Sepsis timer needs to die.

[sylvanthus]

This damn thing goes off every 8 seconds on everyone for random assed reasons. I spend way too much time stopping it and documenting why an elevated lactic acid is not sepsis. Someone hack epic and eliminate this annoying waste of time. Rant off.

---

Members don't see this ad.

 

[EMhawkeye]

Cerner has a similar feature. Every wbc of 12.5 and one sirs vital sign makes you have to go through a pop up window and document whether sepsis is present and a source before it lets you open the chart.

 

[RustedFox]

Meditech has no sepsis timer. I can imagine what one might be like.

I would hate that nonsense.

 

[dadaddadaBATMAN]

The medical equivalent of “please remove item from bagging area.” Ours fires 40% of the time every time, and secretly fires without a popup another 20% of the time to f up your metrics

 

[p1acebo]

This damn thing goes off every 8 seconds on everyone for random assed reasons. I spend way too much time stopping it and documenting why an elevated lactic acid is not sepsis. Someone hack epic and eliminate this annoying waste of time. Rant off.

Do you get dinged if you don’t document why it’s not sepsis?

 

[BoardingDoc]

This damn thing goes off every 8 seconds on everyone for random assed reasons. I spend way too much time stopping it and documenting why an elevated lactic acid is not sepsis. Someone hack epic and eliminate this annoying waste of time. Rant off.

Stop ordering lactates?

 

[Rekt]

Stop ordering lactates?

It's pretty shocking how many docs just dump out a lactic. Unless I'm looking for ischemic bowel or my patient is likely going to ICU or or step down I'm not getting a lactic. See so many lactics ordered on healthy young pts slightly bumped from some vomiting or whatever inappropriate nonsense then your stuck for hours waiting for the repeat to normalize

 

[Mount Asclepius]

Do you get dinged if you don’t document why it’s not sepsis?

Not to my knowledge (at least in Epic). I've been clicking 'sepsis not suspected' routinely for a long time even in septic patients that I do the entire bundle on immediately prior to clicking 'sepsis not suspected' just to get the warning to disappear.

 

[dadaddadaBATMAN]

Stop ordering lactates?

The sepsis gestapo may ding you for this as well. Mine would

It's pretty shocking how many docs just dump out a lactic. Unless I'm looking for ischemic bowel or my patient is likely going to ICU or or step down I'm not getting a lactic. See so many lactics ordered on healthy young pts slightly bumped from some vomiting or whatever inappropriate nonsense then your stuck for hours waiting for the repeat to normalize

As above. Some places require you to (it’s a scam to call a lactic 2.1 “severe sepsis” which increases complexity index and billing for the system) and I’m guessing the people you are talking about have holdover practices from this.

Also, takes a teeny bit of balls but you can just discharge the cbd hyperemesis patient with the lactic of 3.8 or admit the cirrhotic with a lactic of 17 billion to the floor (pending a hospitalist with a sense of reason). You just have to be right. When discharging I just document that they didn’t want to hang out for 3 more hours of fluids, that it’s obviously crap, and generally I call them next day. If they want to stay I don’t usually pry them out though, because it’s real awkward if you’re wrong.

And if you’re experiencing doubt when you consider that plan, I generally think the lactic was probably reasonable even if I wouldn’t have done it

 

[RustedFox]

It's pretty shocking how many docs just dump out a lactic. Unless I'm looking for ischemic bowel or my patient is likely going to ICU or or step down I'm not getting a lactic. See so many lactics ordered on healthy young pts slightly bumped from some vomiting or whatever inappropriate nonsense then your stuck for hours waiting for the repeat to normalize

Yeah, just because they give you an order set doesn't mean that you have to use every item in the order set.
I use our Sepsis Bundle but uncheck the dumb stuff all the time.

 

[EMhawkeye]

The sepsis gestapo may ding you for this as well. Mine would

As above. Some places require you to (it’s a scam to call a lactic 2.1 “severe sepsis” which increases complexity index and billing for the system) and I’m guessing the people you are talking about have holdover practices from this.

Also, takes a teeny bit of balls but you can just discharge the cbd hyperemesis patient with the lactic of 3.8 or admit the cirrhotic with a lactic of 17 billion to the floor (pending a hospitalist with a sense of reason). You just have to be right. When discharging I just document that they didn’t want to hang out for 3 more hours of fluids, that it’s obviously crap, and generally I call them next day. If they want to stay I don’t usually pry them out though, because it’s real awkward if you’re wrong.

