We’re Thinking About It Wrong: The CTA vs. LP Debate for Subarachnoid Hemorrhage The SimKit Podcast

A debate continues in the subarachnoid hemorrhage (SAH) work up on what is the best follow up test – CT angiogram (CTA) or lumbar puncture (LP)? This is the wrong question. Find out what question you should be asking and how to best answer it. 


Hello everybody and welcome back to the SIM Kit podcast. And we’re thinking about it wrong. The CTA versus LP debate for subarachnoid hemorrhage.

I can think of very few things in emergency medicine that are so split in our approach to the care of patients. Yes, we battle consultants, Glucagon for food impactions, Ng tubes for bowel obstructions, TNK for strokes. We have a difference of opinions with many consulting services, but there’s very few circumstances in which there is practice variation to such a degree as the workup and rule out for subarachnoid hemorrhage CTA versus LP’s. The only one I can think of that’s akin to that is maybe succinylcholine versus rocuronium for RSI.

So I want to talk today about why we are framing the conversation about CTA versus LP wrong and how we really should be thinking about So I want to frame this conversation around a hypothetical case. And as we do, we are going to throw in some pearls about the workup, treatment, diagnosis, etc. for subarachnoid hemorrhage. So let’s introduce you to Lisa. Lisa is a 48 year old female. She actually works in HR in your hospital. She has two kids.

She is incredibly nice and helpful, so immediately you gotta raise the red flags, right? Nice people tend to have bad diagnosis, but let’s blow that out of her mind for a moment. Lisa has a headache. She actually developed her headache, uh, she was a little reluctant to say after she had a heated argument with her significant other about finances. We’ve all been there before.

The pain was sudden in onset. She’s had prior headaches, maybe considered some of them migraines, usually after her menstrual cycle, but this is different. It reached a maximum intensity within a matter of minutes and was very intense in nature. She tried some Tylenol because she had some childcare responsibilities at home, didn’t really help her. About an hour or so later she called EMS, they evaluated her and wanted to transport, but she declined because she had these care responsibilities, needed to wait for her husband to come home.

So she gets her childcare all situated. She has a friend to drive her to the emergency department and comes in through the front door in the waiting room. Unfortunately, there is that transport bias. We try to triage based on chief complaint and all that, but she spends a good number of hours in the emergency department waiting room waiting to be seen. She’s eventually seen by one of the residents who presents the case to you, 48 year old female with a sudden onset of headache after an argument with her significant other. This astute resident has the differential diagnosis of #1 subarachnoid hemorrhage, #2 subarachnoid hemorrhage, #3 subarachnoid hemorrhage, and so you guys move forward with CT imaging. So you guys order the CT you are aware of the potential diagnosis. Unfortunately, between her childcare responsibilities, her waiting for transport with her friend, her getting to the emergency department and then waiting in the waiting room, she has had a multiple hour delay in getting her imaging done. You try to push her through the CT as quickly as you can. Again, it’s an incredibly busy day. There is some time delay in doing that and she has her scan at the 9 hour mark. The non contrast head CT is indeed Negative. Now just as a stopping point here, we do want to review quickly the six-hour rule which many of us in emergency medicine are familiar with. If a CT scan is ordered for rule out subarachnoid hemorrhage and is completed within six hours with a competent radiologist, a modern generation scanner and some patient characteristics which are important like their hemoglobin which will be included in the show notes if that imaging is negative, that can be considered a rule out test for subarachnoid hemorrhage much of the literature comes from.

Harry, there’s also the Dubash meta-analysis in 2016 that showed that this is about or greater than 98% sensitive. The confidence intervals on the Dubosh menta-analysis was 97.1 to 99.4, and you are backed up as well by an ACEP clinical policy in 2019. Unfortunately, once we get past the six hour mark, there is a little less clarity about the value of CT.

Generally speaking I like to think of about an 85% sensitivity, but obviously that’s going to be variable by time of presentation and time from scan. You can think that within the 1st 24 to 48 hours a non-contrast head CT is probably greater than 90% sensitive for the diagnosis of subarachnoid hemorrhage.

