Idiopathic intracranial hypertension

[Bougiebuster]

Typical overweight female with classic symptoms. POCUS with large ONSD.
How many of you guys are doing the LP in the ER? Any evidence that this needs to be done emergently?
Do you admit for MRI/MRV?
Do you start them on acetazolamide before confirming dx?

---

Members don't see this ad.

 

[southerndoc]

ONSD?

I refer them to neurology unless their headache is either worrisome for a SAH or is intractable. If they get pain relief, then it's an outpatient workup facilitated by IR-guided LP since these people are usually obese and difficult for blind LP's. I don't start acetazolamide unless I diagnose with an elevated opening pressure as it can cloud the diagnosis when someone actually does an LP.

 

[thegenius]

Optic Nerve Sheath Diameter?

 

[Darth Doc]

I don't think optic nerve sheath diameter measurements are standard of care for diagnosis - at least not yet. Non-emergent diagnosis. I'm in a tertiary care center, so we consult neurology before discharge, but it's outpatient diagnosis without acetazolamide. (Neurology does try to get us to do the LP, but we don't do it.)

 

[Bougiebuster]

yeah optic nerve sheath diameter to suggest papilledema...I wouldn't trust my ophthalmoscope skills to see this otherwise

 

[thegenius]

I don't consider most cases of IIH a medical emergency. I view this as assessing severity of symptoms. If patients are really having a hard time seeing, their VA is low, and have bad headaches it might be worth doing the LP now.

Some people have mild symptoms and I think it's probably OK to start diamox (if neg CT Head) and refer to Neuro / Ophtho.

However where I work...it's hard to get patients to see anybody like that within several days.


I had a young woman come in with weird history of headaches....I opted to do an LP and I can't remember why. I don't think I was looking specifically for IIH. But I can't remember. Anyway her pressure was > 55, the CSF came out the top of the manometer!!! I took off about 20-25 cc, she felt better...and I think I admitted her because I didn't know what to do and didn't have Neuro available.

Neuro saw her the next day, started diamox, and discharged her.

She saw ophtho as an outpatient and about 3-6 months later, despite maximal medical therapy and weight loss attempts, her vision got markedly worse and got shunted with 100% resolution of symptoms.

 

[Mushdoc]

I’ve had ophtho tell me (as a resident) that they think any visual changes/loss should be seen pretty emergently due to risk of vision loss and they would recommend admission for MRI/MRV to eval for other potential causes - tumor, CVST, etc

 

[RoyBasch]

I’ve had ophtho tell me (as a resident) that they think any visual changes/loss should be seen pretty emergently due to risk of vision loss and they would recommend admission for MRI/MRV to eval for other potential causes - tumor, CVST, etc

Agree, any visual changes and the disease is now in the urgency/emergency zone. That being said, I rarely do the LP as it's not really a 'right this second' emergency but I would usually phone-consult neurology, admit, and arrange for MRI/MRV and IR guided LP to be done inpatient that day or the next.

I would not start diamox prior to confirmatory LP as this could confound the diagnosis.

There is a softer call for "intractable headache" to admit and do the same. May depend on how attainable is prompt neuro follow up.

Everyone else who just has a headache, gets better, but the diagnosis is considered (longer standing symptoms, headache is worse in the morning or with bending over, obese woman in her 20s-30s, etc.) gets neuro follow up.

This is coming from the community btw. Academic may be different with lower threshold to do the LP and do everything (and 8 hour length of stay in the ER). I find in the community the threshold to LP is fairly high, and reserved mostly for suspected cases of bacterial meningitis.

 

[carbonizedeyesockets]

I’ve had ophtho tell me (as a resident) that they think any visual changes/loss should be seen pretty emergently due to risk of vision loss and they would recommend admission for MRI/MRV to eval for other potential causes - tumor, CVST, etc

This has been my experience. I've been told the same by neuro - vision changes make this emergent due to risk of permanent vision loss. However, I do NOT do the LP in the ED. Admit for MRI, LP by IR, and start acetazolamide if recommended by neuro.

 

[thegenius]

This has been my experience. I've been told the same by neuro - vision changes make this emergent due to risk of permanent vision loss. However, I do NOT do the LP in the ED. Admit for MRI, LP by IR, and start acetazolamide if recommended by neuro.

yea that's fine as others have wrote. but it's not like your vision goes from 20/50 to 20/FC or 20/handwave in 3 days if not diagnosed properly.