And if you’re experiencing doubt when you consider that plan, I generally think the lactic was probably reasonable even if I wouldn’t have done it

You ever notice the sepsis gestapo are never physicians and can't intelligently speak about the pathophysiology of sepsis?

 

[RustedFox]

You ever notice the sepsis gestapo are never physicians and can't intelligently speak about the pathophysiology of sepsis?

I was asked to step down from the sepsis committee at my first job because I "wasn't a champion of sepsis".

I pointed out that I was the only physician in the room, the only one actually making decisions regarding patient care, and the only one who could demonstrate an understanding of medical literature.

No joke. This happened.

 

[southerndoc]

From the official sepsis police, simply documenting "no sign of severe sepsis" will eliminate the need for a lot of unnecessary cultures, lactate, etc. Obviously you can't just use it to avoid SEP-1 measures, but we over test for sepsis.

 

[Mount Asclepius]

Agree. Many (including physicians) forget the basic premise that you have to have BOTH an infection AND end organ damage. I see too many patients incorrectly called septic with just infection + SIRS, or with end organ damage from a variety of causes without an infection. Just chart no bacterial infectious process suspected or identified and you are in the clear of the timer and the gestapo.

 

[sylvanthus]

Do you get dinged if you don’t document why it’s not sepsis?

Emails inevitably ensue.

 

[sylvanthus]

Stop ordering lactates?

Not just the lactates is the problem. The algorithm includes a bunch of nonsense and the timer will just start automatically regardless if lactic is ordered or resulted. wbc elevated? HR elevated? etc etc etc

 

[thegenius]

It's pretty shocking how many docs just dump out a lactic. Unless I'm looking for ischemic bowel or my patient is likely going to ICU or or step down I'm not getting a lactic. See so many lactics ordered on healthy young pts slightly bumped from some vomiting or whatever inappropriate nonsense then your stuck for hours waiting for the repeat to normalize

Agree. Although it is hard to admit a patient to the hospital for most infections without a lactate and blood cultures, despite how useless they appear to be except for the very sick

 

[CoolDoc1729]

See so many lactics ordered on healthy young pts slightly bumped from some vomiting or whatever inappropriate nonsense then your stuck for hours waiting for the repeat to normalize

One of my coworkers puts CC on every one of these lmao

 

[emergentmd]

I do not miss working in the hospital. The countless time I just ignored the sepsis bell, then the trauma bell, then the stroke bell, then the CP bell. Holy cow, I am glad docs can actually make decisions on how to run an ER. I have not followed a protocol in 3 years and my pts all seem to do just fine.

Maybe all these bells allows the hospital to increase billing so maybe we should be doing this.

 

[bravotwozero]

One of my coworkers puts CC on every one of these lmao

If you’re going to do more unnecessary work might as well score a few rvus…

 

[pkwraith]

An issue with the sepsis detection algorhthm in Epic is that the sensitivity is often turned too up in the ED. For those who not familiar with it, it is not based on SIRS. It is based on a computer multi variable regression model that plugs in hundreds of discrete data and associates it with patient that ended up with an ICD code of sepsis. So the sepsis timer is based on things (that includes vitals), but also on if a patient is getting IV antibiotics, how many IVs they have, their past medical history, labs such as lactic acid but also absolute monocyte count.

A patient that a history of cirrhosis will automatically get a high sepsis score (just by existing), and all they need in many cases is an IV and and abnormal vital and it may meet your department threshhold of sepsis score.

In fact it's easy to trigger the sepsis score after you have already identified sepsis, because by ordering fluids and IV antibiotics, you have increased their sepsis score. Which is why many times a septic patient may not trigger sepsis, but after you have identified and treated it, the sepsis alert will activate.

In theory it is actually very interesting, but in practice, has a lot of flaws. There is a UM paper on the lack of accuracy of the model in practice.

 

[RustedFox]

An issue with the sepsis detection algorhthm in Epic is that the sensitivity is often turned too up in the ED. For those who not familiar with it, it is not based on SIRS. It is based on a computer linear regression algorithm that plugs in hundreds of discrete data and associates it with patient that ended up with an ICD code of sepsis. So the sepsis timer is based on things (that includes vitals), but also on if a patient is getting IV antibiotics, how many IVs they have, their past medical history, labs such as lactic acid but also absolute monocyte count.

