Using high sensitivity troponin in chest pain dispo

[SpacemanSpifff]

The CMS thread from @DocEspana sparked some discussion surrounding chest pain disposition and how to incorporate high sensitivity troponin assays and clinical decision tools into that framework. I have encountered three different pathways at three different shops. Some of the cardiologists I've talked to don't really seem to know how to deal with the new troponin assays. Reading some of the discussion on the other thread, this seems to be an area of significant practice pattern variation. I have a some questions for y'all:

1. If you have high sensitivity troponin in your shop, what is your department approved pathway guiding chest pain disposition for patients with suspicion of ACS?
2. Do you follow the ACEP clinical practice guideline to a t (i.e. if HEART >3, the patient can never be "low risk"?)
3. What situations will you accept a single troponin as sufficient evidence that chest pain is low risk?
4. How easy is it to routinely admit non-low risk patients to the hospitalist for further eval in the absence of clear ECG abnormalities or "elevated" troponin? Assuming you don't have an observation unit to punt these folks into.

---

Members don't see this ad.

 

[RustedFox]

The CMS thread from @DocEspana sparked some discussion surrounding chest pain disposition and how to incorporate high sensitivity troponin assays and clinical decision tools into that framework. I have encountered three different pathways at three different shops. Some of the cardiologists I've talked to don't really seem to know how to deal with the new troponin assays. Reading some of the discussion on the other thread, this seems to be an area of significant practice pattern variation. I have a some questions for y'all:

1. If you have high sensitivity troponin in your shop, what is your department approved pathway guiding chest pain disposition for patients with suspicion of ACS?
2. Do you follow the ACEP clinical practice guideline to a t (i.e. if HEART >3, the patient can never be "low risk"?)
3. What situations will you accept a single troponin as sufficient evidence that chest pain is low risk?
4. How easy is it to routinely admit non-low risk patients to the hospitalist for further eval in the absence of clear ECG abnormalities or "elevated" troponin? Assuming you don't have an observation unit to punt these folks into.

I'll answers your questions four tomorrow, but regarding the bolded sentence: the cardiologists at my shop have "given up" and order their own, oldschool TnI assay, essentially saying "Eff this hsTNI nonsense."

I hate the new hsTNI stuff, too. They really do seem "meaningless" to me.

 

[WildEMDoc]

1. Depending on which site I'm working at, either there is this complex flow chart that people use to sort out the delta level (university site), or 2 negative levels = home +/- outpatient cards f/u if feeling good or admit if suspicion is high (community sites).
2. none of my sites really use HEART anymore because the cards docs don't think its useful and I generally agree; gestalt and experience and the context of the patient is so much more useful than that score especially when at least half of your undifferentiated chest pain patients have CAD
3. HS-Trop <6 with a low pretest probability of ACS is all I need to d/c on one trop.
4. Since we call so infrequently at my community sites (most go home with outpatient) if the ED doc wants someone observed its typically pretty easy to admit and normally its due to ongoing pain ie unstable angina.

 

[Angry Birds]

1. Depending on which site I'm working at, either there is this complex flow chart that people use to sort out the delta level (university site), or 2 negative levels = home +/- outpatient cards f/u if feeling good or admit if suspicion is high (community sites).
2. none of my sites really use HEART anymore because the cards docs don't think its useful and I generally agree; gestalt and experience and the context of the patient is so much more useful than that score especially when at least half of your undifferentiated chest pain patients have CAD
3. HS-Trop <6 with a low pretest probability of ACS is all I need to d/c on one trop.
4. Since we call so infrequently at my community sites (most go home with outpatient) if the ED doc wants someone observed its typically pretty easy to admit and normally its due to ongoing pain ie unstable angina.

HS-Trop < 6 - what does the <6 mean?

 

[RustedFox]

HS-Trop < 6 - what does the <6 mean?

???

A number less than 6.

 

[xaelia]

???

A number less than 6.

<3

 

[Franzd'Epinay]

???

A number less than 6.

Big, if true.

