lauren-p--black_1_08-12-2025_115919: I think this is one where everybody I work with knows how much I dislike this test in the ER.

I think it's now synonymous with me.

Sam: Hi everyone, and welcome back to another episode of EMplify.

I'm your host, Sam Ashoo.

Before we jump into this month's episode, I wanna say thank you for being a listener.

I want to encourage you to rate us in whatever app you're listening in so that lots of other people can also benefit from listening.

And I wanna remind you that EB medicine.net is your one-stop shop for all things emergency medicine, pediatric emergency medicine, and evidence-based urgent care.

And if you happen to be going to this year's American College of Emergency Physicians Scientific Assembly in Salt Lake City, I highly encourage you to come by and say hello to us.

We will be at Booth 1 3, 7 7.

That's 1377 from September 7th through September 10th.

Come by, say hello.

We'd love to see you face to face.

And now let's jump into this month's discussion on sepsis with Dr.

Lauren Page Black, whom I think you will find to be just an amazing resource for all things sepsis.

lauren-p--black_1_08-12-2025_115919: Hi, my name is Lauren Black.

I'm an assistant professor of emergency medicine at Northwestern University Feinberg School of Medicine where I am regular ER doctor.

I also teach residents and med students.

And I am also a sepsis researcher funded by the NIH.

Sam: Awesome.

And you are one of five authors for the August 2025 emergency medicine practice article on updates and controversies in the early management of sepsis and septic shock.

So this is kind of a specialty niche for you, something you're passionate about.

lauren-p--black_1_08-12-2025_115919: Yes, it was a big group.

Big topic, big group.

And so thankful for all the co-authors.

Hope I represent them well.

It is myself and Dr.

Faheem Guirgis, one of the other senior authors are both NIH funded sepsis researchers.

So, it's pretty dear to my heart, so that's why I keep picking it.

Sam: Good.

Good.

I'm always surprised by the sepsis statistics, honestly.

And when we talk about epidemiology, number of cases a year, mortality rates those things always tend to jump off the page at me.

And so when I'm looking through this article again, no surprise, but the number of cases of sepsis in the nation.

You know, Here we're talking about 850,000 cases in the US coming through EDs that are related to sepsis.

And depending on how sick you are, severe sepsis to septic shock, mortality can be as high as 40%.

That's really significant.

lauren-p--black_1_08-12-2025_115919: It is really significant.

And, in addition to its human toll, with exactly like you said, at least a 10% mortality for general sepsis, as you get to septic shock, four out of every 10 patients will die.

It's also one of the most expensive reasons for hospitalization in the United States and is unfortunately the ultimate common pathway for a lot of people who die in the hospital.

Some estimates are, it's up to one out of every three hospital deaths ultimately are attributed to sepsis.

Sam: And therefore the big effort from government agencies, hospital associations, specialty organizations, to try and focus on things like tools to help us detect sepsis early and early goal directed therapy, and all this effort that we put into detecting it and treating it as quickly and early as possible.

Right.

lauren-p--black_1_08-12-2025_115919: Yes, which has been met with a lot of challenges because sepsis is far more of a syndrome than a discrete disease.

For many diseases, the patients themselves are different.

That's not unique to sepsis, but sepsis is far more of a syndrome, and the fact that the infection types are different, the patterns of organ dysfunction are different.

And ultimately this leads to a lot of heterogeneity and a lot of really different presentations, some subtle organ dysfunction for some patients that makes it really hard to screen for and, honestly, something a lot of the times we can recognize it after the fact, but it is hard upfront sometimes to recognize that the patient themselves is septic.

But certainly, especially for the more severe patients with of hypoperfusion or hypotension, time really is important in recognizing and treating sepsis in those patients.

Sam: And then from the government standpoint and from the definition standpoint, there's been a significant period of change, I'd say in the last 10 or more years, where we kind of keep evolving the definitions.

Where are we now on that?

lauren-p--black_1_08-12-2025_115919: Yeah, that was one of my biggest goals in the paper we wrote that pairs with this podcast is one of my goals was to really reinforce and clarify the difference in consensus definitions and what the most updated consensus definitions are, as well as clarify where those are similar and different than CMS sep-1, which is the government quality bundle that hospitals are held accountable to.

So, first I'll take on consensus definitions.

So, that has changed quite a bit.

Sepsis is a dysregulated host response to an infection characterized by organ dysfunction in the setting of a suspected or confirmed infection.

So what the simple version of that means is if somebody has an infection or suspected infection and evidence of organ dysfunction, they're considered septic.

So prior to sepsis three, sepsis was really defined by SIRS criteria with a second category called severe sepsis and a third septic shock.

In sepsis three, sepsis infection plus organ dysfunction is really considered sepsis now .

Severe sepsis went away.

Septic shock maintains mostly its similar definition.

It's a vasopressor requirement to maintain a MAP above 65 with evidence of hypoperfusion usually considered a n elevated lactate.

As the sepsis three definitions change partially because as we all know, SIRS is not very sensitive or specific.

Old people don't necessarily mount a thermoid or leukemoid response the same way young people do.

SIRS is, in a lot of ways, representative of a appropriate host response to an infection.

If you or I got the flu, we'd have SIRS criteria.

It doesn't necessarily mean we have organ dysfunction, and it certainly doesn't mean we have a dysregulated host response to an infection.

By the same token, the elderly patient or the immunocompromised patient may not mount a white count, may not have a fever, but they may have pretty significant renal dysfunction intermittent hypotension, et cetera.

That patient is clearly septic.

Those of us with the flu need some ibuprofen and, you know, to move on.

And so the definitions changed largely because of that nuance.

I think what makes it difficult is the CMS sep one bundle has some slight differences with the consensus guidelines.

And I think that adds a lot of confusion to the topic.

Though the CMS sep one bundles technically still rely on a SIRS based definition, there's really only two bundles you fall into.

So it's the severe sepsis bundle, which again, very closely mimic sepsis three sepsis definition, as well as the septic shock bundle.

One of the differences there though, is CMS sep one defines septic shock as hypotension, even if they're not on vasopressors or a lactate greater than four, even if they're not on vasopressor.

So that's what triggers the septic shock CMS bundle of the fluid requirement component of the bundle.

But since those definitions are slightly different, I think that does add some confusion, which is one of the reasons we tried to clear that up, in table two a little bit, highlighting the sepsis three definitions, the fact that severe sepsis is no longer a contemporary term, and compare it with the CMS sep one core measure definitions to hopefully add some clarity there.