A patient that a history of cirrhosis will automatically get a high sepsis score (just by existing), and all they need in many cases is an IV and and abnormal vital and it may meet your department threshhold of sepsis score.

In fact it's easy to trigger the sepsis score after you have already identified sepsis, because by ordering fluids and IV antibiotics, you have increased their sepsis score. Which is why many times a septic patient may not trigger sepsis, but after you have identified and treated it, the sepsis alert will activate.

Oh dear God.
It's worse than I ever thought.

 

[Dr Mantis Toboggan]

Oh dear God.
It's worse than I ever thought.

Soon it will become self aware.

 

[dadaddadaBATMAN]

Soon it will become self aware.

But apparently it’s an idiot. If we let it vote it will blend right into the American mainstream.

 

[thegenius]

In theory it is actually very interesting, but in practice, has a lot of flaws. There is a UM paper on the lack of accuracy of the model in practice.

Give us the link!

 

[pkwraith]

Give us the link!

External Validation of a Widely Implemented Sepsis Prediction Model in Hospitalized Patients

​

Basically, the current model poorly predicts sepsis and created large burden of alarm fatigue. For example it generated alarm in 18% of all hospitalized patient but missed identification of 67% of patient with sepsis.

 

[RustedFox]

External Validation of a Widely Implemented Sepsis Prediction Model in Hospitalized Patients

​

Basically, the current model poorly predicts sepsis and created large burden of alarm fatigue. For example it generated alarm in 18% of all hospitalized patient but missed identification of 67% of patient with sepsis.

But that won't stop admins from looking at the title of the paper only and saying: "Look! Its VaLiDaTeD!"

 

[turkeyjerky]

Holy Jeebus that's bad. Is the Epic one like Cerner, where it automatically adds a diagnosis of 'sepsis' to the patient list if it fires? And thus auto-validates it's own model using an incorrect outcome measure each time, in effect forever becoming less accurate?

 

[pkwraith]

And for further clarification, I'm not actually anti Epic or anti predictive models. In fact, I'm the opposite. I think there is a very real use for EMR, machine learning, and it will eventually reach a point where it can be a real asset. In fact, I think the Epic predictive model is a good (first) step, taking large amount of available data and trying to make sense of it in a practical sense. And right now it is a proof of concept that has a lot of flaws, and a lot is limited by how current EMR data structure, as the quirks of statistical models And when it is used (like any other tools) incorrectly such as for compliance that a lot of the flaws are evident, and may of the flaws should be fixed.

The way a multi regression database such as the Epic sepsis predictive model is that sometimes you end up with variables that are self fulfilling or sometimes spurious.
For example, the computer model identified that "sepsis" patients have high association with IV cephalosphorin orders and receive a lot of IV fluids. Therefore, IV cephalosporin and number of IV fluid orders must predict a patient that will be septic. When in actually, these patients have been identified as septic by nature of already receiving antibiotics and fluids. Warning someone of something that had already happened is not helpful, and is a failure of a predictive model. This is also somethign the JAMA article points out, if you identify sepsis after sepsis has already been identified, it should count as as a Miss in a predictive model.

Another example is when you churn through thousands of variables blindly, it is entire possible to find some spurious variable just out of random chance. And you end up with a random lab value that has never been studied (or even proposed) to be related to sepsis. I can't remember right now if it is the RDW or the MCHC (from the diff) that is part of the sepsis score. It does not get a big weight, but it is nonetheless part of what determines if a patient gets flagged as sepsis. I am fairly certain no one here has ever looked at RDW or the MCHC to determine if a patient is sepsis or not.

The end result is a tool that is being used for and treated as something more than it is in the current metric-obssesed world. Although, if you feel it is firing too much, your department does have the ability to turn down the threshold that it fires off.

 

[xaelia]

Machine learning is magic, and the folks popping up little unicorn startups utilizing such dark magic don't want you to question it.

Fascinating work being done to examine the issues with ML models – including such interesting findings as CXR imaging ML being able to identify ethnicity from even luminance-normalised radiographs, or "healthcare support outreach" models being systematically biased against low-SES patients because they don't have any healthcare dollars spent on them.

 

[RustedFox]

Machine learning is magic, and the folks popping up little unicorn startups utilizing such dark magic don't want you to question it.