 

[Angry Birds]

Ok I was thinking different units. .06 at my shop

 

[southerndoc]

Ok I was thinking different units. .06 at my shop

That's a traditional troponin.

 

[Groove]

The CMS thread from @DocEspana sparked some discussion surrounding chest pain disposition and how to incorporate high sensitivity troponin assays and clinical decision tools into that framework. I have encountered three different pathways at three different shops. Some of the cardiologists I've talked to don't really seem to know how to deal with the new troponin assays. Reading some of the discussion on the other thread, this seems to be an area of significant practice pattern variation. I have a some questions for y'all:

1. If you have high sensitivity troponin in your shop, what is your department approved pathway guiding chest pain disposition for patients with suspicion of ACS?
2. Do you follow the ACEP clinical practice guideline to a t (i.e. if HEART >3, the patient can never be "low risk"?)
3. What situations will you accept a single troponin as sufficient evidence that chest pain is low risk?
4. How easy is it to routinely admit non-low risk patients to the hospitalist for further eval in the absence of clear ECG abnormalities or "elevated" troponin? Assuming you don't have an observation unit to punt these folks into.

1) We have converted to HsTnT. The dept approved pathway is 2 trop, 90 mins apart, 2 EKGs, HEART score with traditional disposition based on score.

2) I don't think HEART is perfect, but I do use it and document it on any ACS rule out. That being said, I don't always rely on it. My entire chest pain work ups and dispositions are almost entirely gestalt based. I'm also one of the few in my group that almost never gets a 2nd trop/EKG. I'm usually prepared to disposition my patient after the initial trop. I think holding these people in the ED for hours is ridiculous. Either admit them or don't. This bastardized amalgamation of HEART and ADAPT kills metrics and overloads WR.

3) Atypical features, pain that's been over 4-6 hours, etc.. (See above, I almost always rely on a single trop.)

4) Very easy. That being said, I do a lot of informed decision making with the pt. Sometimes, bringing them in is reasonable for a stress test but I find people just don't want to be in the hospital. I discuss risks/benefits and let them make their own decision as to whether they want to be in the hospital or f/u outpatient and document appropriately. Luckily, we have a really good outpatient system where I can send off a message with pt linked and request outpatient stress and they will set it all up within 3-5 days as well as cardiology appt.

 

[Angry Birds]

That's a traditional troponin.

Hmmm I’ll be damned

 

[RoyBasch]

The utility of the repeat troponin in my mind is based entirely on the onset and course of the chest pain. If the chest pain has been constant for more than 6 hours, I do not feel the second troponin is going to tell you much the first one didn't.

If the pain started 30 mins ago, repeat troponin even in low risk chest pain is necessary.

If the pain is intermittent, I go from the start time of their current pain. I.e. if patient has "chest pain" for 3 days but they really mean they have had 4-5 episodes of pain lasting 30 minutes each over the last 3 days and now in the ER they have an episode of pain that started 30 minutes ago, I would consider 30 minutes ago--not 3 days ago--the "start time" with regards to the current troponin's sensitivity to evaluate ischemia.

In the event of a true ACS/MI: troponin rises with time. The rate of rise is going to be influenced by the severity of the event and amount of tissue infarcted, but its not instantaneous. We have all seen STEMIs with negative troponin drawn 15 mins after the pain started, but the second troponin a few hours later is 100s/1000s.

Although high sensitivity troponin (my shops use them exclusively now) are somewhat frustrating because of poor specificity, they are quite sensitive and I do feel confident that if you have normal limits hsTn more than 6 hours after chest pain started, things are very reassuring that it is not an ACS. I do think they are useful when they are negative. (When they are positive is a different story...)

All that being said, risk factors matter. And the reason is, even if a patient is NOT really having an ACS when you see them in the ER with chest pain, if you send a high risk chest pain patient home and they die of an MI a days, weeks, months, later from an actual ACS, people will say you "missed" their ACS because in the mind of the public/QI/lawyers, "Chest pain = ACS." If you send a high risk chest pain home who dies of an MI 1month later, you "missed the ACS." My problem is these patients even if I think they are not having an ACS today, may have one in the near future because they do have objective risk factors. If you are the last physician with hands on this patient, you are f'd. Close follow up helps mitigate this risk, but in many systems, being hospitalized is the only way to get a stress test in a somewhat timely fashion.

