Do You Admit TIA's?

[WilcoWorld]

I've long wondered about the benefit of admitting TIA's. This study* is limited in that it's retrospective, but it casts further doubt on the thought that inpatient TIA workups are helpful.


* Impact of Hospital Admission for Patients with TIA

---

Members don't see this ad.

 

[southerndoc]

ACEP clinical policy: MRI if you're going to discharge. They went against the ABCD2 recommendation for risk stratification. Basically ACEP says if you don't MRI and also image the carotids in some way, they aren't safe for discharge and should be admitted. I don't agree with this.

 

[xaelia]

We admit for higher ABCD2 scores for the expedited evaluation, less-convincing and/or low scores get urgent outpatient – but we have the integrated health system where we can schedule these in a tidy fashion.

Outside of this system, would offer admission/observation – with full disclosure to lower-risk patients their testing would be cheaper done outpatient, but non-zero risk of potentially preventable events.

 

[shoal007]

we admit. wait time is too long. will soom be in our ED observation unit

Sent from my Pixel using Tapatalk

 

[migm]

If i think it's a TIA, I'll admit it. Our neurologists have asked for this - at this point, follow up is not sufficiently speedy and the PCPs don't reliably handle it (some feel comfortable, others don't). Not sure I understand the disagreement regarding not MRI'ing as southerndoc points out. Even if symptoms have resolved, MRI findings constitute a better predictor than even abcd2 regarding progression to stroke.

 

[Angry Birds]

What if it's a complete softie, like "I had numbness and tingling on the left side of my face and left hand, and it's gone..." All work up is negative. Patient wants to go home. Then what?

(Maybe I also need to brush up on what exactly constitutes a TIA. Any neurological symptoms that resolve within 24 hours? What are you telling the admitting doc? I say, "I have a TIA patient who needs an MRI +/- neuro eval..." And what diagnosis do you put as an admitting diagnosis? I put TIA, r/o stroke. Is that good?)

 

[southerndoc]

If i think it's a TIA, I'll admit it. Our neurologists have asked for this - at this point, follow up is not sufficiently speedy and the PCPs don't reliably handle it (some feel comfortable, others don't). Not sure I understand the disagreement regarding not MRI'ing as southerndoc points out. Even if symptoms have resolved, MRI findings constitute a better predictor than even abcd2 regarding progression to stroke.

I don't think I was very clear with what I was trying to say. If you have someone with an ABCD2 of 1 and plan to discharge them (i.e., maybe they had 5 minutes of slurred speech that is now resolved, no weakness, etc.). Even if you can arrange for <72 hour follow-up for MRI, new ACEP clinical guideline from 2016 says you really need to get an MRI brain and imaging of the carotids (CTA, MRA, or ultrasound) before discharging the patient. I'm not arguing they don't need an MRI (many TIA's are actually small strokes), but whether they need this in the ER prior to discharge is what I'm arguing.

 

[migm]

What if it's a complete softie, like "I had numbness and tingling on the left side of my face and left hand, and it's gone..." All work up is negative. Patient wants to go home. Then what?

(Maybe I also need to brush up on what exactly constitutes a TIA. Any neurological symptoms that resolve within 24 hours? What are you telling the admitting doc? I say, "I have a TIA patient who needs an MRI +/- neuro eval..." And what diagnosis do you put as an admitting diagnosis? I put TIA, r/o stroke. Is that good?)

I think hand and face qualifies, it's a TIA until proven otherwise - alternate a more likely explanation such as complex migraine. "r/o xyz" is not a a billable complaint. Appropriate here would be TIA or stroke like symptoms. And yes, neuro and/or IM needs to see these patients, just like someone with a mild rise in trop and a concerning story would need to see cardiology. I am not getting lipid panels, evaluating carotids, getting echos, doing telly monitoring in the ED. Someone needs to minimize the patients risk factors and make sure they don't have the big one.

I don't think I was very clear with what I was trying to say. If you have someone with an ABCD2 of 1 and plan to discharge them (i.e., maybe they had 5 minutes of slurred speech that is now resolved, no weakness, etc.). Even if you can arrange for <72 hour follow-up for MRI, new ACEP clinical guideline from 2016 says you really need to get an MRI brain and imaging of the carotids (CTA, MRA, or ultrasound) before discharging the patient. I'm not arguing they don't need an MRI (many TIA's are actually small strokes), but whether they need this in the ER prior to discharge is what I'm arguing.