Fascinating work being done to examine the issues with ML models – including such interesting findings as CXR imaging ML being able to identify ethnicity from even luminance-normalised radiographs, or "healthcare support outreach" models being systematically biased against low-SES patients because they don't have any healthcare dollars spent on them.

Maybe we need DEI for machines to go to machine learning schools. Wait.... its not supposed to matter in the first place.

 

[thegenius]

Maybe we need DEI for machines to go to machine learning schools. Wait.... its not supposed to matter in the first place.

I think we need Battle Bot Wars between a DEI machine and a Machine Learning machine.

On the Discovery Channel, Tue at 10:30 PM PST.

 

[Janders]

Why do they care so hard about SEP-1? It is NOT pay for performance. Margins are tight, why do they waste cash paying someone to be the sepsis czar. Is it purely to enable upcoding under the guise of “quality”?

 

[southerndoc]

Why do they care so hard about SEP-1? It is NOT pay for performance. Margins are tight, why do they waste cash paying someone to be the sepsis czar. Is it purely to enable upcoding under the guise of “quality”?

SEP-1 is concentrated on for several reasons:

1. It's a publicly reported measure.
2. Although not tied to CMS reimbursement (yet), some private insurers will deny sepsis DRGs if SEP-1 compliance was not met on an individual case.
3. Some private insurers will reduce all sepsis DRG payments by a certain percentage if your SEP-1 compliance is low (as opposed to #2, denying on a case-by-case basis).
4. CMS has proposed that it be tied to reimbursement. This has been in the works for a while now and likely will take effect in 2024.
5. Adherence to SEP-1 measures (primarily antibiotics in a timely manner) has been associated with decreased mortality.

Most people think SEP-1 equals requirement to give IV fluids. A 30 mL/kg fluid bolus is ONLY required for a lactate >4 OR hypotension.

If a patient is hypotensive for other reasons (GI bleed, dehydration, etc.) or if the lactate is elevated from a non-sepsis reason, then simply document it.

The problem comes with abstractors being unable to figure out what you were thinking. You have to clearly document it.

The hypotensive GI bleed with a UTI doesn't need cultures and a fluid bolus. Simply document "hypotension due to anemia and not due to sepsis/infection" and it excludes it. If I were reviewing your chart, I would understand that. An abstractor isn't always medically trained and if they are, they still aren't allowed to connect the dots.

 

[Mount Asclepius]

Most people think SEP-1 equals requirement to give IV fluids. A 30 mL/kg fluid bolus is ONLY required for a lactate >4 OR hypotension.

As of earlier this year there are also now multiple reasons you can use to not give 30 ml/kg IV fluids as long as you document that reason and the alternative amount given instead. I found previously that the majority of physicians were primarily frustrated with SEP-1 because they felt like they had to give 30 ml/kg IV fluids to patients with septic shock that they felt were already fluid overloaded.

 

[hundreddaysoff]

I suspect that regardless of the theoretical medical benefits of these ML models or the actual underlying science, things will trend toward whichever medical decisions make hospitals the most money, until/if another big non-healthcare stakeholder like the Justice Dept or whatever figures out what's going on and intervenes.

Dig: The organizations that make more money by selling more saltwater and antibiotics (for example) will use their money to get bigger and bigger. They will in time make enough money to ditch their cheap little Meditech from 1985 and buy Epic, which will allow them to fire more of these BS sepsis triggers, thus (via subconscious repetition both electronic and verbal) warping the common knowledge of terms like "severe sepsis", and thus resulting in more and more saltwater and antibiotics getting sold.

The biggest ones, like HCA, will become big enough to, by subtle and ingenius means, effectively create public policies that let them sell more saltwater and antibiotics. Eg, One-Hour Sepsis, which is now a protocol recommended by the insurer of my current little rural hospital, which has nothing to do with HCA. (Although One-Hour Sepsis is not, amazingly to my knowledge, yet recommended by my CMG USACS! At least, not until the equity to debt conversion cometh...)

The organizations that do not go this way, either because they only hire ethical and sensible docs or they have ethical and sensible policies, will not make as much money. They will also go against the grain of the big insurance and CMS and other bureaucracies, whose marching orders will be more and more created by... you guessed it... the big saltwater+antibiotics sellers.

Ultimately, those "sensible" organizations will die, so in retrospect they were not so sensible. And so the feedback will continue. Tail, dog, wag. Rich get richer, money goes to money.