 

[SpacemanSpifff]

Although high sensitivity troponin (my shops use them exclusively now) are somewhat frustrating because of poor specificity, they are quite sensitive and I do feel confident that if you have normal limits hsTn more than 6 hours after chest pain started, things are very reassuring that it is not an ACS. I do think they are useful when they are negative. (When they are positive is a different story...)

All that being said, risk factors matter. And the reason is, even if a patient is NOT really having an ACS when you see them in the ER with chest pain, if you send a high risk chest pain patient home and they die of an MI a days, weeks, months, later from an actual ACS, people will say you "missed" their ACS because in the mind of the public/QI/lawyers, "Chest pain = ACS." If you send a high risk chest pain home who dies of an MI 1month later, you "missed the ACS." My problem is these patients even if I think they are not having an ACS today, may have one in the near future because they do have objective risk factors. If you are the last physician with hands on this patient, you are f'd. Close follow up helps mitigate this risk, but in many systems, being hospitalized is the only way to get a stress test in a somewhat timely fashion.

The last paragraph is the key to my, and I would imagine most docs', frustration. As others have said, some folks walk around with a HEART of 4 every day, and if they ever have random chest pain and come in it's hard to discharge them unless you a) don't care about malpractice risk and b) have a good follow up system in place to minimize the hot potato phenomenon (i.e. you touched them last, you're left with the bill). That said, these folks are always going to be at relatively high risk for MACE in the 30 day period after discharge. And, does a low risk stress or a period of negative serial troponins change that risk? No. Although it might change the risk of a lawsuit.

1. Depending on which site I'm working at, either there is this complex flow chart that people use to sort out the delta level (university site), or 2 negative levels = home +/- outpatient cards f/u if feeling good or admit if suspicion is high (community sites).
2. none of my sites really use HEART anymore because the cards docs don't think its useful and I generally agree; gestalt and experience and the context of the patient is so much more useful than that score especially when at least half of your undifferentiated chest pain patients have CAD
3. HS-Trop <6 with a low pretest probability of ACS is all I need to d/c on one trop.
4. Since we call so infrequently at my community sites (most go home with outpatient) if the ED doc wants someone observed its typically pretty easy to admit and normally its due to ongoing pain ie unstable angina.

Fair enough on the bolded, see above. I've never talked to a cardiologist or internist who liked the score because it inevitably creates work for them. Any 65 year old who smokes, has HTN, obesity and DM is a 4 and therefore not low risk. Throw in a left bundle, and boom, you've got a consult or admission for risk stratification.

 

[thegenius]

All that being said, risk factors matter. And the reason is, even if a patient is NOT really having an ACS when you see them in the ER with chest pain, if you send a high risk chest pain patient home and they die of an MI a days, weeks, months, later from an actual ACS, people will say you "missed" their ACS because in the mind of the public/QI/lawyers, "Chest pain = ACS." If you send a high risk chest pain home who dies of an MI 1month later, you "missed the ACS." My problem is these patients even if I think they are not having an ACS today, may have one in the near future because they do have objective risk factors. If you are the last physician with hands on this patient, you are f'd. Close follow up helps mitigate this risk, but in many systems, being hospitalized is the only way to get a stress test in a somewhat timely fashion.

I wouldn't send home a HEART score of 6-9. Even though the Kaiser data shows even a 6! has a 30 day MACE of 1.82%.

I send up up to 4 all the time, and often send home 5. I let the patient know there is about a 1% chance of having a heart attack in 30 days. They almost always want to go home. If you see a patient and a bad thing happens to that patient in the coming weeks, you might to be sued. Chest pain is no different.

Chest Pain (to D/c or not to D/c)

After having a lengthy conversation with a few other attendings, it was really funny and interesting to see how 4 different doctors approached chest pain disposition all differently. Some use a hard cutoff heart score, others a gestalt of 1 or 2 trops and dispo home with followup... some just...

forums.studentdoctor.net

Just look at that graph in the link above. 30 day MACE for intermediate HEART (4-6) is just 1%. That's it.