For the sake of argument here, let’s think about Lisa herself. She resents with a sudden, severe, different than typical headache that is concerning for subarachnoid hemorrhage.

We have a non-contrast head CT that was completed in greater than six hours. We said nine. It could be 10, it could be 12. The question here is in a rule out subarachnoid hemorrhage patient with a negative non-contrast head CT at greater than six hours, what do you do? So tape up your knuckles and put it on your boxing gloves, because you’re about to dive into that LP versus CTA debate for subarachnoid hemorrhage.

Let’s talk about lumbar puncture first. This is the tried and true, question mark? In my opinion. When we’re doing these lumbar punctures, we are again looking for RBCS or Xanthochromia, which is the yellow discoloration of the CSF with RBC lysis. Interestingly, there’s not great information to say when exactly xanthochromia forms. There have been arguments in the past about waiting for the lumbar puncture to when we know Xanthochromia should be present, but that leaves this patient in a limbo state with a potentially serious or life-threatening diagnosis. And you should not be delaying your lumbar puncture if that’s the test that you choose to do. But you should be looking for either RBCS or Xanthochromia. As another aside, an important and interesting question is what is your RBC threshold for the diagnosis of subarachnoid hemorrhage versus traumatic tap? Now interesting, the numbers cited for cup points here are actually all over the board. My particular favorite and I think the most well-done study on the topic matter was a Perry study in BMJ in 2015. In this study, they had 1739 patients, all of whom got lumbar puncture to rule out subarachnoid hemorrhage.

They found that a cut point of 2000 x 106 RBCS in addition to nose xanthochromia excluded aneurysmal subarachnoid hemorrhage with a sensitivity of 100% specificity of 91, but was very effective at excluding the vast majority of traumatic taps. So for me that number is 2000 x 106  RBCS.

But coming back to the CTA lumbar puncture debate for subarachnoid hemorrhage, let’s talk about the pros and cons of lumbar puncture. First, the pros. Of course, there is no additional radiation to doing a lumbar puncture. There’s also the potential benefit for some alternative diagnosis, most importantly meningitis. Bacterial meningitis of course, being an important one for us to diagnose. Now, does bacterial meningitis commonly present with a sudden severe headache without other red flags like fever? No, but it is a potential cause for a Thunder clap style headache. The cons of lumbar puncture, of course, include the time commitment of the provider to complete it, the post lumbar puncture headache, and a potential false positive rate with traumatic taps. Collectively, it can be said that a non-contrast head CT plus a lumbar puncture has a sensitivity of greater than 98% for ruling out subarachnoid hemorrhage.
Now, people in the lumbar puncture camp will often cite a Perry study that was done in 2008 on the topic matter. That showed that it had 100% sensitivity of a non-contrast sensitivity plus a lumbar puncture. Interestingly, they had 5000 x 106 as their cup point, but this study predated the one looking at traumatic tap rates. In this 2008 study in the Annals of Emergency Medicine, they did find a sensitivity of 100% of CT then lumbar puncture, which is fantastic.

What is often overlooked in this study though is the nearly 30% false positive rate with this approach. Now, is the traumatic tap rate truly 30% in all comers for lumbar puncture? No, I don’t think so. But it there is a rate and it must be considered in the pros and cons debate.

Now what about CTA? This is the new-ish, kind of newer kid on the block. Pros and cons. The pros for CTA is that is noninvasive, it is fast, and it also can consider alternative diagnosis.
The cons here are the radiation dose exposure, the incidental aneurysm recognition, and that it misses smaller aneurysms, usually ones that are less than 3 millimeters in size, which of course could potentially be culprit. But we know that larger aneurysms are more likely to rupture.

Now, coming back to that incidental aneurysm rate, this is a very important potential con for CTA. Incidental aneurysms that have not had anything to do with the patient’s headache can be recognized on CTA. Depending on where you live in the world, what is that rate of incidental aneurysms in the general population? You can see that it’s about 2 to 5%. Depending again, Asian populations may have a higher incidence than other areas, but let’s say 5%, nice round easy number, 5% incidental aneurysm rate. Now, the CT-CTA route also has fantastic test characteristics for ruling out aneurysmal subarachnoid hemorrhage. Studies on it have shown that sensitivity to be 97.3 to 99.2 and it is an ACEP approved approach for ruling out subarachnoid hemorrhage.