I hope I am unrealistically cynical and naive about all this and someone will correct me that this is not how our medical system really operates these days.

Source: I'm pretty good at operating small-scale flywheels. I fed myself by writing and publishing scientific papers for 10 years before and during med school. In fact, by far the most important consequence of my publishing all these papers was that I was able to craft a clutch narrative that got me into med school. All expenses paid, in fact!

 

[Mount Asclepius]

I suspect that regardless of the theoretical medical benefits of these ML models or the actual underlying science, things will trend toward whichever medical decisions make hospitals the most money, until/if another big non-healthcare stakeholder like the Justice Dept or whatever figures out what's going on and intervenes.

Dig: The organizations that make more money by selling more saltwater and antibiotics (for example) will use their money to get bigger and bigger. They will in time make enough money to ditch their cheap little Meditech from 1985 and buy Epic, which will allow them to fire more of these BS sepsis triggers, thus (via subconscious repetition both electronic and verbal) warping the common knowledge of terms like "severe sepsis", and thus resulting in more and more saltwater and antibiotics getting sold.

The biggest ones, like HCA, will become big enough to, by subtle and ingenius means, effectively create public policies that let them sell more saltwater and antibiotics. Eg, One-Hour Sepsis, which is now a protocol recommended by the insurer of my current little rural hospital, which has nothing to do with HCA. (Although One-Hour Sepsis is not, amazingly to my knowledge, yet recommended by my CMG USACS! At least, not until the equity to debt conversion cometh...)

The organizations that do not go this way, either because they only hire ethical and sensible docs or they have ethical and sensible policies, will not make as much money. They will also go against the grain of the big insurance and CMS and other bureaucracies, whose marching orders will be more and more created by... you guessed it... the big saltwater+antibiotics sellers.

Ultimately, those "sensible" organizations will die, so in retrospect they were not so sensible. And so the feedback will continue. Tail, dog, wag. Rich get richer, money goes to money.

I hope I am unrealistically cynical and naive about all this and someone will correct me that this is not how our medical system really operates these days.

Source: I'm pretty good at operating small-scale flywheels. I fed myself by writing and publishing scientific papers for 10 years before and during med school. In fact, by far the most important consequence of my publishing all these papers was that I was able to craft a clutch narrative that got me into med school. All expenses paid, in fact!

You’re aren’t wrong to ‘follow the money.’ However, the one hour sepsis bundle isn’t entirely wrong either. Surprise, you give antibiotics faster to someone with a bacterial infection, they may be more likely to do better. Saltwater and pressors may also help. Sure, CVPs, cultures and bundles may be more money than medical benefit, but there is something in the muck.

 

[3rdJob]

Machine learning is magic, and the folks popping up little unicorn startups utilizing such dark magic don't want you to question it.

Fascinating work being done to examine the issues with ML models – including such interesting findings as CXR imaging ML being able to identify ethnicity from even luminance-normalised radiographs, or "healthcare support outreach" models being systematically biased against low-SES patients because they don't have any healthcare dollars spent on them.

Folks,

So in one of my other jobs, I work in (with co founders) a medical device startup (never do this). The algorithm that interprets data from the device uses a neural-network technique; neural networks are a subset of Machine Learning. Having now been in the ML space for about a decade now I can say that there are mutiple pitfalls, rabbit holes, errors and misinterpretations that can occur using these techniques. One can really fool oneself as to the validity of the results. Badly. Been there, done that; wrote the book (well, not really...).
Although there are good techniques to help one stay out of the rabbit holes; I suspect many of these startups probably don't know and/or use them as they can be time/$$ intensive. Or conversely, they don't understand our Emergency Medicine space very well.

Interpret results from ML devised health-care algorithms with great caution......just my $0.04

 

[RustedFox]

Well, I can put the EKG leads on a patient, take 3 EKGs each one minute apart, and get 3 different computer interpretations (the morphology will remain the same, for you pedantic ones out there), so I'm not surprised?

 

[hundreddaysoff]

You’re aren’t wrong to ‘follow the money.’ However, the one hour sepsis bundle isn’t entirely wrong either. Surprise, you give antibiotics faster to someone with a bacterial infection, they may be more likely to do better. Saltwater and pressors may also help. Sure, CVPs, cultures and bundles may be more money than medical benefit, but there is something in the muck.