I guess the point is, even if the symptoms have resolved, if hte MRI is positive for ischemic findings, you need to admit them to the hospital (as the risk for progressing to CVA is much larger), so in that it changes your disposition they would need it prior to leaving. Similarly, if they have symptomatic carotid stenosis, they probably need a roto-rooter with surgery on an urgent basis. That was my understanding, anyway.

 

[WilcoWorld]

I don't think I was very clear with what I was trying to say. If you have someone with an ABCD2 of 1 and plan to discharge them (i.e., maybe they had 5 minutes of slurred speech that is now resolved, no weakness, etc.). Even if you can arrange for <72 hour follow-up for MRI, new ACEP clinical guideline from 2016 says you really need to get an MRI brain and imaging of the carotids (CTA, MRA, or ultrasound) before discharging the patient. I'm not arguing they don't need an MRI (many TIA's are actually small strokes), but whether they need this in the ER prior to discharge is what I'm arguing.

I think hand and face qualifies, it's a TIA until proven otherwise - alternate a more likely explanation such as complex migraine. "r/o xyz" is not a a billable complaint. Appropriate here would be TIA or stroke like symptoms. And yes, neuro and/or IM needs to see these patients, just like someone with a mild rise in trop and a concerning story would need to see cardiology. I am not getting lipid panels, evaluating carotids, getting echos, doing telly monitoring in the ED. Someone needs to minimize the patients risk factors and make sure they don't have the big one.


I guess the point is, even if the symptoms have resolved, if hte MRI is positive for ischemic findings, you need to admit them to the hospital (as the risk for progressing to CVA is much larger), so in that it changes your disposition they would need it prior to leaving. Similarly, if they have symptomatic carotid stenosis, they probably need a roto-rooter with surgery on an urgent basis. That was my understanding, anyway.

The study that I linked at the beginning of the thread would appear to support SouthernDoc's point. "Although admitted patients were more likely to receive diagnostic procedures, we did not identify improvements in outcomes among admitted patients."

 

[migm]

The study that I linked at the beginning of the thread would appear to support SouthernDoc's point. "Although admitted patients were more likely to receive diagnostic procedures, we did not identify improvements in outcomes among admitted patients."

I think discharging high risk patients from the ED has been proven to be associated with worse outcomes, no? +MRI in TIA is one of those high risk findings.

 

[Venko]

We routinely discharge people with TIA and use the ABCD3i system. Basically ABCD2 score plus imaging of carotid. We also have an ED observation pathway for TIA though it's not used very much. For people where follow up or follow through can be predicted to be problematic we admit to neuro.


Sent from my iPhone using SDN mobile

 

[migm]

I think there's a lot to be said for standard of care in your environment, so long as it's reasonable, it's probably defensible.

@Venko - how does the carotid ultrasound change your score? What is your cutoff? what does follow up for less than cut off look like? do you admit all greater than x score on abc3i?

 

[goodoldalky]

We routinely discharge people with TIA and use the ABCD3i system. Basically ABCD2 score plus imaging of carotid. We also have an ED observation pathway for TIA though it's not used very much. For people where follow up or follow through can be predicted to be problematic we admit to neuro.


Sent from my iPhone using SDN mobile

Would you mind sharing this? PM if necessary.

 

[WilcoWorld]

I think discharging high risk patients from the ED has been proven to be associated with worse outcomes, no? +MRI in TIA is one of those high risk findings.

I have not exhaustively reviewed the literature. I thought the article (Impact of Hospital Admission for Patients with TIA) had interesting (though, admittedly, preliminary) findings. Hence this thread. If you have evidence that comes down on the other side, please post a link. I am genuinely interested in a discussion.

 

[Zebra Hunter]

I think discharging high risk patients from the ED has been proven to be associated with worse outcomes, no? +MRI in TIA is one of those high risk findings.

A positive MRI for ischemic findings means it's not a TIA.

 

[migm]

I have not exhaustively reviewed the literature. I thought the article (Impact of Hospital Admission for Patients with TIA) had interesting (though, admittedly, preliminary) findings. Hence this thread. If you have evidence that comes down on the other side, please post a link. I am genuinely interested in a discussion.

My understanding is that there is a subset of patients who would in fact benefit from admission, abcd is probably not enough to figure out who those are or aren't. That's why you're seeing combined scores as a way to increase the fidelity. See this study regarding discharge of high risk TIA's

Is hospitalization after TIA cost-effective on the basis of treatment with tPA?

 

[migm]

A positive MRI for ischemic findings means it's not a TIA.