Now this is all well and good, and this is information that addresses the question of “what is the next appropriate test for rule out subarachnoid hemorrhage in a patient with a negative non-contrast head CT that was completed at greater than six hours?”

I argue, though, that we are thinking about this wrong. This is not the approach that we should be having, and we have a much more important question to ask. That question is “what does this patient have?” What does Lisa, in our circumstance have, or what is she suffering from? To answer this question, we have to think about the differential diagnosis for causes of Thunderclap headache. The number one, two and three as our astute resident says is subarachnoid hemorrhage, but the list does not end there. We have to include intracerebral hemorrhage, cervical artery dissection, cerebral venous sinus thrombosis, and reversible cerebral vasoconstrictive syndrome.

Interestingly, when we look at all-comers, people that present with a thunderclap headache, what would you say is a percentage of those that are eventually diagnosed with subarachnoid hemorrhage? Thundercloud headache presentation, final diagnosis of subarachnoid hemorrhage. What’s your guess for that percentage?

It is 11 to 25%.

That leaves a good portion of our population as not having that final diagnosis.

And Dubosh, the one who has that meta-analysis proving the less than six-hour rule has this to say. And I quote first. “Our analysis refers only to subarachnoid hemorrhage, Thunder clap headache has a differential diagnosis. If the clinical presentation or epidemiological context suggest another non subarachnoid hemorrhage diagnosis, further testing beyond CT may be indicated.”

So I asked that we all reframe our thinking from “what is the next best test to rule out subarachnoid hemorrhage?” To “what is the best tests in the workup of thunderclap headache?”

In our hypothetical Lisa patient, she gets her noncom head CT, which comes back as negative. Her symptoms have been abating with some therapy and while the resident and yourself the attending are debating the next test you want to do in the rule out process, the nurse calls you and says that Lisa is in severe pain again. She has a severe onset of similar symptoms that were at its maximum when she initially had that argument and you go and see her and she’s in excruciating pain. At that point, you decide to move forward with a CT-angiogram and you see a “bead on a string” patterning consistent with RCVS, reversible cerebral vasoconstrictive syndrome.

Now, many of us may not be as familiar with RCVS as we would like to be or as we should be, because if you look at that diagnosis, reversible cerebral vasoconstrictive syndrome, it is the second most common cause of thunderclap headache that is not subarachnoid hemorrhage. Now, interestingly, true incidence of the disease is yet to be determined, but it is recognized as a very prominent cause for this type of headache. The name itself was only coined in 2007, which is part of the reason why we don’t have great data about as true incidence and epidemiology.

Now, at least half of the cases of RCVS are secondary to a vasoactive substance like illicit drugs, sympathomimetics or serotonergic drugs, and it can also happen in the postpartum period.
Now, while the radiologist will recognize this as likely a benign disease, neurologists may differ in their opinion slightly because there is an association with both hemorrhage and ischemic stroke in these patients, so recognition of this disease process is very important.

So going back a few steps and thinking about headache and subarachnoid hemorrhage in the emergency department, we have some numbers to crunch. Headache is a very common chief complaint. It represents about 2% of all Ed presentations. Of those presenting with headache, about 1% will be eventually diagnosed with subarachnoid hemorrhage. And as we mentioned, about 11 to 25% of thunderclap headaches will carry the final diagnosis of subarachnoid hemorrhage, but the conversation does not stop there. As we mentioned, Thunder clap headache carries a differential diagnosis, subarachnoid #1 #2 and #3, intracerebral hemorrhage- we’re going to recognize that on our non-contrast head CT that can be taken off the conversation-cervical artery dissection, cerebral venous sinus thrombosis and reversible cerebral vasoconstrictive syndrome.