Not denying that for legit bacterial sepsis with a lower-case s, IVF+abx +/- Solumedrol +/- pressors are essential. But the way ¡Sepsis! (TM) is defined, legit bacterial sepsis is a needle in a haystack of all the things I am supposed to push these drugs for.

The question for me has always been how much overall harm we're doing with all the false negatives, both with first-order medical adverse effects and also with billing patients for stuff they never needed and can't afford. And the trend has been to take this decisionmaking away from doctors and turn the feds and the wolves (ie, the people who sell the products we push) into nudgers. Thus increasing the likelihood in the real world that I as a doctor will do harm.

Last I checked ~2020, there was no literature validating 1-hour sepsis over the standard 3-hour SEP-1 bundle. Does convincing literature now exist?

It scares the willies out of me if it's true that CMS will start punishing for sepsis "fallouts" in 2024. (As opposed to just punishing for failure to document all the SEP-1 stuff, as I understand they do now.) Because, if all their weaselly definitions stay the same, this increases the incentives of the bad actors (eg HCA) to act poorly to make more money. But at the same time, if CMS's apparent incentive to find ¡Sepsis! (TM) everywhere does not change, then CMS still has no obvious incentive to, eg, punish HCA nudgers/docs/whoever does the old trick of pushing saltwater+antibiotics in a 19-year old with Strep throat whose numbers happen to be randomly crunked.

Ie, without a responsible adult intelligently rethinking sepsisology, this CMS policy change would just put things even more out of balance in favor of more bigger pointlesser workups.

 

[Mount Asclepius]

Not denying that for legit bacterial sepsis with a lower-case s, IVF+abx +/- Solumedrol +/- pressors are essential. But the way ¡Sepsis! (TM) is defined, legit bacterial sepsis is a needle in a haystack of all the things I am supposed to push these drugs for.

The question for me has always been how much overall harm we're doing with all the false negatives, both with first-order medical adverse effects and also with billing patients for stuff they never needed and can't afford. And the trend has been to take this decisionmaking away from doctors and turn the feds and the wolves (ie, the people who sell the products we push) into nudgers. Thus increasing the likelihood in the real world that I as a doctor will do harm.

Last I checked ~2020, there was no literature validating 1-hour sepsis over the standard 3-hour SEP-1 bundle. Does convincing literature now exist?

It scares the willies out of me if it's true that CMS will start punishing for sepsis "fallouts" in 2024. (As opposed to just punishing for failure to document all the SEP-1 stuff, as I understand they do now.) Because, if all their weaselly definitions stay the same, this increases the incentives of the bad actors (eg HCA) to act poorly to make more money. But at the same time, if CMS's apparent incentive to find ¡Sepsis! (TM) everywhere does not change, then CMS still has no obvious incentive to, eg, punish HCA nudgers/docs/whoever does the old trick of pushing saltwater+antibiotics in a 19-year old with Strep throat whose numbers happen to be randomly crunked.

Ie, without a responsible adult intelligently rethinking sepsisology, this CMS policy change would just put things even more out of balance in favor of more bigger pointlesser workups.

You make some valid points. However, I think like a lot of physicians still miss a key basic tenet. You have to have a patient with a bacterial infection, SIRS criteria and end organ damage who is getting admitted to fall into the sepsis metric. Most 19 year olds with strep throat are going home. The metric doesn’t apply. The ones with SIRS criteria often don’t have end organ damage. Even if they have a lactate of 2.3, I’d bet most still get discharged home. The ones that meet all of the criteria are probably sick enough that they should be receiving antibiotics quickly.

You usually know within 1 hour. Sure, there are often system challenges that prevent that from being accomplished. There isn’t much you can do about that other than getting involved at the local level to try to fix those challenges. I think that’s the bigger focus of the sepsis police. It’s usually not trying to micromanage physician autonomy.

The key is to understand the rules to the simple game. Then it’s easy enough to meet the metric when appropriate, and document correctly when not so that you don’t have to deliver inappropriate care.

 

[hundreddaysoff]

You make some valid points. However, I think like a lot of physicians still miss a key basic tenet. You have to have a patient with a bacterial infection, SIRS criteria and end organ damage who is getting admitted to fall into the sepsis metric. Most 19 year olds with strep throat are going home. The metric doesn’t apply. The ones with SIRS criteria often don’t have end organ damage. Even if they have a lactate of 2.3, I’d bet most still get discharged home. The ones that meet all of the criteria are probably sick enough that they should be receiving antibiotics quickly.