In the pt I had above, she had some blurry vision and was shunted ~4 months later.

I would defend myself in a court of law LOL

 

[Porfirio]

I just had one of these recently, and I did the LP in the ED. Pressure was 30. I normally don’t do them. I felt bad for the patient and it was a holiday weekend (aka no Neurology or IR till 3 days from when she came). She felt a lot better afterwards.

 

[WilcoWorld]

I don't think it's necessary, but it can be quite satisfying. I had a patient with a convincing story who felt pretty lousy and I wasn't swamped - did the LP & couldn't measure the opening pressure accurately, because the CSF started to overflow out the top of the manometer! I took off several CC's and afterwards the patient felt MUCH better. Dc'd her with an acetazolamide Rx and the Neuro clinic number.

 

[Brigade4Radiant]

No LP for an ED doc it’s important to know when not to do a procedure especially one h That is time consuming. If it’s not sub arachnoid or meningitis. ICH LPs should be done by IR

 

[WilcoWorld]

No LP for an ED doc it’s important to know when not to do a procedure especially one h That is time consuming. If it’s not sub arachnoid or meningitis. ICH LPs should be done by IR

But sometimes you don't know if it's SAH/meningitis/IIH until you do the LP, so you might as well measure the opening pressure while you're getting your CSF sample.

 

[GeneralVeers]

But sometimes you don't know if it's SAH/meningitis/IIH until you do the LP, so you might as well measure the opening pressure while you're getting your CSF sample.

Well actually you do. The NNT for CT-negative SAH is so low, as to not be worth doing the procedure.

 

[WilcoWorld]

Well actually you do. The NNT for CT-negative SAH is so low, as to not be worth doing the procedure.

Not if you're > 6 hours after onset.

I almost never LP for SAH anymore, but I can imagine a scenario where I would.

 

[Mushdoc]

just skip to CTA, forget the LP

 

[southerndoc]

just skip to CTA, forget the LP

I tend to follow this philosophy, but there is a small percentage of aneurysms that are missed on CTA because of vasospasm. That and a negative non-con head CT makes the number missed extremely low.

 

[Mushdoc]

Real question: what would you do with a positive LP and a negative CTA? Get NSGY to do a real Angio? MRA? Just call it perimesencephalic?

 

[thegenius]

Real question: what would you do with a positive LP and a negative CTA? Get NSGY to do a real Angio? MRA? Just call it perimesencephalic?

nothing.
I would admit and let them figure it out.

In all seriousness though, I would probably call NSGY and then admit. I mean...what is there to do if you have an SAH without an aneurysm? Just watch them in the hospital to make sure they don't get worse.

 

[Brigade4Radiant]

But sometimes you don't know if it's SAH/meningitis/IIH until you do the LP, so you might as well measure the opening pressure while you're getting your CSF sample.

Even for meningitis antibiotics and steroids early is the most important thing and with these people you are going to admit.

For SAH you do a CTA lumbar can tell you if it’s bleeding. But if the tap is traumatic and even if it clears in the fourth tube having blood you are still going to call neruosurgery.

 

[GeneralVeers]

Prob a CTA to see if there's aneurysm. Honestly this should be done instead of LP anyway. Havnt done an LP for headache in 10 years.

 

[WilcoWorld]

Even for meningitis antibiotics and steroids early is the most important thing and with these people you are going to admit.

I am not suggesting that we should withhold antibiotics until we get an LP, but I do think you are taking better care of your patient by getting an LP as early as is feasible when you're worried about meningitis, and that means attempting an LP in the ED. Delaying the LP can make it significantly harder to take good care of the patient down the road (Kanegaye et al 2001).

 

[turkeyjerky]

Real question: what would you do with a positive LP and a negative CTA? Get NSGY to do a real Angio? MRA? Just call it perimesencephalic?

Admit to Nueuro ICU for monitoring and prn-BP control and usually a formal angio the next day. I've always had the understanding that CTA will miss a not-insignificant number of aneurysms in the setting of an acute SAH (either to due vasospasm, obscuration from blood or distal location).