Sam: Yeah, so, the CMS world is where hospital core measures come from and how hospitals are ranked and how their performance is measured.

And still very important to us in the emergency department because if a hospital's being measured by it, we will be measured by it as well as willing participants in that relationship.

But then we also have the specialty society definitions, kind of like the clinical side of medicine.

And that's more, I wanna say that's more research based.

But is it more really just consensus opinions that have changed over the last 10 years, or is there good evidence that's driving that change?

lauren-p--black_1_08-12-2025_115919: I think that is a loaded question.

I think both.

I think a lot of it is consensus based.

And then I do think there's some really good evidence that drives some of it, but a lot of it certainly is consensus based especially where the evidence just doesn't completely answer our questions.

And they're both controversial in their own regards and actually they both had a lot of commentary by our emergency medicine societies because they are a bit challenging.

I do think they mostly overlap in the patient cohorts they're describing in both, but I do think it is a little bit confusing that they use slightly different terminology.

I think that does make it hard.

But yes, I do think we are obligated to be aware of both what the consensus guidelines and the society group say as well as the core measures.

Sam: Yeah so then when we're doing our charting and we're talking about hospital core measures, and we want to clarify with our hospital administrator colleagues what we're talking about, we're usually referring to CMS sep one or sep one criteria, which haven't changed in over a decade and are still based on very early consensus opinions of what sepsis is.

And then when we're talking with, say, our critical care colleagues and we're talking about early detection of sepsis, more clinical things, we're using the sepsis three definition.

This is now as of 2021.

Is that the right year?

lauren-p--black_1_08-12-2025_115919: 2016.

There have been some updates since then.

Though I do think SIRS is just hard to you know, separate ourselves from this.

I think the ease of its use early on, I think makes it hard for people let it go

Sam: Yeah that's right.

lauren-p--black_1_08-12-2025_115919: So I think we still see a lot of reticence to sort of let that one have had its time in the sun

Sam: Yeah.

lauren-p--black_1_08-12-2025_115919: But again, I do think, CMS only holds you to a severe sepsis and a septic shock bundle.

And severe sepsis is essentially the sepsis three consensus definition of sepsis.

So I think the confusion is mostly semantic rather than actually describing terribly different groups of people.

Sam: Yeah, and I think you guys did a great job explaining that in the article.

So if you're listening and you have access to this article I highly recommend you go just read that one little section on definitions and terminology.

It's very well laid out.

Now in there, there's a discussion about screening tools as well.

So today, in 2025, do we yet have any good screening tools for sepsis?

lauren-p--black_1_08-12-2025_115919: No, I think they all have strengths and weaknesses.

We still don't have a perfect screening tool.

I'm sure all of us have a hospital BPA that goes off all the time.

And think that's a hard sell to tell clinicians these things aren't working perfectly.

They're not working perfectly.

It's one of the areas I do think will hopefully be addressed by advanced analytic methods and machine learning and some EHR based predictive algorithms.

Though some of those are running and currently available, their performances have been largely mixed

Sam: Mm-hmm.

lauren-p--black_1_08-12-2025_115919: And so I don't think any one of those is functioning perfectly.

I'm optimistic that will change in a few years.

What I will say went away as a screening tool and the most recent updates were qSOFA.

So if you read this paper or were hearing about that a little bit ago, that has largely been retired due to its suboptimal performance.

But I do think all of the screening tools have some degree of something left to be desired with regard to sensitivity and specificity.

Sam: And I see like a common problem with them is that some rely heavily on data that we're not gathering in the ED.

And others that seem to parse that out then become less sensitive or less specific and just not as helpful.

And most of our EHR or AI related tools seem to just scrape for SIRS criteria, which in and of itself is not sufficient either.

So, there's no sweet spot yet that we have.

lauren-p--black_1_08-12-2025_115919: What I will say is I think clinicians are pretty good at it.

And I think part of the problem is none of these have really outperformed clinician gestalt, which we've seen for other diseases and syndromes as well.

And what I tell the residents when I'm working with them is if you think somebody has an infection, ask yourself, do they have new organ dysfunction?

And if they do, you should treat 'em as septic, and then if they've got new organ dysfunction, ask yourself if you think it's secondary to an infection.

And if they do, you should presume they're septic or at least go on a expedition for why they're not.

And clinicians seem to be pretty good at that.

The tools have not performed better than our ability to do that.

Sam: Yeah.

Yeah.

They're not making us better.

Not yet.

lauren-p--black_1_08-12-2025_115919: Not yet.

Sam: Okay.

lauren-p--black_1_08-12-2025_115919: In my opinion.

Some people might have more optimistic outlooks, but it's not one that I personally hold.

Sam: Good.

All right.

Well that's table three, the comparison of the screening tools, if you've got the article.

And so that's things like SOFA and qSOFA and MEWS and a multitude of others listed there along with their sensitivities and specificities, none of which are really all that great.

And so qSOFA, we kind of held up as maybe something that really was going to have a lot of good potential.

And it turns out that even that is not all that sensitive for the ED.

Is that right?

lauren-p--black_1_08-12-2025_115919: Exactly.

So I think where that came from is when they operationalize, and by they I mean the surviving sepsis campaign, when they operationalize the new sepsis, sepsis three definition, a dysfunctional host response to an infection, characterized by organ dysfunction in the setting of presumed infection, they clinically operationalize that as a SOFA score of two or more that was new in the setting of a presumed infection.

I think what's hard is the SOFA score is hard to use in the ED.

We don't always obtain all of those components, and I'm certainly not saying you should.

So I think the desire behind qSOFA was to make something that was easier to operationalize in the ED if you weren't getting PDF ratios on people and it just didn't live up to that ability.

Sam: It didn't do it.

It did not do it.

Okay.

Well, so if we're working in the emergency department today and someone is insisting that we use one of these particular scoring methods, there's not really one that seems to stand out to me to be the best.

So it's more like, use what you're being asked to use, but know that it's going to have limitations and maybe understand those limitations.

Make it no better than your own judgment, maybe worse than your own judgment.

lauren-p--black_1_08-12-2025_115919: Hey, that's exactly what I would personally, yeah.

Great summary.

Sam: Okay.

Alright, well then let's talk about the clinical side of sepsis then.