 

[southerndoc]

I wouldn't send home a HEART score of 6-9. Even though the Kaiser data shows even a 6! has a 30 day MACE of 1.82%.

I send up up to 4 all the time, and often send home 5. I let the patient know there is about a 1% chance of having a heart attack in 30 days. They almost always want to go home. If you see a patient and a bad thing happens to that patient in the coming weeks, you might to be sued. Chest pain is no different.

Chest Pain (to D/c or not to D/c)

After having a lengthy conversation with a few other attendings, it was really funny and interesting to see how 4 different doctors approached chest pain disposition all differently. Some use a hard cutoff heart score, others a gestalt of 1 or 2 trops and dispo home with followup... some just...

forums.studentdoctor.net

Just look at that graph in the link above. 30 day MACE for intermediate HEART (4-6) is just 1%. That's it.

Keep in mind that most HEART studies were done with traditional troponins and not high-sensitivity troponins. You run a mile and your hsTrop will be elevated. I wouldn't discharge a HEART of 6 with a regular troponin. HEART 6 with hsTrop is a different deal. 75 year old with HTN, obesity, and dyslipidemia is a HEART of 4 just walking around. Throw in a non-specific EKG and you get another point. People could be overzealous with their clinical history and give a point there.

The real issue is that people misapply HEART scores. HEART scores aren't for every chest pain. It's for people you are suspicious are having acute coronary syndrome. Hence 0 points is low suspicion. If HEART was intended for every chest pain, you would have a no suspicion category. Pleuritic pain, positional pain, etc. is unlikely to be cardiac in origin. Not impossible but highly unlikely.

 

[Tenk]

I have yet to see two high sensitivity troponin in the setting of ongoing chest pain miss ACS and I have been using them for years now.

That being said, if negative trop but older with risk factors, high heart score, etc, I just offer admission. Most patients nowadays refuse, I then document they were offered and refused and move on.

 

[southerndoc]

If you've ruled out an MI, docs usually say they want to admit to get a stress test. It's been a while since I've looked through the literature, and granted it wasn't an exhaustive search, but inpatient stress tests do not reduce 30-day mortality.

 

[pollyspockets]

If you've ruled out an MI, docs usually say they want to admit to get a stress test. It's been a while since I've looked through the literature, and granted it wasn't an exhaustive search, but inpatient stress tests do not reduce 30-day mortality.

Completely agree based on literature. However in my region it’s still local standard of care and I can imagine a lawyer ready to sue me based on timi or heart. It really seems like there is not a lot of evidence for stress and even cath in non-occlusive MI. All those stents are really useless- doesn’t improve mortality or anginal symptoms (again no OMI cohort).

 

[RoyBasch]

I have yet to see two high sensitivity troponin in the setting of ongoing chest pain miss ACS and I have been using them for years now.

Ok but what is your gold standard for confirmation of a miss/no miss? Patient bounces back to your er? Are you checking the EMR for a few days/weeks after every chest pain discharge? Does QI notify you? By definition: you missed the cases you miss. You don’t necessarily know about the patient you discharge who goes home and dies or goes by EMS to a different ER with arrhythmia, ACS, arrest, etc.

 

[SpacemanSpifff]

If you've ruled out an MI, docs usually say they want to admit to get a stress test. It's been a while since I've looked through the literature, and granted it wasn't an exhaustive search, but inpatient stress tests do not reduce 30-day mortality.

Totally agree. Does it reduce the likelihood of being named in a lawsuit for an unrelated cardiac event in the next 30 days? I think that’s the outcome this practice pattern is aiming for.

What is the delta that qualifies as “positive” for this of you with hs-trop? My current employer has it at 5, previous was 7.

 

[Tenk]

Ok but what is your gold standard for confirmation of a miss/no miss? Patient bounces back to your er? Are you checking the EMR for a few days/weeks after every chest pain discharge? Does QI notify you? By definition: you missed the cases you miss. You don’t necessarily know about the patient you discharge who goes home and dies or goes by EMS to a different ER with arrhythmia, ACS, arrest, etc.