You usually know within 1 hour. Sure, there are often system challenges that prevent that from being accomplished. There isn’t much you can do about that other than getting involved at the local level to try to fix those challenges. I think that’s the bigger focus of the sepsis police. It’s usually not trying to micromanage physician autonomy.

The key is to understand the rules to the simple game. Then it’s easy enough to meet the metric when appropriate, and document correctly when not so that you don’t have to deliver inappropriate care.

Sorry, I should've given more background on the actual problem. I haven't slept in a while.

I know all these excellent things you are saying and agree with them and chart to them. I personally don't run into trouble with these metrics. In fact I have been invited to be a Champion Of Sepsis by (minor) Powers That Be but refused because I value my sanity.

The problem is that many of the clinicians I've worked with don't know, or say they know when I tell them but then smile and nod and keep doing what they've been doing, in particular NPs and PAs. Many of them believe, eg based on simplistic emails from my medical directors, that they do need to do the sepsis stuff on someone with (eg) a bacterial infection who does not have end-organ damage and in any case is going home. The more stuff these clinicians order, the safer they feel and the more job security they think they have. At my HCA shop, the scariest thing is that they were probably right. At my latest USACS shop, it turned out not to matter because USACS just laid off all the midlevels anyway when we hit the COVID lull and never hired them back.

I educate my RNs about sepsisology night after night. Different set of travelers, same willful ignorance and needless OMG SEPSIS!!! LARPing. I educate the good ones, and then they leave, and then a new batch comes in who does all the same naive Sepsis (TM) stuff.

Why should I listen to what my RNs tell me about how to treat patients? I generally don't, and I usually don't run into problems because I've designed my schtick around avoiding actual harm from all this nonsense. But there are often triage protocol orders designed by RN admins that I can't change. I fight to override them, but sometimes it's just too busy, and more importantly many docs I know just give in and take the path of least resistance.

And, patients hear nurses talking about what "should" happen and overhear the word "sepsis", and then they tell their friends and family they were septic, and then patients get some unhealthy expectations as well. It's all like a big game of Whack-A-Sepsis.

What I'm saying is that there is a culture of ignorance, groupthink, and fear about Sepsis (TM) in most of the ERs where I've worked. This works to the benefit of those who stand to profit from overworking patients. Like insurance companies who purposely design protocols to grind down doctors who request appropriate but expensive patient care, I suspect that the powers that be at these places know exactly what they're doing with the protocols and the culture. They nudge us and enrich themselves.

I am changing jobs in a couple months to a place that looks more stable and less traveler-y on paper and more stable personnel wise. It is not owned by HCA or any other behemoth. I am an optimist and I hope things will be better there. But they were pretty similar at my last 2 shops.

TL;DR I'm not discussing first-order, ie what's right or wrong patient care. I'm discussing Common Knowledge among ER staff and patients, and expectations, and how stakeholders have consciously or unconsciously shaped this common knowledge to benefit themselves.

I dunno, maybe I'm just crazy. Thanks for making me clarify from my original rant

 

[RustedFox]

You make some valid points. However, I think like a lot of physicians still miss a key basic tenet. You have to have a patient with a bacterial infection, SIRS criteria and end organ damage who is getting admitted to fall into the sepsis metric. Most 19 year olds with strep throat are going home. The metric doesn’t apply. The ones with SIRS criteria often don’t have end organ damage. Even if they have a lactate of 2.3, I’d bet most still get discharged home. The ones that meet all of the criteria are probably sick enough that they should be receiving antibiotics quickly.

You usually know within 1 hour. Sure, there are often system challenges that prevent that from being accomplished. There isn’t much you can do about that other than getting involved at the local level to try to fix those challenges. I think that’s the bigger focus of the sepsis police. It’s usually not trying to micromanage physician autonomy.

The key is to understand the rules to the simple game. Then it’s easy enough to meet the metric when appropriate, and document correctly when not so that you don’t have to deliver inappropriate care.

So, you're missing one thing here (and this is the thing that really makes me mad)
Any abnormal vital sign + a bacterial infection later on discovered also counts as "sepsis present on admission" and will hose you.

I had a sepsis "fallout" for acute MI. How?

1. HR = 96 at triage.
2. EKG is a STEMI.
3. Gone to cath lab. No abx given.
4. Foley'd at some point.
5. UTI

IM ostrich writes *sepsis present upon admission*.