So there are some people who listen to the podcast who are on the pre-hospital side of it, and that seems to be an important part in the handoff when they come to the emergency department.

Are there things that can help them recognize sepsis in the pre-hospital setting so that they can help alert us to it when they arrive in the ED?

lauren-p--black_1_08-12-2025_115919: Sure.

First of all, I admire our pre-hospital colleagues 'cause if sepsis is hard to screen for in the ED with all of our labs and et cetera it's exponentially in the pre-hospital setting.

I would say things that are really helpful are histories if the patient is not acting right, they have some evidence of organ dysfunction that is obtained on history as well as vital signs in route being abnormal.

And I'm certainly not saying that just because SIRS isn't sensitive enough doesn't mean we should be ignoring fevers or elevated heart rates.

Those are super important.

But, I would say parts of the history where mom and dad haven't been acting quite right for a few days, they haven't been going to the bathroom as often or something l ike that can be really helpful aspects of the history.

But it's also just hard.

If it's hard in the ED with our labs and I think it's super hard in the pre-hospital setting as well.

The other things that can also help are whether or not they're taking meds that may mask of those SIRS criteria.

So if there's that bag of meds sitting there, it's sometimes helpful to bring in, 'cause then you can find out if they're immunocompromised and we didn't know, or they were on a beta blocker or something like that.

And those things can be really helpful

Sam: Yeah, so that's a common theme for our pre-hospital personnel on this podcast is helping us obtain an accurate history from anyone who's on scene and then gathering medications just so we all have the same picture when you get to the emergency department.

Those are excellent tips.

There was a mention of local screening tools, but again, we don't even have great screening tools in the ED.

Is there anything that's been proven to work pre-hospital?

lauren-p--black_1_08-12-2025_115919: To my knowledge, no, because those would also have to rely only on vital sign based arrangements.

And I certainly think they can be helpful in identifying the sickest of the sick.

I also think our prehospital colleagues have great clinical gestalt themselves and if somebody's hypotensive I think everybody knows they're sick.

So I, again, am a little skeptical of the ability of these things to supplement Gestalt in some ways.

People are certainly trying to look at those.

I think it would be helpful if they could work in the pre-hospital setting as well.

I think the linkage of the pre-hospital vital sign data would be ideal if that could happen in the emergency department especially like were they actually really hypotensive?

They got a later fluid and they were responsive and those type of history things are important and it would be ideal I think, if we could link those things to our environment.

'cause I think that provides a lot of rich information.

Especially if it wasn't then their first hypotensive ED vital sign is not really their first hypotensive vital sign.

It was actually 30 minutes earlier when they got picked up.

And I think those things are super helpful.

But I certainly think if we're having challenges in the ED of running screening tools, it's gonna also be pretty challenging in the pre-hospital environment.

I think it's something where, honestly education about sepsis and what populations really do need fluids or prompt antibiotics.

I think that probably is helpful as any screening tool could be.

Sam: All right, so one more loaded question.

I have some EMS agencies that, especially in the rural settings where there's long transport times, talk about early administration of antibiotics en route to the hospital before even ever getting to an ED for suspected sepsis cases.

Is there good evidence behind that kind of approach, or do you think, I mean, your opinion's fine, do you think there's enough in the way of information at your disposal in that kind of setting to go ahead and start antibiotics in that kind of pre-hospital arena?

lauren-p--black_1_08-12-2025_115919: I think perhaps in some populations.

So this I also think gets to a way where some of the literature has kind of been propagated in a way that may not actually represent what the literature actually says.

So the literature for time to antibiotics is fairly clear, although I will say there's always a few papers that say something else.

But in general, broad strokes, time to antibiotics has really reproducibly only shown to have a mortality benefit in patients who are hypotensive or with hypoperfusion.

By that I mean like an elevated lactate.

The original Kumar study, which was fabulous, showed this profound time to antibiotic benefit, but it wasn't from triage, it was from time of onset of hypotension.

We're never gonna be able to redo that study again because it would be totally unethical to hold you know, a largely safe intervention for most people.

But in general, it is in patients who are hypotensive or who have evidence of hypoperfusion.

I'll also say in that Kumar study, the median time to antibiotics was fairly long because this was in the two thousands, but we don't wait six, 12 hours anymore to give antibiotics to most septic people ideally.

So, I will say in the rural environment I do, particularly if somebody is hypotensive, definitely think there could be a role for early antibiotics.

The evidence does not really support that.

Somebody who may just be like a little more confused from a urinary tract infection or maybe a little dehydrated, the evidence is not largely supported.

The emergency, though still urgent nature of antibiotics in that subpopulation.

But I certainly think for somebody who is hypotensive it would be very reasonable in the rural setting to start some antibiotics.

The hard thing is you wanna target those ideally to a suspected source of infection.

So I, do think this stuff gets a little complicated.

Especially if they're that sick, it's hard to know some allergies, et cetera, et cetera.

But I do think in some rural critical access areas, there probably is a role for targeted early initiation.

Sam: That makes sense.

Alright, so let's talk about then when they finally get to the ED and we're trying to get a history besides from our pre-hospital personnel who hopefully have gathered as much information as they can, is there anything new in this arena that we should be asking or looking for?

lauren-p--black_1_08-12-2025_115919: I always just say, well, my mentor said, I'm gonna steal Dr.

Guergis' line myself, treat these people as a trauma, the sicker they are.

So the less they can tell you, the more we need to make sure we expose the patient, make sure we roll them, make sure they don't have a sacral decub.

And I think not forgetting aspects of the history and physical, especially when they could change management.

So we give pretty similar antibiotics to most of these people.

But, we wanna be thinking about things like endocarditis, prostate abscesses, STIs, things that really might change management or things that might lead to a source control procedure and targeting our history and physical exam to those things.

Sam: Good.

And so that goes for physical, like you just mentioned as well, obviously the ideal scenario is you're not going to be examining them in triage, in a chair fully clothed but hopefully, hopefully in a stretcher in a private area where you can fully undress them and look for that source control and see if you can find that open wound or that sacral decub or whatever it is that they might have.

So that part remains the same.

If you can, you should kinda strip them naked and actually get a look at their skin and still go looking for the source.

And is that something that's commonly missed, you think when we're talking about sepsis?

I mean, barring the fact that we're doing examinations and triage.

lauren-p--black_1_08-12-2025_115919: I think the environment in which we're all practicing emergency medicine is hard these days.