My gold standard is I’ve been doing this for years and haven’t heard a peep. We do have systems in place that would likely notify me of such a big miss.

That being said, I didn’t follow discharges but I did follow almost all of the admitted patients with negative trops. None of them popped positive and also none of them got cathed.

 

[RoyBasch]

My gold standard is I’ve been doing this for years and haven’t heard a peep. We do have systems in place that would likely notify me of such a big miss.

That being said, I didn’t follow discharges but I did follow almost all of the admitted patients with negative trops. None of them popped positive and also none of them got cathed.

For the record: I agree with you, hsTn, is very sensitive. If they are negative x2, that is very reassuring the actual episode at the time of evaluation is not a true ACS. I mean for chrissake they elevate with most patients just walking from the parking lot into the building.

 

[southerndoc]

Totally agree. Does it reduce the likelihood of being named in a lawsuit for an unrelated cardiac event in the next 30 days? I think that’s the outcome this practice pattern is aiming for.

What is the delta that qualifies as “positive” for this of you with hs-trop? My current employer has it at 5, previous was 7.

If you admit a patient and they have a stress test, if they go home and die you're probably going to be named in the litigation anyhow.

Delta >3 gets further workup.

 

[aafisahar]

Stress testing finds lesions unlikely to be the cause of ACS. By almost definition plaque that can cause symptoms of angina is going to be calcified, fibrous plaque that won't rupture and cause an acute occlusion. This is what stress testing picks up. ACS is caused by the 30% lesion the ruptures to form thrombus and myocardial injury.

Troponin in the setting of ACS is looking for the latter while stress testing is looking for the former. It's mostly a historical practice patten that leads us to stressing every chest pain rule out when the likelihood of finding an acute plaque rupture is quite minimal. If they had troponin they should have been cathed. Except for UA, thrombus should cause troponin especially with the hs assay. Stressing with two negative hs troponins is basically trying to find a thrombus big enough to cause symptoms but small enough to not cause injury. And this is probably why every look at stress testing for suspected ACS has come up snake eyes.

My decision making is basically, is this potentially ACS, if yes, then cath. If no, investigate alternatives (eg chf, af, valvular dz, etc) and if absent, not my problem. Stress testing is for mostly outpatients or for other doctors who want to "do something" in the inpatient setting.

 

[turkeyjerky]

Anyone use, for reasoning or just to document, the EDACS score? I kind of like it in the (at baseline) moderate to high risk HEART score patients, since it allows some stratification in the elderly cohort. Eg Mee-maw, the 80 year old diabetic female w/ R sided pleuritic pain and chest wall tenderness is still low risk.

I'm always a little perplexed by the admontation that, if a patient is having persistent chest pain, then they must be admitted. I'm typically far more confident of my diagnostic tools (eg history, exam, ekg, hsTrop 2 hours from now) if the patient is still having chest pain during the encounter. Anyone care to correct my thinking on this?

 

[BoardingDoc]

I'm always a little perplexed by the admontation that, if a patient is having persistent chest pain, then they must be admitted. I'm typically far more confident of my diagnostic tools (eg history, exam, ekg, hsTrop 2 hours from now) if the patient is still having chest pain during the encounter. Anyone care to correct my thinking on this?

I don't think anyone I work with still thinks this. I remember having to do this in residency but it never made sense to me and I certainly don't do it as an attending. I'll also take it one step further and DC constant CP for several hours with a single trop, no repeat. This of course depends on clinical gestalt, but if you tell me you have had atypical CP constantly since you woke up, a reassuring EKG and a single negative trop at 2pm is perfectly fine with me for you to go home with ongoing CP.

 

[thegenius]

Anyone use, for reasoning or just to document, the EDACS score? I kind of like it in the (at baseline) moderate to high risk HEART score patients, since it allows some stratification in the elderly cohort. Eg Mee-maw, the 80 year old diabetic female w/ R sided pleuritic pain and chest wall tenderness is still low risk.