Now, in crunching some of the numbers on the incidence of these alternative diagnosis, subarachnoid hemorrhage has an incidence of about 6 to 8 per 100,000. We’ll have a table in the show notes showing some of the incidents for these other diagnosis, but if you add them together, we start to get pretty darn close to that 6 to 8 per 100,000 we see for subarachnoid hemorrhage. When looking at these disease processes, we recognize CTA as beneficial in ruling them out and LP having very low if any utility.

So, to wrap this all up-Lumbar puncture and CTA both have great test characteristics for subarachnoid hemorrhage. I cannot fault you on either side of the fence when you are choosing lumbar puncture or CT angiogram. Her rule out subarachnoid. They both have fantastic test characteristics. They both have great data behind them and societal support for their use in subarachnoid hemorrhage. But the differential diagnosis for thunderclap headache is probably broader than you think.

But CTA is a better test for these alternative diagnoses. When you choose to ask the question “what does this patient with thunderclap headache have and how do I assess for these diagnoses?” – CTA is by far the better test.

I am sure there is some controversy in this. I’m sure people are going to have their strong opinions one way or the other. There are great references in the show notes both to the papers that I cited today and other background reading for you. If you have contrary arguments, I would love to hear them. I would love to have an educated discussion about this. You can reach out to us through the comments section or find us on Twitter @SimKitCo. Thanks so much for listening.

The 6 Hour Rule

There is strong data to support a non-contrast head CT as a stand-alone test to rule out SAH if completed within 6 hours, with the following test characteristics.[1][2]

CT Within 6 Hours for SAH Requirements

Patient Characteristics
Thunderclap headache concerning for SAH, isolated
No loss of consciousness, seizures, or neck pain
Hematocrit >30%
Imaging Characteristics
Rule out SAH or thunderclap HA listed in the indication
3ʳᵈ generation or newer CT scan
Neuroradiologist or General Radiologist who routinely reports CT head imaging

After 6 hours

After 6 hours, no census exists on the test characteristics for non-con head CT. More recent literature is pushing the sensitivity higher and higher as technology (CT imaging) improves. One study by Suarez et al. [3] notes a 93% sensitivity up to 24 hours, but this is not a widely accepted number and given the seriousness of the diagnosis considered, a 7% miss rate is not acceptable. Hence the need for additional work up after non-contrast head CT.

The Next Test

After non-contrast head CT, the two options for work up SAH include LP and CTA. Let’s think about these tests and their pros and cons individually.

Lumbar Puncture

No RadiationTime Intensive
Alternative DiagnosesPost-LP headache
False & with traumatic taps

The CT-LP combination approach has a >98% sensitivity for SAH and is a very reasonable approach in the rule-out SAH algorithm.

CT Angiogram

FastIncidental Aneurysms[6][7]
Alternative DiagnosesMisses small (<3mm) Aneurysms

The CT-CTA combination approach also has great test characteristics with a sensitivity >98% and is supported by an ACEP clinical policy[8].

The Real Question

These approached for rule-out SAH are both acceptable and literature supported. But they answer the question “what is the next appropriate test for rule out subarachnoid hemorrhage in a patient with a negative non-contrast head CT that was completed at greater than six hours?” This is not the question we should be asking ourselves. It is simply “What does this patient have?”

This brings up the differential diagnosis for thunderclap headache. Which, for the ED provider, should be:

  1. Subarachnoid Hemorrhage
  2. Subarachnoid Hemorrhage
  3. Subarachnoid Hemorrhage
  4. Intracerebral Hemorrhage (ICH)
  5. Cervical Artery Dissection (CeAD)
  6. Cerebral Venous Sinus Thrombosis (CVT)
  7. Reversible Cerebral Vasoconstrictive Syndrome (RCVS)

When we look at all-comers with a thunderclap headache, only 11-25% will be diagnosed with SAH[9]. This leaves the vast majority without that diagnosis, and a differential that requires further evaluation. Interestingly, RCVS is recognized as the second most common cause of thunderclap headache.

At least half of the cases of RCVS are secondary to a vasoactive substance like illicit drugs, sympathomimetics or serotonergic drugs, and it can also happen in the postpartum period.
Now, while the radiologist will recognize this as likely a benign disease, neurologists may differ in their opinion slightly because there is an association with both hemorrhage and ischemic stroke in these patients, so recognition of this disease process is very important.