And so I, I do think it can be missed, and I'm not saying we need to do that for absolutely every pick 'cause a lot of patients who are septic are not that sick and can talk to us.

But I do think for the very sick patient, we should be doing everything we can to roll and expose them, but we all know how hard it is to roll a bunch of patients when you have a bunch of patients coming in and sometimes it's logistically challenging if it's only you or one other person in the room.

Where I trained for residency we did a lot of rectal temps and I would say probably the best part about that was it made us actually

Sam: Roll the patient, right?

lauren-p--black_1_08-12-2025_115919: Roll every single patient.

I'm not saying we should be doing it as much as we used to when I trained, but I do think one of the helpful parts about that was just you got eyes on your patient.

Sam: Yeah, I will put in the plug for communication with nursing.

If you have an open line of communication with your nurses, you'll find out quickly about infections from Foley catheter insertions or rectal temps all the time.

lauren-p--black_1_08-12-2025_115919: Exactly.

I also think ultrasound can be kind of a useful adjunct to the physical exam here in many cases.

Sam: Okay.

Tell me more about that.

lauren-p--black_1_08-12-2025_115919: I think.

One, I think it can be super helpful to assess cardiac function and see how their heart looks.

I also think we can use it to look for evidence of cellulitis in places where clinically it may be harder to see.

So I think there's a lot of evidence that it can help kind of improve our gestalt.

Sam: Good.

Good.

All my fellow ultrasound colleagues will be very happy to hear that.

So more uses for the uh, the handy ultrasound you carry around your neck instead of the stethoscope, for sure.

Okay, so if we get past history of physical, and now we're looking at diagnostics and we're gonna get some labs, we're gonna have some sepsis bundle we're gonna order in our EMR.

But of those numerous tests, what do we know is helpful?

Now, you've mentioned lactate level a few times as a surrogate marker for organ dysfunction.

Right?

lauren-p--black_1_08-12-2025_115919: Yeah, I think most of this is part of our routine ordering stuff.

Just a CBC and a BMP, I would get on most of these patients really for subtle signs of organ dysfunction.

So there's the floridly septic patient that walks in that you can tell on vital signs alone, they're hypotensive, they look like garbage.

But then, there's a lot of patients where they have really more subtle signs of organ dysfunction and they're septic, they have a creatinine of three and it's usually 0.5 and they have other laboratory evidence of kind of more subtle organ dysfunction.

And in those patients, it's fine to then initiate your sepsis protocol at the time that the sepsis declares itself, which may be then at the time some of their labs come back.

I do think a lactate is helpful if you're suspecting sepsis, but I also don't think you should beat yourself up if you wait to order it until the BMP comes back.

'Cause I think it's okay for our differential to change over the course of their ED course.

I do think lactate still can be helpful.

I don't think it is the holy grail.

We thought it was a handful of years ago.

Nonetheless, it can be helpful.

CMS wants you to do it.

And I do think although there are populations in whom it functions less well, it can be a reasonable evidence of hypoperfusion.

Sam: Great.

And now tell me, you just mentioned this, but what kind of specific populations does it tend to not be as helpful in.

lauren-p--black_1_08-12-2025_115919: So patients with liver dysfunction may not clear their lactate as well.

Patients who are on a bunch of albuterol may have or any beta agonism.

So certainly, if you end up on the third pressor and you start 'em on epi or something, you can't really trend your lactate levels.

That said, the recent guidelines is less towards lactate clearance and more towards lactate improvement in a more generalized setting.

Just because of the way the evidence about lactate normalization played out over the past few years.

Sam: So if I happen to be in the unfortunate scenario of resuscitating a patient with cirrhosis, then a simple improvement in that lactic acid may still be what I'm looking for as opposed to complete resolution and clearance.

lauren-p--black_1_08-12-2025_115919: Exactly.

What I will say is also the CMS bundle, you know, requires for all patients with sepsis, and I'm using the modern definition for that, so the entity formerly known as severe sepsis, get a lactate and they require it be repeated if it's greater than two.

And we can talk about the bundle again in a little bit.

The exception to that though is if the second one is higher, they need a third.

Sam: Hmm.

lauren-p--black_1_08-12-2025_115919: And we can talk about the bundle more, but you know, if it goes from 4.1 to four you're technically done though that patient is probably still really sick

Sam: Right.

lauren-p--black_1_08-12-2025_115919: But technically if it goes from two to 2.1, you need a third to be compliant with, you know, and I think we can put compliance and sort of, what we think the patient needs in hopefully overlapping buckets.

But it's just worth knowing that any lactate above two requires clearance.

If it's less than two, you can stop and you don't need another repeat.

Sam: Okay.

lauren-p--black_1_08-12-2025_115919: But I do think it still provides valuable information.

I just don't think it's a surrogate for all the information we need

Sam: Perfect.

All right.

And then one of my inpatient colleagues' favorite tests, the procalcitonin.

We should get all the time everywhere for everything.

lauren-p--black_1_08-12-2025_115919: Exactly.

I think this is one where everybody I work with knows how much I dislike this test in the ER.

I think it's now synonymous with me.

Um, But I think this is a perfect example of a great test in a different location.

I think procalcitonin can be really helpful, particularly to our upstage colleagues for antibiotic deescalation and a few other scenarios.

However.

As a surrogate decision maker for antibiotic initiation, it has never been shown to perform well in the emergency department.

The long answer is part of that's for a number of reasons.

They peak 12-48 hours after the onset of infection.

Granted, I don't know when onset of infection was.

Nobody really does in the ER, but the point is it may rise too late for it to be helpful in the ER.

It also has really limited sensitivities in really important subgroups for us.

So like though, what's really good for lung infections, it's not really great for atypical lung infections, and we know plenty of people who have gotten sick from those.

It also doesn't perform as well for other infections, like skin soft tissue and immunocompromised patients.

And several studies looked at this and looked at, how could it augment gestalt?

And it has never once been shown to outperform physician gestalt, especially when studied in the emergency department environment.

I will put a caveat here that I'm talking about adults.

The pediatric world, that discussion is, very different.

But in adult patients, if you think they need antibiotics, like you should order antibiotics.

And if you don't think they're septic, you can just document that.

If you think their vital sign abnormalities are 'cause they have a GI bleed or tamponade or something else, and I don't think you need a procalcitonin.