I had a true CT-negative SAH a couple of years ago. Granted it wasn't a subtle presentation. Thunderclap onset during exercise, came in 2 or 3 days later w/ ill appearing with a continued headache and obvious meningismus on exam. Had something like 40k RBCs + xanthrochromia on LP. Guy actually thanked me after I did the LP.

Of course both the transfer center and paramedics acted like I must be a half-wit for transferring a patient with a negative head CT.

Back to the OP though. Unless there's objective visual findings, truly intractable headache (eg patient in distress) or a cranial nerve palsy, I just send these patients home with the phone number for neurology.

 

[RustedFox]

Admit to Nueuro ICU for monitoring and prn-BP control and usually a formal angio the next day. I've always had the understanding that CTA will miss a not-insignificant number of aneurysms in the setting of an acute SAH (either to due vasospasm, obscuration from blood or distal location).

I had a true CT-negative SAH a couple of years ago. Granted it wasn't a subtle presentation. Thunderclap onset during exercise, came in 2 or 3 days later w/ ill appearing with a continued headache and obvious meningismus on exam. Had something like 40k RBCs + xanthrochromia on LP. Guy actually thanked me after I did the LP.

Of course both the transfer center and paramedics acted like I must be a half-wit for transferring a patient with a negative head CT.

Back to the OP though. Unless there's objective visual findings, truly intractable headache (eg patient in distress) or a cranial nerve palsy, I just send these patients home with the phone number for neurology.

This is the clincher, here. CTA is only good for so long after the initial event (exactly how long; I don't know - I've read different things from different papers).

 

[southerndoc]

This is the clincher, here. CTA is only good for so long after the initial event (exactly how long; I don't know - I've read different things from different papers).

One of my former colleagues had a patient with a negative CT, negative CT angiogram, AND a negative LP (8 hours post headache) who later died of a subarachnoid hemorrhage. Keep in mind that I work in a very busy ER that is a comprehensive stroke center (with one of the busiest aneurysm repair centers in the country). So this is unlikely to happen to anyone in a smaller ER. She was torn up about the case. After all the workup, the guy was being discharged when his headache came back suddenly. He then had altered mental status. When she repeated the CT, he had a subarachnoid hemorrhage seen on CT. A repeat CTA showed the aneurysm (that had vasospasm'd down).

 

[Bougiebuster]

Here's a different Q -> how many of you get CTH in elder with vertigo that sounds peripheral in nature?
Are you boxing yourself into admitting for MRI if normal?

 

[southerndoc]

I read somewhere (maybe for board review?) that MRI misses 6% of cerebellar strokes in the first 24 hours.

 

[RustedFox]

If I think it's likely peripheral in an older adult, but have any doubt, then I just skip the CT and go straight to MRI in the ED as if negative it saves an admission. If I think it is probably central with objective deficits outside of the window for tPA/thrombectomy, then I admit for MRI and further evaluation.

You must work somewhere nice, with nice people in the radiology department.

 

[lymphocyte]

I read somewhere (maybe for board review?) that MRI misses 6% of cerebellar strokes in the first 24 hours.

Would you believe 15-20%? A good HiNTs exam is better than a MRI in excluding stroke. (Quoted from my absolute favourite paper on stroke in dizziness and vertigo!).

https://www.ahajournals.org/doi/full/10.1161/STROKEAHA.117.016979

 

[southerndoc]

Would you believe 15-20%? A good HiNTs exam is better than a MRI in excluding stroke. (Quoted from my absolute favourite paper on stroke in dizziness and vertigo!).

https://www.ahajournals.org/doi/full/10.1161/STROKEAHA.117.016979

Yea, but even neurologists can't do the exam properly. It takes a neuroophthalmologist (yes, they exist) to do them.

 

[Brigade4Radiant]

Yeah the thing with the Hints exam Is that if most providers can’t do them properly it’s not useful

 

[turkeyjerky]

Yea, but even neurologists can't do the exam properly. It takes a neuroophthalmologist (yes, they exist) to do them.

Plus, it's nearly impossible to identify the correct patient population in which to apply the exam.

That figure of 15-20% of MRIs being falsely negative is curious. It seems like Dr. Newman-Toker has a habit of citing his own work. I'm not saying it's inaccurate, but it would be nice to see external replication. Me thinks the good doctor might have an intellectual bias.

I also don't get why these guys always talk about TPA for these patients, considering that they weren't included in the relevant studies.