I'm hosed.

 

[Dr Mantis Toboggan]

So, you're missing one thing here (and this is the thing that really makes me mad)
Any abnormal vital sign + a bacterial infection later on discovered also counts as "sepsis present on admission" and will hose you.

I had a sepsis "fallout" for acute MI. How?

1. HR = 96 at triage.
2. EKG is a STEMI.
3. Gone to cath lab. No abx given.
4. Foley'd at some point.
5. UTI

IM ostrich writes *sepsis present upon admission*.

I'm hosed.

Major BS.

 

[CoolDoc1729]

So, you're missing one thing here (and this is the thing that really makes me mad)
Any abnormal vital sign + a bacterial infection later on discovered also counts as "sepsis present on admission" and will hose you.

I had a sepsis "fallout" for acute MI. How?

1. HR = 96 at triage.
2. EKG is a STEMI.
3. Gone to cath lab. No abx given.
4. Foley'd at some point.
5. UTI

IM ostrich writes *sepsis present upon admission*.

I'm hosed.

The stemi was from a septic embolus! /s

 

[Arcan57]

So, you're missing one thing here (and this is the thing that really makes me mad)
Any abnormal vital sign + a bacterial infection later on discovered also counts as "sepsis present on admission" and will hose you.

I had a sepsis "fallout" for acute MI. How?

1. HR = 96 at triage.
2. EKG is a STEMI.
3. Gone to cath lab. No abx given.
4. Foley'd at some point.
5. UTI

IM ostrich writes *sepsis present upon admission*.

I'm hosed.

Of course if IM doesn’t put that down then they get dinged for a never event (CAUTI). I’m working in a system now where we routinely fail the SEP-1 measure bundle. On review a third of the patients we fail on flag for opening a deceased pt’s chart when we go back to review the case. I feel like sepsis is still stuck in the same place it was during the EGDT days. Devoting attention and resources to sepsis improves outcomes compared to not doing so, but we don’t have a good handle on which resources and how much attention actually matters. Furthermore, the people involved in the conversation are not agnostic to the answer since their jobs largely depend on having a complicated enough algorithm that it justifies their FTEs.

 

[Mount Asclepius]

So, you're missing one thing here (and this is the thing that really makes me mad)
Any abnormal vital sign + a bacterial infection later on discovered also counts as "sepsis present on admission" and will hose you.

I had a sepsis "fallout" for acute MI. How?

1. HR = 96 at triage.
2. EKG is a STEMI.
3. Gone to cath lab. No abx given.
4. Foley'd at some point.
5. UTI

IM ostrich writes *sepsis present upon admission*.

I'm hosed.

Nope, not missing anything (other than perhaps your hospital/system needs your help in herding the ostriches, but sometimes it's easier to just stick your head in the sand).

#1 How did you get a urine prior to cath lab?!

#2 Did they really have a UTI, or just asymptomatic bacteruria? (rhetorical, as here is potentially the first bad documentation by IM).

#3 That patient wasn't septic per information provide. Need infection, 2 SIRS criteria and end organ damage. This is poor documentation by IM.

#4 There is grey area in abstraction. In our system, 'sepsis on admission' means earlier between time of admit order to the hospital or H&P time not 'ED arrival' time. Some systems abstract that differently and count 'ED admission.'

#5 It's all in the documentation and abstraction. Your group should meet with the sepsis folks and inpatient physicians to improve. You also shouldn't be dinged for a fallout or hosed based upon this case. This is where your group or medical director needs to defend you without you ever really even hearing about this being a 'fallout.'

 

[RustedFox]

I didn't get a urine.
Someone shortly thereafter slapped a Foley in Grandma and -surprise- called it a UTI.

I'll write more later; but it's easy to see how this could happen. In similar fashion, someone coming in under the auspices of stroke alert that has abnormal vitals and ends up on a vent also can have IM slap "sepsis present on admission" on the chart.

 

[sylvanthus]

Antibiotics resistance is gonna be a bitch in the future. Noones tracking how much damage overtreating for sepsis is doing.

 

[VA Hopeful Dr]

Antibiotics resistance is gonna be a bitch in the future. Noones tracking how much damage overtreating for sepsis is doing.

I don't think its the sepsis stuff that's going to cause that. Its us mildly burnt out PCPs giving it away to every "I've had the sniffles for 3 hours fix me" patient that walks through our doors.