In fact, I think that's a inappropriate use of a procalcitonin.

In that scenario.

I mean, just like outside of a few specific situations, you don't really need a VBG on coding patients to tell you that their pH is bad.

Um, you know, I think this is one of the things where you just have to anchor on your , clinical gestalt and it's okay to do that,

Sam: Good.

lauren-p--black_1_08-12-2025_115919: And the evidence suggests that's the best in this scenario.

Sam: And then when it comes to imaging, this is all just guided by your examination and your concern for occult infection areas where you might not be able to see it as best as you can get from history and physical.

lauren-p--black_1_08-12-2025_115919: Yeah, nothing has really changed about this in the past few years.

There was, I think a brief period of time people were talking about medical pan scans.

I certainly don't think the evidence supports us doing that right now.

Who knows what I'll be saying in a handful of years.

What I will say is I do think judicious use of imaging can really help us find sources of infection.

The particular population I do think we should have a pretty low threshold for, in particular abdominal imaging, is the elderly patient.

And I think we all know that.

And I think if they have a pretty unremarkable UA and it's like, eh, that could be a UTI.

But they look like absolutely terrible.

Those are the patients where I think anchoring on the UA may be inappropriate.

And it's probably best to at least consider abdominal imaging, especially because, sicker elderly patients, immunocompromised patients may not localize infections as well.

I mean, I think we all probably have several cases of this we've seen over the years.

So I'd say I would have a low threshold, especially with the general ease in most ERs of obtaining CT imaging, of considering abdominal imaging in elderly and immunocompromised patients bellies.

Particularly 'cause often there's a source control procedure there that can change management, or I won't say often, but enough of the time that it's relevant.

Sam: It's relevant.

Perfect.

Okay, then let's get into that CMS bundle for a second.

'cause now we're into that kinda initial management, that first few hours period.

Where are we in that requirement for the bundle?

Has that changed at all in the last few years?

lauren-p--black_1_08-12-2025_115919: It has changed a little bit in the past few years.

What I first wanna say about the bundle is, not everybody with the bundle requires 30 ccs per kilo by CMS, which I think sort of the way the bundles are written, it can be confusing.

But essentially there's a severe sepsis bundle, which is what we now call sepsis.

So if you think somebody has organ dysfunction in the setting of an infection, you need to order blood cultures prior to giving them antibiotics.

You need to order a lactate and you need to repeat it.

That's the six hour part of the bundle, if it's greater than two.

And you need to give antibiotics as reasonably early as possible.

It's really within the first three hours.

Though, they have this caveat, ideally within the first three hours, what they hold you to is three hours, and I think that's reasonable for that patient population.

Then, if they're hypotensive or if they have a lactate greater than four, that's what triggers the 30 ccs per kilogram fluid bolus.

So if they have pneumonia and they have a new oxygen requirement and you know that's technically sepsis, that's organ dysfunction, hypoxia in the setting of an infection, that patient doesn't necessarily need 30 ccs per kilo.

Then if they become hypotensive or lower lactate comes back at four, that would trigger the 30 cc per kilo fluid bolus.

In the past few years, CMS has pushed out some updates that give us some ability to have some discretion there.

So resuscitation based on ideal body weight is acceptable for patients whose BMIs are 31 or higher, so greater than 30, and you just have to document resuscitation per ideal body weight given BMI greater than 30.

They also permit documentation of a lesser volume of fluid when accompanied by clinical reasoning.

So for example, if somebody you know has an ICD because their EF is 10%, it is reasonable and acceptable to document that, but it has to be pretty clear.

So you have to document the amount of fluid you're giving and why.

So 500 ccs of LR given instead of the 30 ccs per kilogram bolus due to concerns for congestive heart failure, EF less than 10%, something along those lines or ESRD.

But in ESRD, fluid overload, congestive heart failure, they give you some ability to document

Sam: Good.

lauren-p--black_1_08-12-2025_115919: What I will say though is I also think this pendulum has swung again with fluids in the past few years where I think people are like terrified of fluids now and don't wanna give any at all.

And I think, you know, I think we'll hopefully see where we used to just give a bajillion liters to everybody.

And I think we'll see the pendulum switch back a little bit, but that is what from a CMS criteria, they hold you to.

Within the first six hours, they ask you to remeasure that lactate if it was greater than two.

And if the patient still has a lactate greater than four, or they're still hypotensive within the first six hours they ask that you start vasopressors and reassess their volume status with like a clinical note.

That's pretty much the extent of the bundle, it hasn't changed dramatically.

Sam: So if you have someone whose say, initial lactate is high, let's say it's five for the sake of this conversation, and then I repeat it, and now it's four and a half, but they're not hypotensive.

In the CMS six hour bundle, it says you should be thinking about vasopressors at this point, but this person is not clinically hypotensive.

It's not somebody I would normally start a pressor in.

Could I have to go write something in there saying this is why I did not start vasopressors?

lauren-p--black_1_08-12-2025_115919: No uh, the vasopressor part does say just to maintain a MAP above 65.

So if they're maintaining that on their own, that's acceptable.

I also always think we should do the right thing for the patient and, just fight it on the back end.

So if they don't need pressors, they don't need pressors.

And I think we should always have our first responsibility be to the patient when those disagree with guidelines.

But no it's just vasopressors to maintain a map above 65.

Sam: Okay.

lauren-p--black_1_08-12-2025_115919: It's not terribly different than what consensus guidelines are, it's just that the lactate above four requires the fluid bolus and they call that shock, even though we don't call that shock clinically.

And I do think that's the teeniest bit confusing.

I'd love to see some more clarity around that.

Sam: Yeah.

lauren-p--black_1_08-12-2025_115919: But, you do not have to start vasopressors just for a lactate.

Sam: Yeah.

And when you say they call that shock, you mean CMS in their definition?

Calls them?

Calls anybody with a lactate greater than four, that's persisting to be in shock when that's clinically not the case for us.

lauren-p--black_1_08-12-2025_115919: Exactly.

And it's really just that they use it to trigger the septic shock bundle, which is really the fluids and then the, if they're persistently hypotensive, vasopressor requirement.

But yes, that would trigger their septic shock bundle, and those people would be captured by the septic shock abstraction methods

Sam: Gotcha.

Okay.

Let's talk about fluid type.

One conversation that people love to have.

So we're talking about some kind of fluid.