How about this. If a patient can walk, is in sinus rhythm, and has a normal CTA, is it really necessary to diagnose a posterior fossa CVA? If a delayed MRI shows a sub-centimeter infarct, does it make a sound?

 

[thegenius]

Yeah the thing with the Hints exam Is that if most providers can’t do them properly it’s not useful

The head impulse test is very hard to do on people with acute vertiginous symptoms, and you are likely to snap off grandmas' head too. Oops! Sorry daughter. Here...let me suture her head back on. It will just take me a few minutes.

 

[thegenius]

How about this. If a patient can walk, is in sinus rhythm, and has a normal CTA, is it really necessary to diagnose a posterior fossa CVA? If a delayed MRI shows a sub-centimeter infarct, does it make a sound?

Remember bears don't **** in the woods if nobody sees them taking a **** in the woods.

 

[RoyBasch]

just skip to CTA, forget the LP

The problem is there is a significant percentage of the population with asymptomatic aneurysms (1-2%), if someone comes in with "the worse headache of their life" and the CTA shows an aneurysm (despite negative non con HCT), now you kind of HAVE to tap them to prove it's not ruptured. Where I agree a neg CTA/HCT gives you a very low risk patient, you may actually wind up taping more patients if you do CTAs and start diagnosing asymptomatic aneurysms in headache patients.

I would argue the population with an asymptomatic aneurysm may be larger than the number of CT negative true SAHs. So doing the CTA may result it more overdiagnosis and overtreatment than doing a non con HCT and just deciding clinically to do or not do an LP.

 

[Rekt]

I read somewhere (maybe for board review?) that MRI misses 6% of cerebellar strokes in the first 24 hours.

Agree and agree it's likely higher. These cases are essentially just passing the hot potato. At least at my shop they get a MRI without and discharge. I feel these patients at least need a concomitant MRA head/neck but whatever.

 

[dadaddadaBATMAN]

The problem is there is a significant percentage of the population with asymptomatic aneurysms (1-2%), if someone comes in with "the worse headache of their life" and the CTA shows an aneurysm (despite negative non con HCT), now you kind of HAVE to tap them to prove it's not ruptured. Where I agree a neg CTA/HCT gives you a very low risk patient, you may actually wind up taping more patients if you do CTAs and start diagnosing asymptomatic aneurysms in headache patients.

I would argue the population with an asymptomatic aneurysm may be larger than the number of CT negative true SAHs. So doing the CTA may result it more overdiagnosis and overtreatment than doing a non con HCT and just deciding clinically to do or not do an LP.

In general when I’m doing a cta head on someone looking for sah I’m pretty worried about them or they have a relatively bad story.

Although I acknowledge there is a significant portion of people with a symptomatic aneurysms, if they’re part of the sub population that
1. presents to the ed with severe headache
2. Have a convincing enough story that I am ctaing

I’m feeling good about not ignoring that “incidental” aneurysm. Some of them might be bs, but I bet it’s more dangerous than the general aneurysm population. Maybe they’re expanding, or had a small sentinel bleed that terminated, etc. I sleep alright with this one.

That doesn’t mean that I’m happy about the folks that are ordering these for obvious migraines

 

[thegenius]

In general when I’m doing a cta head on someone looking for sah I’m pretty worried about them or they have a relatively bad story.

Although I acknowledge there is a significant portion of people with a symptomatic aneurysms, if they’re part of the sub population that
1. presents to the ed with severe headache
2. Have a convincing enough story that I am ctaing

I’m feeling good about not ignoring that “incidental” aneurysm. Some of them might be bs, but I bet it’s more dangerous than the general aneurysm population. Maybe they’re expanding, or had a small sentinel bleed that terminated, etc. I sleep alright with this one.

That doesn’t mean that I’m happy about the folks that are ordering these for obvious migraines

Agreed...

As a side note CT Head is one the most useless (e.g. low yield) CT's we do. It is really not particularly helpful for the vast majority of the non-traumatic headaches.

 

[sloh]

The head impulse test is very hard to do on people with acute vertiginous symptoms, and you are likely to snap off grandmas' head too. Oops! Sorry daughter. Here...let me suture her head back on. It will just take me a few minutes.

