis, we know where the source is, so how do we fix these three things? Order the antibiotic, get it to the floor, and administer it. So we have a little fishbone for you. So out of everyone here who's not a pharmacist, do you know what happens after you hit Enter on your order before the antibiotic gets to the patient? Raise your hand if you know. Okay, well, obviously you know, but does anybody else who's not? Okay, we got some people back there that know. So here's the question for antibiotic ordering. Does your code sepsis team have a laptop? Do they have access to a computer to put the order in for the patient? Did they order the right drug? Is it the right dose? Is it starting at the right time? Did they order it stat? What else is your pharmacist doing that should be processing those orders? Need an antibiotic delivered to the floor, again, does the central pharmacy know it's a stat order? How long does it take to remix the antibiotic? Did you move the patient? Did you order it on the internal medicine floor and then transfer the patient to the ICU? That antibiotic is going to where the patient was when you actually put the order in. Can you use a tube system or does it have to be hand-carried? And where does the drug go when it actually hits the hospital unit? And then finally, antibiotic administration. Does your nurse have enough line access? Do they know you put this order in? How long does it take to infuse the antibiotic? What are their competing priorities for getting the antibiotic in? And we're going to go through all of these in more detail. So let's go back to the antibiotic ordering question. So again, consider where the sepsis is identified. Do they have access to actually get the antibiotic ordered? Do they have verbal ordering? Can the nurse in the room put it in to get the process started? Is it the right drug and dose? Every pharmacist in here can think of 1,000 times when someone's ordered a stat order and it's just not the right thing. They're allergic to it. The dose is wrong. And if we don't have collaborative practice agreements, legally we're not allowed to fix that. We have to call you and tell you, you need to change the order to this. What is the timing of the order start? And was it ordered stat? So it's not ordered stat. No one knows that it's stat. Electronic health records are not infallible. A common problem that I deal with every single day is somebody says, I want this ceftriaxone to be every 24 hours. I'm going to order it daily. Turns out daily defaults to 9 a.m. So if you're not paying attention, someone could need a dose of ceftriaxone at 6.30 at night. If it's ordered as daily, it's not going to get there until 9 o'clock the next morning. And then finally, what is the time to pharmacist verification? Do you have a trigger for awareness of order importance? Okay, I've got another raise hands question for everybody. So anybody who's not a pharmacist, do you know what the pharmacist actually does after you put the order in? Like, have you seen the screen? I've got a couple of people. So I've got kind of a nerdy example, because I wouldn't be a pharmacist if I wasn't a nerd, but all right. Here is my, I'm the pharmacist, this is my verification queue. So let's say I'm an internal medicine pharmacist and I'm covering three floors of patients, so I've got about 60 people that I'm responsible for. And this is what I see in my queue. I see a patient name, a location, how long the order's been in the queue, and who ordered it. It doesn't, I don't know what is within the order. I don't know that it's an antibiotic. And you see the red bar across the top there as a representation of if you put something in stat, it'll flip to the top of the queue, so I know I should do that first. But people order a lot of things stat. I can't tell you how many times I've had to process a stat docusate order. We might disagree about how relevant that is in a stat situation, but if you don't have the order support for your pharmacist and they have too many patients that they're taking care of, they don't go on the code sepsis call. They don't know that Mary Brandybuck in room 456 half an hour ago was septic and needed the antibiotic as quickly as possible. And so this is why you need pharmacy operations input to see is there a way you can work with your EHR provider to trigger, this is a septic patient, or this is an antibiotic order, or something along those lines. Then we think about order to the floor. So again, was it placed stat? How long does it take to mix up? Where is the patient? Like I mentioned before, if you put in an order for piperacillin-tazobactam on the floor because it's your code sepsis team and you've responded and you know it's sepsis, the label will print in the central pharmacy with the patient location of the floor, and there's no way for them to go back and fix it until that drug is delivered to the unit the patient is no longer at. And then was it shift change? Does anybody on that unit know the patient is now in the ICU and needs this antibiotic? It's questionable. The other thing is on the side of the receiving end, let's say in the ICU. Drugs can be delivered either by hand carry. A lot of hospitals have pneumatic tube systems, but they also have little trains that will just deliver a whole bunch of stuff to the unit. So who's responsible for checking the tubes and the trains? Is it a nurse coordinator? Is it a health unit coordinator? Do they know that the antibiotic is important and should be given to the patient? Do they know that it's stat? Or do they go put it in the patient's cubby for the room number because, oh, here's a drug. The nurse knows she'll get to it when she can. And then for everybody who's not a pharmacist, I wanted to go through and just let you know the actual step-by-step process of what getting an antibiotic to the patient looks like once you've hit enter on your order and I've verified it. Best case scenario, it's an antibiotic dose that is standardized that your hospital has pre-purchased. So we've got little bags or, you know, vials that have the right dose and ready to go. The label prints in the pharmacy. The correct pre-mixed antibiotic is identified and labeled and then is checked by a pharmacist and goes on its way to the patient. If you're doing anything non-standardized or you had to do a specific dose composition or like, for example, many vancomycin doses will need to be compounded, the label prints in the pharmacy. They have to find the components. So this is powder drug, diluent for the drug, and whatever the bag is that you're going to deliver it in gets passed into a sterile hood. They put the drug in a solution, which can take, if you're doing ceftazidime, it's going to take you probably 10 minutes to get that into solution. Then the drug solution is put into an IV carrier and then it's checked by the pharmacist and goes up the stairs. So it really isn't an instantaneous process of, well, I put the order in. I'm good. My job's done. There's no other problems that could come up to get this drug