I'm always a little perplexed by the admontation that, if a patient is having persistent chest pain, then they must be admitted. I'm typically far more confident of my diagnostic tools (eg history, exam, ekg, hsTrop 2 hours from now) if the patient is still having chest pain during the encounter. Anyone care to correct my thinking on this?

yea up to a certain age. I don't think I'll be sending home a 75 yo mee maw with chest pain w/ neg trops / EKGS....reason is I'll see them 4 hours later in my ER again for the same crap

 

[RustedFox]

yea up to a certain age. I don't think I'll be sending home a 75 yo mee maw with chest pain w/ neg trops / EKGS....reason is I'll see them 4 hours later in my ER again for the same crap

Yeah. There's so many different ways to think about this topic, and they're all valid.

It's agitating.

Send home the meemaw and they come back "because".

Send home the 37 year old with ongoing pain and it's a big OOF if the outcome is bad.

I had the 55 year old guy with middle of the road risk factors and two stone cold negative hsTnI values go into a v-fib arrest a few weeks back. I would have sent him home with cardio follow up.

 

[thegenius]

I had the 55 year old guy with middle of the road risk factors and two stone cold negative hsTnI values go into a v-fib arrest a few weeks back. I would have sent him home with cardio follow up.

And it would have been fine 99.6% of the time too.

 

[RustedFox]

And it would have been fine 99.6% of the time too.

I know, right ?!
You can argue that this is "unstable angina", or you can argue that its "a nothing".
And both arguments can be reasonably well-supported.
It's a hard clinical question, no doubt.

I say this with sincerity; I'm not just trying to stir the pot, as I so often do.

 

[RustedFox]

So, this is our flowchart.

Let's throw out the red section, as obvious trops are obvious.

Let's talk about the green section. To qualify, you have to roll less than a six to start.
Not six. LESS THAN SIX. Not less than or EQUAL TO six. Less than six.
LOL @ that. Nobody rolls less than a six.
Your mom rolls a 9. Ooooh.
Now everybody is in the yellow section.

The yellow section means... nothing. Observed. The asterisk says "think about things and maybe a third trop. Think about HEART score."

Ok, now I'm right where I was before hsTnI.
Can I just go back to normal trops and a 0-2 hour window? That made more sense to me.

Open to criticism.

 

[RustedFox]

That's a traditional troponin.

You mean a REAL troponin?

 

[thegenius]

Have a hard time believing that a change of 10 (high sensitivity troponin) is a marker for AMI.

Using old troponins...
0.02, then 0.03 -> basically equivalent given the statistical ability to detect troponin. This is not a heart attack. No occlusive myocardial infarctions go 0.02, then 0.03

Using high sensitivity troponins
20, then 30 -> possible AMI? That's a joke.

Must be this 1 hr thing. My shop we wait two hours. I think the delta is like 30 or 35 or something to "rule in"

 

[turkeyjerky]

Have a hard time believing that a change of 10 (high sensitivity troponin) is a marker for AMI.

Using old troponins...
0.02, then 0.03 -> basically equivalent given the statistical ability to detect troponin. This is not a heart attack. No occlusive myocardial infarctions go 0.02, then 0.03

Using high sensitivity troponins
20, then 30 -> possible AMI? That's a joke.

Must be this 1 hr thing. My shop we wait two hours. I think the delta is like 30 or 35 or something to "rule in"

I think this is the whole crux of this lab assay. "high sensitivity" is a misnomer, in my understanding. They really should be called "highly precise" troponins, as the assay is so precise that that small change is actually real (whether or not related to type 1 ACS is a different matter...)

Fwiw, I have had a few patients in the recent past where their 1 hr repeat went from 40 -> 50 or some small change like that, with both numbers black on the computer screen, whose levels subsequently spiked up to several hundred/thousand at 6 hrs after admission.