Most of us are using either normal saline or some kind of lactated ringer.

Is there actually a preference for one over the other or evidence that might swing us in one way versus the other?

lauren-p--black_1_08-12-2025_115919: So this is still quite controversial.

I will say, the last time we wrote this, SALT-ED and SMART had come out and those two studies were pragmatic, randomized, controlled trials where they switched out the fluids every month and looked at mortality and then MAKE 30 and did see a slight difference in MAKE 30, which was a composite outcome in favor of balanced crystalloids.

However, and since I think the last version of this paper was published, they then did a subgroup analysis of patients with sepsis that I think was well done.

It was pre-planned, it was clearly pre-planned in their protocol, and they found a mortality difference in favor of balanced crystalloids rather than normal saline.

So that's pretty compelling to me, even though it wasn't a randomized study in that group in particular, it's a well-planned secondary analysis of a population that makes sense to look at this in, and there's a mortality benefit in favor of LR.

So I think, in general I think LR does outperform normal saline based on the best quality of the evidence we have.

I think it makes sense physiologically.

It's no longer has a huge difference in cost.

So , in my personal practice I lean on LR.

I'm also married to a surgeon and there's nothing they hate more than normal saline.

So maybe maybe it's partially that, but for sepsis, I feel far more strongly I think this is a case where we should be using balanced crystalloids instead of normal saline.

Sam: And that subgroup analysis you talked about in sepsis patients, was that those with septic shock or just all comers sepsis.

lauren-p--black_1_08-12-2025_115919: It was their ICU cohort arm.

So they were the sicker patients.

So whether or not they were intubated , in the ICU or in septic shock, it was the sicker patients that they did show that mortality benefit in.

But mortality benefits are hard to see in sepsis, just because of the number of patients you have to enroll.

And so I feel like it's a pretty compelling result, however, it is in the sickest patients

Sam: Okay.

All right.

So lactated ringers it is.

And then there is a part of the bundle that requires a repeat assessment.

So what does that have to look like for CMS to be satisfied?

And then what should it look like for us clinically?

lauren-p--black_1_08-12-2025_115919: For CMS it is a documentation of intravascular volume status and tissue perfusion.

I think for us clinically, I think it's just going to the bedside and reassessing the patient, not the computer, and make sure they actually look better or don't look worse.

Which I think, sometimes you go in and, maybe none of the fluids are going in, maybe it's in their AC and they've got it kinked up and the fluids are largely still in the bag or something.

So I always think it's wise to go back in a few times and check on these patients that are a bit tenuous and have a high probability of decompensating.

Sam: All right.

And then let's talk about antibiotics.

So, timing of antibiotic administration, you mentioned earlier, but there is a requirement in the CMS bundle, at least some requirement for time.

And then clinically, do we have good evidence that even for a subset of populations, it might be best to just give 'em as soon as we can?

lauren-p--black_1_08-12-2025_115919: Yeah.

So I think that's a great question.

CMS will hold you to three hours.

That's still what the guidelines hold you to.

I think that gives us a fair amount of time to get some idea of a source of infection.

So I actually think that's pretty reasonable.

Sam: Can I clarify there for one second?

Is that three hours from when they arrive, or three hours from when you recognize them to have sepsis

lauren-p--black_1_08-12-2025_115919: Three hours from sepsis recognition

Sam: recognition?

Okay.

lauren-p--black_1_08-12-2025_115919: And there are some nuances here, just like on the backend with what that's called.

So you know, sometimes if a BPA goes off and you click like sepsis suspected, I think that can start your timer actually in some scenarios.

But it's the time of sepsis recognition in general.

So you have three hours from that to order antibiotics.

I will say the time to antibiotics mortality difference is in the patients with hypoperfusion or hypotension.

So in those patients, I would go ahead and start them early on because they have a clear mortality benefit.

And so if they're very sick, I would either start broad spectrum antibiotics or target to what your best idea is clinically.

But those patients, I think we should be pushing whether that means, physicians are doing lines themselves, et cetera, in order to get them antibiotics early on.

But for the rest of the patients, you have some time to determine the most appropriate antibiotic.

I will say there is some change as the surviving sepsis campaign also re-put out our one bundle, which is confusing again because it's nomenclature is fairly similar to the CMS bundle where I think the goal was to administer antibiotics earlier on if the patient has evidence of shock.

And I do think that's important.

I think the nomenclature around that is a little confusing.

I also think it perhaps misjudges the complexity of the emergency department where one hour is pretty hard.

I think that may be easier to do in environments where people already have lines, et cetera.

But if you think they're septic from the time they walk in, an hour is still sometimes a little hard to get all this stuff going.

But there is some more discussion from consensus guidelines, but again, not the CMS metrics, that are pushing for an hour.

If sepsis is definite or probable, I would personally just keep doing what we all think is best for the patient.

But I think that nomenclature is a bit confusing because it so closely mimics CMS step one, but CMS step one has not changed their guidelines off of three hours.

Sam: Okay, and that's three hours regardless of whether they're in shock or whether they're just have plain old sepsis and they're stable as can be.

lauren-p--black_1_08-12-2025_115919: Exactly.

I mean, it does say like within one hour if possible.

And I think some of that is because in those sicker patients we should be pushing for earlier initiation of antibiotics.

But even if you're pretty sure they're septic, if they don't have shock or hypoperfusion, you really do have time .

The evidence suggests you have some time to figure out the appropriate antibiotic choice.

And I think that's also beneficial to the patient.

'cause we all know the patient who's gotten like four different antibiotics what we initially thought and then it changes.

And so I think there does have to be some sort of judicious balance between antibiotic stewardship in the right patients and prompt antibiotic administration.

And I think those things are intention sometimes, and I think that's where some of the controversy is.

And I think the benefit of early antibiotics clearly outweighs the risk in the sicker patients who have low blood pressures, who even if they're not requiring vasopressors, but if they're intermittently hypotensive or they have evidence of severe hypoperfusion.

I do think we should promptly give those people antibiotics, but I think we should consider antibiotic stewardship in some of the other populations.

And we can wait two hours for the x-ray to actually come back.

Sam: Perfect.

Okay.

And then when it comes to picking the correct antibiotic, there's a great table.

It's a very large table, but table five in the article talks about the infection by type and then recommended antibiotics.

And if they're penicillin allergic, or anaphylactic, really , then some alternate choices.