Numbers to Crunch

Headache is a very common chief complaint in the Emergency Department. In looking at this chief complaint we have some numbers to crunch:

  1. 2% of Emergency Department presentations will be for headache
  2. 1% of these presentations will be diagnosed with SAH
  3. 11-25% of thunderclap headaches will be a SA

When we look at the incidence of the diseases that cause thunderclap HA, when we start to tally the likelihood of the some of the other disease processes (even without true incidence for RCVS), we start to get rather close to the incidence of SAH its. Moreover, when we compare the utility of LP and CTA for these alternatives, it is clear CTA has a greater utility.

Note: It is important to recognize that the test of choice for CVT is not a traditional CTA, but a CT venogram. The point still stands that, in the right clinical context, a CT with some form of contrast is of greater utility than LP.

Average Age (years)Incidence Rate (per 100k)LPCTA


  • LP and CTA both have fantastic test characteristics for rule-out SAH
  • The Differential Diagnosis for thunderclap headache is probably broader than you think
  • When you choose to ask the question “what does this patient with thunderclap headache have and how do I assess for these diagnoses?” – CTA is by far the better test.

With so much to know and do in Emergency Medicine, you shouldn’t have to worry about your procedural competency… but you need to. Are you really ready to place that Blakemore or do a cric? Let SimKit help you bring your skills up to snuff in a fun and engaging way with simulation that delivers right to your door. Practice at home, at work, in the office and on your time.

References & Further Readings

[1] Dubosh NM, Bellolio MF, Rabinstein AA, Edlow JA. Sensitivity of Early Brain Computed Tomography to Exclude Aneurysmal Subarachnoid Hemorrhage: A Systematic Review and Meta-Analysis. Stroke. 2016 Mar;47(3):750-5. doi: 10.1161/STROKEAHA.115.011386. [pubmed]

[2] Perry JJ, Stiell IG, Sivilotti ML, Bullard MJ, Emond M, Symington C, Sutherland J, Worster A, Hohl C, Lee JS, Eisenhauer MA, Mortensen M, Mackey D, Pauls M, Lesiuk H, Wells GA. Sensitivity of computed tomography performed within six hours of onset of headache for diagnosis of subarachnoid haemorrhage: prospective cohort study. BMJ. 2011 Jul 18;343:d4277. doi: 10.1136/bmj.d4277. [pubmed]

[3] Suarez JI, Tarr RW, Selman WR. Aneurysmal subarachnoid hemorrhage. N Engl J Med. 2006 Jan 26;354(4):387-96. doi: 10.1056/NEJMra052732. [pubmed]

[4] Perry JJ, Alyahya B, Sivilotti ML, Bullard MJ, Émond M, Sutherland J, Worster A, Hohl C, Lee JS, Eisenhauer MA, Pauls M, Lesiuk H, Wells GA, Stiell IG. Differentiation between traumatic tap and aneurysmal subarachnoid hemorrhage: prospective cohort study. BMJ. 2015 Feb 18;350:h568. doi: 10.1136/bmj.h568. [pubmed]

[5] Perry JJ, Spacek A, Forbes M, Wells GA, Mortensen M, Symington C, Fortin N, Stiell IG. Is the combination of negative computed tomography result and negative lumbar puncture result sufficient to rule out subarachnoid hemorrhage? Ann Emerg Med. 2008 Jun;51(6):707-13. doi: 10.1016/j.annemergmed.2007.10.025. [pubmed]

[6] Brisman JL, Song JK, Newell DW. Cerebral aneurysms. N Engl J Med. 2006 Aug 31;355(9):928-39. doi: 10.1056/NEJMra052760. [pubmed]

[7] International Study of Unruptured Intracranial Aneurysms Investigators. Unruptured intracranial aneurysms–risk of rupture and risks of surgical intervention. N Engl J Med. 1998 Dec 10;339(24):1725-33. doi: 10.1056/NEJM199812103392401. [pubmed]