 

[Mount Asclepius]

In the past with regular troponins I used to treat troponin testing as normal even if minimally uptrending if still within the normal limits. By definition, ACS has an exponential rise in troponin testing. However, with many of the newer protocols in place, ACS is only excluded if the second high sensitivity troponin only rises by a certain amount even if still within normal limits. If a second high sensitivity troponin is still within normal limits, but rising beyond the protocol’s ability to exclude ACS, are folks still counting as normal and discharging patients if not otherwise concerned, trending a third troponin in the ED, or admitting the patient?

 

[thegenius]

I don't get third troponins anymore, unless there are extenuating circumstances. I probably do it 1-3x/year. And I see chest pain 1-3x/shift.

 

[Mount Asclepius]

I don't get third troponins anymore, unless there are extenuating circumstances. I probably do it 1-3x/year. And I see chest pain 1-3x/shift.

So if a high sensitivity troponin is still within normal limits, but minimally rising with inability to exclude ACS, do you discharge or admit?

 

[thegenius]

So if a high sensitivity troponin is still within normal limits, but minimally rising with inability to exclude ACS, do you discharge or admit?

I use 2 hr High sens Troponins. So I need a good bump up to admit. If it goes from 20 to 30 in 2 hours. I'll probably discharge. In old troponin land that's 0.02 - 0.03. That's a nonsense rise. I tend not to get a third troponin because I could then get 20, 30, and 40. What do I do with that? It's clearly not an occlusive myocardial infarction. It's some other nonsense that likely isn't life threatening.

Sometimes I'll admit a 20 -> 30 because the patient is scared or i don't want to think about it, or whatever. We at some point have to make a decision about what to do with these patients. if they are sitting in the ER sleeping, or on their iphone and over 3-4 hours their trop goes from 20 to 30, they are not "sick" and you can just do what you want with them. They might go into vfib in 4 hours, they might croak the next day, but there really isn't anything you can do to prevent it if they are sleeping or watch "Days Of Our Lives" on the TV in the room.

I think if ER docs accept the fact that they will never be 100% sensitive in ruling out chest pain, and that's OK, they would be more OK with these decisions.

 

[thegenius]

For all but the high risk patients:
it's very easy to just tell them "there is about a 1% chance you'll have a heart attack in the next 30 days. that's what the data says." Make it a nice an even percentage. don't say 0.6%, 1.2%...just "about 1%". Let the patient decide re: admit or d/c. BTW, I don't let the patient decide if they are a 1, 2, or even a 3.

 

[Mount Asclepius]

I use 2 hr High sens Troponins. So I need a good bump up to admit. If it goes from 20 to 30 in 2 hours. I'll probably discharge. In old troponin land that's 0.02 - 0.03. That's a nonsense rise. I tend not to get a third troponin because I could then get 20, 30, and 40. What do I do with that? It's clearly not an occlusive myocardial infarction. It's some other nonsense that likely isn't life threatening.

Sometimes I'll admit a 20 -> 30 because the patient is scared or i don't want to think about it, or whatever. We at some point have to make a decision about what to do with these patients. if they are sitting in the ER sleeping, or on their iphone and over 3-4 hours their trop goes from 20 to 30, they are not "sick" and you can just do what you want with them. They might go into vfib in 4 hours, they might croak the next day, but there really isn't anything you can do to prevent it if they are sleeping or watch "Days Of Our Lives" on the TV in the room.

I think if ER docs accept the fact that they will never be 100% sensitive in ruling out chest pain, and that's OK, they would be more OK with these decisions.

I don’t overall disagree, but now there are protocols in place with these high sensitivity troponins that hang you out to dry with the lawyers when you “didn’t follow the protocol.” In the past there wasn’t a protocol. The regular troponin was either abnormal or normal, even if detectable (I’m aware every lab test is dependent on multiple factors, not necessarily binary and only one point in time). I typically discharge patients with two high sensitivity troponins that are within normal limits, but occasionally it is against the protocol due to a ‘trivial’ rise. I’m just curious what others are doing.

 

[RoyBasch]

I don’t overall disagree, but now there are protocols in place with these high sensitivity troponins that hang you out to dry with the lawyers when you “didn’t follow the protocol.” In the past there wasn’t a protocol. The regular troponin was either abnormal or normal, even if detectable (I’m aware every lab test is dependent on multiple factors, not necessarily binary and only one point in time). I typically discharge patients with two high sensitivity troponins that are within normal limits, but occasionally it is against the protocol due to a ‘trivial’ rise. I’m just curious what others are doing.