So it's all laid out there for you.

Everybody likes to give rocephin just off the bat.

It stops the clock and it's what we have in the Pyxis machine and it's very easy to administer.

But there are times where you might consider giving something else.

And so having at least one or two other medications in your armamentarium is helpful.

And I think this table does a good job laying that out by source if you happen to know the source or at least suspect something specific.

lauren-p--black_1_08-12-2025_115919: Thank you.

We hope it's helpful.

We spent a fair amount of time on it.

Let's say the only big changes there in the past few years are, HCAP has kind of gone away in favor of still largely giving cap coverage to severe community acquired pneumonia, unless they have prior MRS A respiratory isolates or hospitalization and, you know, some of the other risk factors.

I don't really know clinically that our clinical practice has changed a whole lot because of that.

But it's worth knowing that there has been some changes to those definitions and those blanket recommendations.

I would also say the big thing that I do think clinically has changed is there is a whole lot more ESBL, like community acquired extended spectrum beta lactamase UTIs than there were, I feel like even five years ago.

And so if you think this suspected source is a uti, it's still so clunky to look at old cultures, I think in most EMRs, but that is one place where I really do think looking at old cultures is particularly important.

Because if they had an ESBL infection, I would not give them a beta-lactam.

I wouldn't give them rocephin or cefepime.

I would try to target one of the antibiotics to which they had shown to be sensitive in the past.

Sam: Yeah, I will say I think that adds kind of two extra steps to your decision making.

Like one, can you access an old culture of any sort?

It seems like when I am consulting my infectious disease colleagues, that's always just, the most significant thing they're doing is going and just looking at a prior culture data and then basing their best guess on that.

And second, what it is you're allowed to give in the emergency department.

A lot of us have the broadest spectrum drugs under lock and key by pharmacy.

And so sometimes it requires a little call to your ID colleague and say, Hey, their last culture showed this and I really want to give ertapenem or whatever it is, or meropenem and I need your blessing to do it.

And it seems silly, but you know, in the days of antibiotic stewardship, it's just kind of another hoop you have to jump through.

So if you have that data, that's great.

And by all means use it.

'cause yes, the resistance , is getting ridiculous and it's only getting worse as time goes on.

lauren-p--black_1_08-12-2025_115919: The only other thing I wanted to say about the table to be clear is it's for patients who are septic.

I'm not saying that these should be our first line antibiotics for patients who don't have organ dysfunction, and maybe they just have a few SIRS criteria and they're going home.

But it is for patients who are being admitted for sepsis.

Sam: good great clarifying point.

Okay, let's talk about vasopressors.

So the.

article mentions norepinephrine, dopamine, vasopressin, epinephrine what is your favorite starting agent?

Do you have one?

And tell me why.

Convince me why.

lauren-p--black_1_08-12-2025_115919: Norepinephrine should be everybody's first line vasopressor for septic shock.

I think the evidence is very strong that it is superior certainly to dopamine for septic shock and has not performed worse than vasopressin in any of the other studies.

So I think norepinephrine really should be our first line agent, and I think that's about the easiest part of septic shock and there's certainly some nuances in the research world, but as far as the overwhelming conglomeration of the evidence, the answer is levophed.

And then vasopressin should be your second additional agent.

Sam: Okay.

Then if you are debating whether or not you could give this through a peripheral line because you haven't placed a central line yet and your nursing colleagues have this tenuous line, is it okay to just go ahead and start it?

lauren-p--black_1_08-12-2025_115919: It is absolutely okay to give it peripherally and I think we should.

When we had to start a central line for everybody I think that also actually delayed some people's more appropriate shock treatment.

But evidence is very clear that through a large, fairly proximal peripheral IV, it is very reasonable to give vasopressors.

And so I personally do routinely start them peripherally.

My personal practices is, if they're on pretty high doses I still put in the central line, and this changes person to person in place to place, but if they're just requiring five of   levophed I think it's actually perfectly fine to just send them upstairs with peripheral pressors.

And I think the evidence supports that.

And I think we are seeing that slowly be more and more common.

And I think it's lovely that I think that's taken one of the cognitive hoops of the decision to start vasopressors.

I think it's perfectly fine to start  levophed peripherally and the evidence largely supports that.

Sam: And if you've only got the 22 gauge in the top of the hand.

lauren-p--black_1_08-12-2025_115919: I would put something

Sam: Alright, just gonna push you outta that one.

That's okay.

lauren-p--black_1_08-12-2025_115919: It's not that magical.

If you've got a reasonable 18 in the AC or something, that's fine.

But yeah, if you've got a 22 in the hand, you gotta get more.

In that case too, what I tell people too is it's our nursing colleagues and upstairs colleagues, if that's all we've got they need more access than that for a variety of reasons.

So, I still certainly think there's a role for a central line.

I just don't think we need to do it as often as we used to.

Sam: All right.

Let's talk about steroids.

So this pendulum has swung several times as well.

Where are we now days with steroids, and what kinds of steroids are we talking about?

lauren-p--black_1_08-12-2025_115919: So I think this is one of the biggest changes in the past few years.

I do feel like this has gone back and forth a whole lot.

Current consensus guidelines now do recommend hydrocortisone either a 50 milligram bolus every six hours or continuous infusion at 200 milligrams for patients in septic shock.

I will say as a caveat to that, in the fine print, you know, it says like if they're still requiring vasopressors after four hours and there is some nuance to that.

So I'm not saying the second you start five of  levophed, they need to be chased with like 50 milligrams of hydrocortisone.

But I do think we all know what boarding is like now.

You know, I do think if they're downstairs for a few hours, then they're still requiring  levophed, I would definitely give that to them.

The evidence for that was not based on any mortality benefit.

What it was based on was, in the meta-analysis, it showed improved time to shock resolution.

Essentially in patients who received steroids, and the consensus guidelines felt that since it's a fairly low cost, low risk intervention, the benefits of steroids outweighed the risk.

But it did not have a mortality benefit.

I will say, as a caveat to a second population, though, a pretty compelling meta-analysis of steroids in patients with severe community-acquired pneumonia did show reduced mortality and mechanical ventilation.

So in those patients, even if they're not requiring pressors if you're intubating somebody for pneumonia and sending them to the ICU, or even if you have them on high flow for pneumonia and they're not requiring vasopressors, I would still give that person some hydrocortisone as well due to the current best state and evidence.