Chiropractors seem to be getting away with it

 

[thegenius]

Chiropractors seem to be getting away with it

Yea on 26 yo hot women.

Not on grandma! Snap!!!

 

[RoyBasch]

In general when I’m doing a cta head on someone looking for sah I’m pretty worried about them or they have a relatively bad story.

Although I acknowledge there is a significant portion of people with a symptomatic aneurysms, if they’re part of the sub population that
1. presents to the ed with severe headache
2. Have a convincing enough story that I am ctaing

I’m feeling good about not ignoring that “incidental” aneurysm. Some of them might be bs, but I bet it’s more dangerous than the general aneurysm population. Maybe they’re expanding, or had a small sentinel bleed that terminated, etc. I sleep alright with this one.

That doesn’t mean that I’m happy about the folks that are ordering these for obvious migraines

I agree with you.

In my post above I was kind of playing the devils advocate.

I in fact do order CTAs for headaches where SAH is considered. I feel kind of the same as you, if my suspicion is lowish, I can be satisfied with the neg HCT alone (and time since onset of sxs, <6 hrs vs. more maters to.)

If my suspicion for SAH is moderate to high (history, general truly uncomfortable appearance of the patient, patient >6 hrs since onset) I get the CTA, and if it's negative that moves the patient back to a lower risk category. If the CTA is positive, this is a patient I was probably considering LPing anyways and I do an LP. Does the CTA result it more LPs for me personally than if I hadn't done one at all after I discover an aneurysm in a patient presenting with headache, probably yes. If I had not done the CTA and never found an aneurysm that would probably be the end of it.

But at least I feel VERY good about the HCT neg, CTA neg patient population as far as ruling out SAH. Even if they are bleeding, if they have no aneurysm or AVM it's probably perimesencephalic and has a fairly benign natural history anyways. Making the dx will not really change the outcome.

I have still LP'd a few neg HCT and neg CTA people (VERY concerned based on appearance) and those have always been negative. As time goes on my threshold to LP these patients has gotten VERY high because they are objectively low risk, and my own personal experience, I have yet to have a positive.

As a side note, I have dx'd a few vertebral and carotid artery dissections whose only sign/symptoms was a severe headache, thus prompting the CTA.

 

[bravotwozero]

I agree with you.

In my post above I was kind of playing the devils advocate.

I in fact do order CTAs for headaches where SAH is considered. I feel kind of the same as you, if my suspicion is lowish, I can be satisfied with the neg HCT alone (and time since onset of sxs, <6 hrs vs. more maters to.)

If my suspicion for SAH is moderate to high (history, general truly uncomfortable appearance of the patient, patient >6 hrs since onset) I get the CTA, and if it's negative that moves the patient back to a lower risk category. If the CTA is positive, this is a patient I was probably considering LPing anyways and I do an LP. Does the CTA result it more LPs for me personally than if I hadn't done one at all after I discover an aneurysm in a patient presenting with headache, probably yes. If I had not done the CTA and never found an aneurysm that would probably be the end of it.

But at least I feel VERY good about the HCT neg, CTA neg patient population as far as ruling out SAH. Even if they are bleeding, if they have no aneurysm or AVM it's probably perimesencephalic and has a fairly benign natural history anyways. Making the dx will not really change the outcome.

I have still LP'd a few neg HCT and neg CTA people (VERY concerned based on appearance) and those have always been negative. As time goes on my threshold to LP these patients has gotten VERY high because they are objectively low risk, and my own personal experience, I have yet to have a positive.

As a side note, I have dx'd a few vertebral and carotid artery dissections whose only sign/symptoms was a severe headache, thus prompting the CTA.

Your post, like Southern Doc's only serve to underscore how hard this job can be.

 

[RoyBasch]

Your post, like Southern Doc's only serve to underscore how hard this job can be.

Sometimes I feel better about it knowing that it is objectively "hard" and that even with experience tough clinical decisions get easier but they never get easy. No matter how much experience you have, you can still make the wrong call just because there are no absolutes in this field, only probabilities. We see huge numbers of patients compared to other specialties, and by the law of large numbers you will see unexpected or improbable (bad) outcomes as time goes on just because the probabilities can get low based on sound judgement and experience, but never go to zero.