Indeed, it does, as ACEP and AAN currently define TIA as the resolution of symptoms without evidence of permanent infarction on imaging. You're definitely in a grey zone here, but DWI findings in patients with resolved symptoms that seem to resolve with time still constitute TIA. http://stroke.ahajournals.org/content/strokeaha/35/5/1095.full.pdf

 

[Noroozi]

Indeed, it does, as ACEP and AAN currently define TIA as the resolution of symptoms without evidence of permanent infarction on imaging. You're definitely in a grey zone here, but DWI findings in patients with resolved symptoms that seem to resolve with time still constitute TIA. http://stroke.ahajournals.org/content/strokeaha/35/5/1095.full.pdf

I think the article you referred is no longer applicable, as it was written in 2004 before the definition of TIA was refined. DWI changes on MRI associated with corresponding ADC means there is infarction of brain tissue and is no longer a TIA.

 

[BJJVP]

In my practice, I tend to place these patients in observation for cardiac monitoring, carotids, echo, MRI and a neuro consult. I agree that hospitalizing these pts probably doesn't help long term outcomes very much. And while I am not one of those EPs that practices defensive medicine, I just don't think that the general public is going to be very understanding if the pt ends up developing a full blown CVA during the short term follow-up period. It is hard to tell pts, "Yes, I know you couldn't talk or move your arm for 10 minutes. There may even be a directly treatable reason why this happened but it is somewhat unlikely that this applies to you. It could happen again and even be permanent the next time, but because you're OK now, I will send you home." I don't really blame the general population for their expectations in this case. Heck, I don't think most doctors and nurses would be forgiving if you discharged their family member home and the patient subsequently stroked out. There is a PCP at my hospital who thinks everything can be done as an outpt. He never wants to admit anyone, even people he sends to the ED. He once argued with me about admitting a pt with a 2 day old stroke (the pt finally couldn't walk anymore and so he came to the ED). He's a pain in my neck. But I can tell you now, if I sent his wife home after a TIA, and she stroked out 2 days later, he would announce to the rest of the hospital, all the administrators, and anyone willing to listen how incompetent the ED is. As frustrating as it can be, I don't even try to fight this one. With these pts, I will admit or place in observation (whatever they call it nowadays) because many doctors, nurses, and the general public have been led to believe that our treatments for cerebral ischemia are far better than they actually are.

 

[Venko]

I think there's a lot to be said for standard of care in your environment, so long as it's reasonable, it's probably defensible.

@Venko - how does the carotid ultrasound change your score? What is your cutoff? what does follow up for less than cut off look like? do you admit all greater than x score on abc3i?

The main concern of TiA is the impending stroke. The preventable causes include carotid stenosis (usually critical is 95% or greater, which is above 200 cm/sec velocity on ultrasound). So if we have benign ECG, no murmur, and clean carotids with low ABCD2...they are not likely to have a preventable cause of impending stroke.

So ABCD2 low, no critical stenosis --> discharge


Sent from my iPhone using SDN mobile

 

[Venko]

Would you mind sharing this? PM if necessary.

There are different versions of ABCD3-I scoring systems. A good overview currently is:

Kiyohara T, et al. ABCD and ABCD3-I are superior score in the Prediction of Short- and Long-Term Risks of Stroke After Transient Ischemic Attack. stroke. 2014;45:418-425

Our version does not include MRI and our cutoff on carotids is "critical carotid stenosis" usually > 95%


Sent from my iPhone using SDN mobile

 

[migm]

I think the article you referred is no longer applicable, as it was written in 2004 before the definition of TIA was refined. DWI changes on MRI associated with corresponding ADC means there is infarction of brain tissue and is no longer a TIA.

The 2016 ACEP guidelines for TIA discuss reversible DWI findings on MRI as compatible with TIA and as being predictive for CVA in a 7 day period, but not definitive of stroke.

 

[DrYoda]

I'm with southerndoc. If you have a normal MRI you can peace out otherwise you need other stuff done.

 

[TheBlueBlazer]

Had a patient that had TIA symptoms, having had the same a couple months prior, where she had the full workup. She was back to her baseline, already on maximal therapy, and after talking with her PMD, there wasn't anything I could offer her in the hospital that she couldn't get at home, other than faster door-to-TPA time, which could be a good thing given that each time this had resolved without intervention. I was a little hesitant, but she was motivated to go home. I used my shared decision-making language instead of my AMA language.

 

[justaregularmed]

....

 

[Birdstrike]

To no one in particular:

"Don't admit TIAs. It's a waste of time and money! Pointless."

"This person stroked out and died after you refused to admit them. Don't you know TIAs are at high risk to have subsequent severe and irreversible strokes, dummy?!"


Sent from my iPhone using SDN mobile