But again this one has changed quite a few times.

As sort of a cutting edge, I think we'll see some more studies about adding fludrocortisone so adding some additional mineral corticoid coverage in addition to hydrocortisone.

But that's very much in the research world right now and not standard of care.

But for patients with persistent shock requiring levophed in the ED or patients intubated intubated high flow, et cetera, for pneumonia, I would give those patients steroids as well.

Sam: Now I'm curious, in your practice, are you doing that as just a protocol event or are you only applying that to say the ones who you know well, they're gonna be down here for a while, there's no ICU beds, I'm just gonna do this as the kind of the final thought?

Or is it just in your protocol like, we're gonna give this and if they happen to be down here, they're getting it.

If they've already gone to the ICU, then they'll just get it there.

lauren-p--black_1_08-12-2025_115919: My personal protocol is, I give it, but I'm a sepsis researcher, and I'm mostly focused on shock, so I care a lot.

I think so much is lost in handoffs regardless of best intentions.

I think it can be hard to tell what happened downstairs and what was or wasn't done.

So I personally usually do it unless there's a reason I don't think it's necessary.

Sometimes I start pressers a little bit earlier than necessary if I think they need it for a certain reason.

And in that patient who, I think they just need some, and this is like personal practice, not necessarily evidence based, but if I'm starting some  levophed to just get them through their fluid boluses, 'cause they're persistently hypertensive and I think they can get off of it once they're more fluid resuscitated, I don't give it to that patient.

But if I think the patient's gonna stay on pressers, I usually go ahead and give it.

But again, the guideline was pretty clear that, they put some four hour line on it, but this is all very consensus based and I think I agree with them the evidence, does favor shock resolution and patients who received hydrocortisone, I think there's probably people who respond really well and people who aren't responders.

And that's why we sort of see this mixed response.

And I think that's probably some of the challenge of the heterogeneity of sepsis in general.

Sam: Okay.

And then let's just lastly touch on blood transfusions.

So, within the last decade we've changed to a pretty conservative, like restrictive blood transfusion policies in most hospitals where, if your hemoglobin is not less than seven, I'm not even thinking about it.

Is that the same for my septic patients?

Or do I need to have a higher threshold for that?

lauren-p--black_1_08-12-2025_115919: No.

So the old like caring capacity days, those have also been retired.

If they don't have some obvious sign, like obvious blood loss I wouldn't give them blood

Sam: Perfect.

All right.

lauren-p--black_1_08-12-2025_115919: Unless their hemoglobin is less than seven, like the regular indications.

Sam: Excellent.

And then there are some special populations that were discussed in the paper, like the elderly, we touched on the one with cirrhosis and end stage renal disease.

Other than the fact that those with cirrhosis may have that persistent lactic acidosis and that they're risk because they're relatively immunocompromised.

What else do we need to know about our cirrhotic patients?

lauren-p--black_1_08-12-2025_115919: I think it's, still maintaining a high index of suspicion because I think sometimes some of their vital signed arrangements in a lot of those more complicated populations can be attributed to their disease state, but may actually not be normal for them.

Sam: Hmm.

lauren-p--black_1_08-12-2025_115919: And I think recognizing that some of our complicated patient populations may have more subtle signs of organ dysfunction and still having a high index of suspicion.

Sam: Yes, I can recall several cirrhotic patients who like live at a systolic blood pressure of 91 and then have like zero capacity for any kind of infection, and then will syncopize immediately.

So I'm always worried about my cirrhotic with syncope and going, well, you live at 91.

I mean, there's really not a whole lot of pressure left.

lauren-p--black_1_08-12-2025_115919: And I think, remembering the unique infections that can happen in those populations.

So I think, making sure you look for SBP in the right patients, considering in ESRD patients who are on peritoneal dialysis considering, whether or not they have SBP or, related to PD I think looking at lines for patients who have indwelling lines for dialysis and just keeping a high index of suspicion for unique sources that impact some of these patient populations.

Sam: Okay, so we're at the end of our time and we spent a lot of time talking about CMS sep one.

Where in the future are there some areas maybe for advancement or improvement in SEP one as we look towards better applying it to populations in sepsis?

lauren-p--black_1_08-12-2025_115919: I would say I think we should at least have some acknowledgement that CMS sep one has some challenges.

And I think as we see it slated to become a pay for performance measure, it's pretty complicated.

And I think there are some things that are intention.

I think we're increasingly seeing this, like one size fits all approach to sepsis being questioned, I think rightly so in the research community.

I think honestly, some of these clinical trials failed not because the interventions, I'm not the only person who thinks this, not because the interventions themselves didn't work, but we just don't know who to use them in.

There's probably some patients who respond better to some of these things and some patients who don't.

And when we study them as a whole, the effect washes out.

But there probably are people who may have some benefits to some of these things.

Same with clovers and restrictive versus liberal fluids.

There might be some patients who really do need a restrictive fluid approach.

We just don't know exactly who those are right now.

So I think that CMS sep one still retains this one size fits all approach, I think is challenging in light of the current evidence.

I think the lack of nuance regarding fluids, though I think most patients can handle some fluids.

But the lack of nuance to that definition, the tension with antibiotic administration, especially as we see some pushes from consensus guidelines to move it earlier, I think doesn't represent the individual nature of patients that well.

And I think emergency medicine providers are smart enough that we can, you know, have some individualization to our approach to the patient.

And I think there is some tension there.

And you've seen our EM societies make some statements, questioning whether or not some of these things need to be updated.

And I think it's at least worth having some of that discussion about what is the best thing for patients in light of what we know in 2025.

Sam: Thank you so much for coming on the podcast, sharing your wisdom with us, and also for being an author.

This is a fantastic article, so thank you and your other co-authors for taking the time to write it.

I think it does a tremendous job summarizing where we are in sepsis today in 2025, and I really look forward to hearing more from all of you.

And I hope you'll come back on the podcast and share your wisdom with us again in the future.

lauren-p--black_1_08-12-2025_115919: Thank you so much for your time.

It was fun.

Sam: And that's a wrap for this month's episode.

I hope you found it educational and informative.

Don't forget to go to ebmedicine.net to read the article and claim your CME.

And of course, check out all three of the journals and the multitude of resources available to you, both for emergency medicine, pediatric emergency medicine, and evidence based urgent care.

Until next time, everyone be safe.