Here's what I do when I get the two hsTn both within normal limits but second is higher than the first (and above the delta considered significant by our protocol--which is very small, its like 6 I believe).

I make a gut call, do I think the patient is legit or not. If they seem legit (older, risk factors, concerning hx) then I just admit the patient for an ACS rule out and invoke the protocol to get no hospitalist push-back.

If I think they are not legit (patient is young, no risk factors, weak hx) I phone-consult cardiology; who invariably tell me they are not concerned about the trivial rise and to discharge the patient. I then document the cardiologist was not impressed and their recommendation to discharge the patient and that they will follow them up closely in clinic (haha right).

I seriously imagine consultants sitting around the doctor's lounge joking with each other: "And then I told the ER we would follow them up CLOSELY in clinic!!!!" (raucous laughter all around).

 

[RustedFox]

I don’t overall disagree, but now there are protocols in place with these high sensitivity troponins that hang you out to dry with the lawyers when you “didn’t follow the protocol.” In the past there wasn’t a protocol. The regular troponin was either abnormal or normal, even if detectable (I’m aware every lab test is dependent on multiple factors, not necessarily binary and only one point in time). I typically discharge patients with two high sensitivity troponins that are within normal limits, but occasionally it is against the protocol due to a ‘trivial’ rise. I’m just curious what others are doing.

Yeah, THIS is exactly my qualm. They gave us a protocol and expect us to follow it... but their protocols are of limited use IRL.

Have a look at the algorithm that I posted above again. I'd rather not have an algorithm to be misrepresented and go back to good clinical sense.

 

[SpacemanSpifff]

In the past with regular troponins I used to treat troponin testing as normal even if minimally uptrending if still within the normal limits. By definition, ACS has an exponential rise in troponin testing. However, with many of the newer protocols in place, ACS is only excluded if the second high sensitivity troponin only rises by a certain amount even if still within normal limits. If a second high sensitivity troponin is still within normal limits, but rising beyond the protocol’s ability to exclude ACS, are folks still counting as normal and discharging patients if not otherwise concerned, trending a third troponin in the ED, or admitting the patient?

You discharged an uptrending classic trop? Interesting - these were blanket admits where I trained. You were either normal (<0.04), or detectable. If not normal, vast majority got admitted.

All three EDs I have worked at since have specific protocols in place, with different “acceptable” deltas for hs-trop. I can’t imagine these are particularly evidence based. Ultimately it would be nice to be able to have admin support you in a judgement call, but that seems like a fantasy.

 

[Mount Asclepius]

You discharged an uptrending classic trop? Interesting - these were blanket admits where I trained. You were either normal (<0.04), or detectable. If not normal, vast majority got admitted.

All three EDs I have worked at since have specific protocols in place, with different “acceptable” deltas for hs-trop. I can’t imagine these are particularly evidence based. Ultimately it would be nice to be able to have admin support you in a judgement call, but that seems like a fantasy.

Yes, in the past I fairly frequently discharged low risk HEART score patients with detectable classic normal troponins that were within normal limits. Always seemed that every patient with ESRD had a detectable classic troponin at baseline. This was the practice pattern where I trained and in other locations since. Even if detectable, it's still normal, not an exponential rise, and not consistent with ACS. I could never understand other EP's practice pattern of trending out 3-5+ troponins all essentially unchanged. They also seemed to be the types that would trend CK levels from 200s to 300s to 400s not realizing that is insignificant, or trend lactates seemingly incessantly from 2.3 to 2.2 to 2.1 to 1.8. I've never seen a protocol regarding classic troponins that were detectable, but within normal limits. However, with high sensitivity troponins everyone has a protocol in place regarding inability to rule out ACS if some trivial rise even if still normal. Makes timely ED disposition of chest pain more difficult and leaves you out on a limb if you don't follow a protocol.