[8] American College of Emergency Physicians Clinical Policies Subcommittee (Writing Committee) on Acute Headache:; Godwin SA, Cherkas DS, Panagos PD, Shih RD, Byyny R, Wolf SJ. Clinical Policy: Critical Issues in the Evaluation and Management of Adult Patients Presenting to the Emergency Department With Acute Headache. Ann Emerg Med. 2019 Oct;74(4):e41-e74. doi: 10.1016/j.annemergmed.2019.07.009. PMID: 31543134. [pubmed]

[9] Ducros A, Bousser MG. Thunderclap headache. BMJ. 2013 Jan 8;346:e8557. doi: 10.1136/bmj.e8557. [pubmed]

[10] Connolly ES Jr, Rabinstein AA, Carhuapoma JR, Derdeyn CP, Dion J, Higashida RT, Hoh BL, Kirkness CJ, Naidech AM, Ogilvy CS, Patel AB, Thompson BG, Vespa P; American Heart Association Stroke Council; Council on Cardiovascular Radiology and Intervention; Council on Cardiovascular Nursing; Council on Cardiovascular Surgery and Anesthesia; Council on Clinical Cardiology. Guidelines for the management of aneurysmal subarachnoid hemorrhage: a guideline for healthcare professionals from the American Heart Association/american Stroke Association. Stroke. 2012 Jun;43(6):1711-37. doi: 10.1161/STR.0b013e3182587839. [pubmed]

[11] Mark DG, Hung YY, Offerman SR, Rauchwerger AS, Reed ME, Chettipally U, Vinson DR, Ballard DW; Kaiser Permanente CREST Network Investigators. Nontraumatic subarachnoid hemorrhage in the setting of negative cranial computed tomography results: external validation of a clinical and imaging prediction rule. Ann Emerg Med. 2013 Jul;62(1):1-10.e1. doi: 10.1016/j.annemergmed.2012.09.003. [pubmed]

[12] Meurer WJ, Walsh B, Vilke GM, Coyne CJ. Clinical Guidelines for the Emergency Department Evaluation of Subarachnoid Hemorrhage. J Emerg Med. 2016 Apr;50(4):696-701. doi: 10.1016/j.jemermed.2015.07.048. [pubmed]

[13] Marcolini E, Hine J. Approach to the Diagnosis and Management of Subarachnoid Hemorrhage. West J Emerg Med. 2019 Mar;20(2):203-211. doi: 10.5811/westjem.2019.1.37352. [pubmed]

[14] Hine Jason, Marcolini Evadne, Ashenburg Nick. Aneurysmal Subarachnoid Hemorrhage. In: Mattu A and Swadron S, ed. CorePendium. Burbank, CA: CorePendium, LLC. https://www.emrap.org/corependium/chapter/recTI59VW0TPBpesx/Aneurysmal-Subarachnoid-Hemorrhage. Updated September 17, 2020. Accessed March 7, 2023. [website]

[15] Ashenburg Nick, Marcolini Evadne, Hine Jason. Approach to Headache. In: Mattu A and Swadron S, ed. CorePendium. Burbank, CA: CorePendium, LLC. https://www.emrap.org/corependium/chapter/rec8eRzSrPsVUiMuV/Approach-to-Headache. Updated October 8, 2022. Accessed March 7, 2023. [website]

[16] Marcolini Evadne, Akhter Murtaza. Reversible Cerebral Vasoconstriction Syndrome (RCVS). In: Mattu A and Swadron S, ed. CorePendium. Burbank, CA: CorePendium, LLC. https://www.emrap.org/corependium/chapter/recXhhb5dfxywqyY5/Reversible-Cerebral-Vasoconstriction-Syndrome-RCVS. Updated December 21, 2022. Accessed March 8, 2023. [website]

Cover image source: https://commons.wikimedia.org/wiki/File:Computed_tomography_of_human_brain_%2821%29.png Department of Radiology, Uppsala University Hospital. Uploaded by Mikael Häggström. Uploaded with edits.

Like this? Share it:

Subscribe to the Podcast on

Subscribe to our mailing list

* indicates required

Connect on Social