to the patient. You might recall that earlier I mentioned, you know, well, the ED just put them in the cabinets. Can't we do that everywhere? If you propose to your pharmacy administration that you want antibiotics in every automated dispensing cabinet, they're going to lose their mind. And this is because space is at a premium in these cabinets. And how many of you that are not pharmacists have direct interactions with the big drug dispensing cabinets in your units? We got a few more hands here. Space is at a premium in these things. And you don't want to take up space with drugs that you're not going to be using on a daily basis if you can help it. So there's not really an infection ward in the hospital, right? So if you wanted to put, say, ceftriaxone in every machine in your internal medicine ward or in your ICU, you're taking something else out that will be used every day. There's issues with beyond-use dating and wasting the drugs. There's issues if you just put the vials of drug in the machine, if the nurses are mixing it up and it's a non-sterile preparation, that increases the patient's infectious risk. And the feasibility of doing this really decreases as the hospital system gets larger. And what I've got here is a study from Lowe and colleagues that was published in 2014 that looked at this. And they looked at piperacillin tazobactam, 3.375 grams in all of their automated dispensing cabinets. As a pharmacist and as a stewardship, and really as a person, this hurts me deeply. Because if the drug is there, they'll use it, right? So everybody's going to get piptazone out because it's right there. They looked at 65 pre- and 56 post-implementation patients. And granted, at the time of the study, the hospital used handwritten orders, so that didn't help their timing. But they did find that having a drug available in the cabinet decreased the time to dose from 4.5 hours to 2.9 hours. So there are situations when drugs in the cabinet is a reasonable thing to try. Let's shift over into the antibiotic administration situation. So you've got your nurse at the bedside. Her patient is septic. You say, OK, they've got to get the fluids. They've got to get the antibiotics. But you've got to wait for them to come up from pharmacy. We've got to do pressers now. Oh, and we're going to intubate, so get all the RSI stuff ready. Oh, and we've got to get the labs, we've got to get the blood to get everything else. And so your bedside nurse has a lot of competing priorities. And depending on the nurse's experience, depending on what's happening with the patient, it's very easy for something like, I need to hook the antibiotic up and get it started, it can get missed when you've got somebody who's critically ill. What about time to infuse antibiotics? When you're ordering broad-sepsis antibiotics, let's use the example of vancomycin, priprocyl and tazobactam, and levofloxacin. This chart here shows you the safe administration times for these antibiotic doses, time in minutes is on the y-axis. It takes an hour and a half to infuse Levaquin, an hour and a half to infuse most vancomycin doses. If your staff doesn't know that the beta-lactam needs to be first, you could have a delay of the most important antibiotic by up to three hours just because of line access and a lack of awareness of what should go first. So things like ceftriaxone, you know, it's 15 minutes, priprocyl and tazobactam, cefepime, meropenem, 30 minutes if you're using the IV piggybacks. So it's extremely important that your educational initiatives when you do these sepsis bundles really focuses even at the bedside level in the nurse for something that seems as simple to us as what needs to go first. So again, order is very important when it comes to giving multiple antibiotics at once. And can we optimize these infusion times? Thankfully, out of the emergency department setting, there is some good safety data for a lot of IV push beta-lactams. Primarily this is ceftriaxone, cefepime, meropenem, and erdapenem. Unfortunately for us, penicillins do carry a risk of seizure with rapid administration. So we're not going to be able to get much faster than 30 minutes for that antibiotic. The logistics of this are very complicated. So if you're looking to implement this at your hospital or within your emergency department, I really encourage you to look at this reference by Spencer and colleagues. They went through all of the data, all of the beyond-use dating, all of the compounding logistics to make it feasible. We do this in RED at Mayo Clinic in Rochester. And it is amazing to have cefazoline, ceftriaxone, cefepime ready to go so that we can really shorten that time to administration. And then you might say, OK, well, do bundle options exist then? Because we do bundles for everything else. So surely a bundle is going to be great here. And we do have a study on this. Amalki and colleagues looked at a surgical trauma ICU, and they looked at their barriers to antibiotics in the SICU via a survey and direct observation of pharmacists, physicians, and nurses. And their intervention was put meropenem and perpicillin-tazobactam in the dispensing cabinet. And again, this physically hurts me, because if somebody has meropenem ready to go, they're just going to give it. But they also did monthly education regarding the importance of rapid antibiotics. They did antibiotic dosing tables attached to every computer in the unit. So any time anyone was ordering, they would get the right drug and the right dose. And they also did a follow-up of monthly reports on success. And for their overall results, they found that 26% of antibiotics versus 84% prior to this intervention were given within 60 minutes. However, the time change only improved in the automated dispensing cabinet antibiotics. So it only improved for meropenem and perpicillin-tazobactam. So the solution to this problem of time to antibiotics in the hospital is not only going to be solved by putting antibiotics in a very limited space. So for some key takeaways, obviously, one-hour administration in septic shock is well-established, and no one's going to argue with you. But if you've got that not-as-clear, potentially septic, really focus on that less-than-three-hour goal. You have to understand your specific institutional barriers to antibiotic administration. You need to have multidisciplinary stakeholders for these quality initiatives. Every hospital is a little different. And so really having that operations side of the pharmacy to tell you where they can flex and where they can help you meet some of these steps a little bit faster is going to be very key. Consider IV push-beta-lactams and nursing education for order of administration. And really, the first antibiotic orders should always be stat in these patients. There should never be a new antibiotic that's just ordered to start whenever. And I'll be happy to answer questions at the end of the round of presentations.

improving time to antibiotics

sepsis

timely administration

septic patients

one-hour administration guideline