Patient can have a neg CTA and neg HCT with no good reason to do an LP that still has SAH. That being said LP'ing all these patients is the wrong call to as it is an invasive procedure with a low (improbable) but not zero rate of complication plus pain, expense, time, etc.

 

[turkeyjerky]

On a somewhat related note, I"m wondering what others do with a failed attempt at a low-yield LP. Let's say you have low suspicion for a dangerous process, but want to be able to document 'shared decision making' so offer the patient the LP. Unfortunately, they actually want it done, but, due to habitus or whatever, you can't get it. Do you automatically admit for IR, give them another chance to refuse?

 

[bravotwozero]

Admit and make the procedure someone else's problem. Usually that is IR. Once you've made the decision to attempt the LP, there's no backing out, unless the patient decides against it and wants to leave, then you make them go AMA.

 

[RoyBasch]

On a somewhat related note, I"m wondering what others do with a failed attempt at a low-yield LP. Let's say you have low suspicion for a dangerous process, but want to be able to document 'shared decision making' so offer the patient the LP. Unfortunately, they actually want it done, but, due to habitus or whatever, you can't get it. Do you automatically admit for IR, give them another chance to refuse?

Arrange for IR, usually requires admission (occasionally not though depending on the time of the day, the IR schedule, and the cycle of the moon). Once the decision is made the patient needs an LP, it's not logically consistent to not do it just because it was difficult.

Conceptually I hear what you are saying, risk is low, procedure is hard, so maybe just bail. However, if you have a bad outcome, the question will be:
"IF you were worried enough about the diagnosis (SAH, meningitis, etc.) to do an LP, why did you not assure it could be done and the diagnostic information obtained?"

Especially given now in the era of IR, where there are further options to enhance the procedural technique if standard measures fail. I suppose back in the day if it truly couldn't be done, the only other option would be someone with more "experienced hands" giving it a try (not sure who that would have been before IR, I cant see neurology as being a go-to person for a tough procedure).

 

[thegenius]

Arrange for IR, usually requires admission (occasionally not though depending on the time of the day, the IR schedule, and the cycle of the moon). Once the decision is made the patient needs an LP, it's not logically consistent to not do it just because it was difficult.

Conceptually I hear what you are saying, risk is low, procedure is hard, so maybe just bail. However, if you have a bad outcome, the question will be:
"IF you were worried enough about the diagnosis (SAH, meningitis, etc.) to do an LP, why did you not assure it could be done and the diagnostic information obtained?"

Especially given now in the era of IR, where there are further options to enhance the procedural technique if standard measures fail. I suppose back in the day if it truly couldn't be done, the only other option would be someone with more "experienced hands" giving it a try (not sure who that would have been before IR, I cant see neurology as being a go-to person for a tough procedure).

Yea...you are kind of in a pickle. You do your best. Say you have a patient with an acute severe headache. CT Head non-con negative. CTA shows a aneurysm. You try to do an LP for xanthochromia but you failed. At that point you admit (or transfer) the patient and you pretend that they have an SAH. So HOB > 30, reverse anticoagulation, BP control...q1hr neuro checks, there isn't much more to do. If neuro exam gets worse repeat CT Head to see if there is hydrocephalus or other changes. There wouldn't be much more to do if you did the LP and had xanthochromia. Whether you get the LP now or 6 hrs from now wouldn't make MUCH difference....as the ultimate question is how quickly does NSGY have to clip or coil an aneurysm if your symptom is HA without evidence of increased ICP? Probably not emergently (like right now). It would probably happen in 24-48 hours is my guess from what I remember from medical school and residency.

I suppose you could call IR to come in at 0300 if that is possible but we all know at community hospitals that's probably not gonna happen. I dunno.

 

[helodoc]

Ever since I started doing ultrasound assisted LPs, I haven't missed a single one. Really a game changer. Also way cut down on my bloody taps. I've found if you're admitting patients for a procedure that really should be in your scope of practice, you need to step up your game somehow

 

[RustedFox]

Ever since I started doing ultrasound assisted LPs, I haven't missed a single one. Really a game changer. Also way cut down on my bloody taps. I've found if you're admitting patients for a procedure that really should be in your scope of practice, you need to step up your game somehow

I like this; but if you're really ready for a challenge then come down here to the United States Capital of Old People and try it on these turbo calcified seniors with Grade 8 kyphoscoliosis.

I don't even bother anymore.