The Subarachnoid Enigma

Six hours means so much when it comes to subarachnoid hemorrhage. That is the inflection point when blood may no longer look like fresh blood on a CT. Several studies have shown that a negative ct done within 6 hours of headache onset effectively excludes clinically significant bleeding (extrapolated to mean aneurysmal subarachnoid hemorrhage). Of course there are other things that can cause severe acute headache in an otherwise healthy person: vascular dissection, dural sinus thrombosis to name a few. So it’s not all CT-LP (lumbar puncture), but focusing on the question of “LP yes or no” after a negative third generation (or higher) CT,  the evidence suggests that LP may not be mandatory. Can there still be bleeding? Yes, there can. But bleeding from a source that’s going to kill is unlikely. This is a time for shared decision making with your patient.

CT may miss a small number of bleeds that LP will find, but there is also the issue of a false positive tap being more likely than a bleed and the downstream effects of testing and aneurysm treatment.

Dr. Ran Ran joins the show to examine the literature on The Subarachnoid Enigma

Direct Download

Upcoming 2015 conferences

June: SMACC Chicago

October: Essentials of Emergency Medicine in Las Vegas

December: Atlantis CME Retreat

References for this Podcast

Blok, Katelijn M., et al. “CT within 6 hours of headache onset to rule out subarachnoid hemorrhage in nonacademic hospitals.” Neurology (2015): 10-1212.

Perry, Jeffrey J., et al. “Differentiation between traumatic tap and aneurysmal subarachnoid hemorrhage: prospective cohort study.” bmj 350 (2015): h568.

Perry, Jeffrey J., et al. “Sensitivity of computed tomography performed within six hours of onset of headache for diagnosis of subarachnoid haemorrhage: prospective cohort study.” Bmj 343 (2011): d4277. 

Claveau, David, and Jerrald Dankoff. “Is lumbar puncture still needed in suspected subarachnoid hemorrhage after a negative head computed tomographic scan?.” CJEM 16.3 (2014): 226-228.


  1. Derek Isenberg

    There is more info on this topic from Dave Newman. However, it is not correct to only use the sensitivity of CT for SAH but you must use the negative LR. If the prevalence of SAH is 8%, using the low end of sensitivity and specificity I calculate a negative LR of .11, giving you a post-test probability of about 1%. This assumes all of the SAH were anurysmal and therefore needed to be treated. Since some of the SAH will be peri-mesenceophalic, the post-test probability is less than 1%.

    1. Post
  2. Joe

    This is a can’t miss diagnosis. You will lose everything if you miss it because you followed the negative head CT within six hours of onset all the way to discharge. That being said, I just left a community ER where LP if suspected SAH was on my differential was done following a normal head CT. Now I am at a non academic tertiary referral center where CT perfusion studies of the brain are done if this is suspected and is requested of us in the ER by neurology instead of the LP. I am not so comfortable yet with this idea so when it comes down to it I discuss it with the patient, the LP or CT perfusion (which one do you think they pick) and document the hell out of it. If someone is willing to dedicate a decade of study on this to adequately power a trial to show CT perfusion is equivalent to LP, then I would be excited to change my practice. But, the trials have not been good enough to change my practice yet. I make a crap cup of coffee and hate hipsters, so I am not losing my license to end up working at Starbucks over a procedure I am good at.

  3. srrezaie

    Hey Rob,
    Great podcast and really a huge dilemma for everyone practicing EM. Lets say for the purpose of this discussion we are keeping our decision making binary (i.e. SAH vs no SAH)….I realize that clinically this is not how patients or the real world work, but for the purpose of this response. My practice is as follows:

    Headache 6 hrs

    Headache 99% NPV with an LP? LP is an invasive procedure that has potential complications of making the current headache worse, infection, discomfort, as well as a high false positive rate. I then always document this conversation in the chart.

    Headache 12Hrs: Neg head CT, the next step is LP. If +RBCs and neg xanthochromia then we are done and no SAH, if +RBCs and pos xanthochromia then there is or was a SAH.

    Now for the frustrating 6 -- 12 hour headache: So again they get a head CT and if it is neg the next step for me would be LP, but some would advocate for CTA which is another ball of wax all together. I like things simple so for me its LP. If LP has +RBCs, regardless of decreasing number and neg xanthochromia I think the miss rate of clinically relevant SAH is too high, I take a more conservative approach and consult neurosurgery or transfer to a facility that has neurosurgery.

    Would love to hear what people do for the frustrating 6 -- 12 hour mark….I am not sure there is a perfect answer here, but would love to hear others thoughts.


  4. Yone Amuka

    The literature on this stuff seems to be tainted with a bit of wishful thinking. We keep WANTING the new generation CT scanners to be sufficient to rule out CT, and we keep WANTING to extend the hours since onset, during which we can trust the negative CT read (I remember it being 12-48 hours when I was leaving residency; while 6 seems to be generally accepted now). While I’m pleased about the study that was done in Canada (the one mentioned in the podcast), it seems to me what we need is a study large/high powered/definitive enough so that we can once and for all update our ACEP guidelines. After all, how strongly can the information in this podcast be relied upon if our official academic society hasn’t yet put the seal of approval on it?

  5. MP

    Great podcast. Since reviewing this literature a year ago I started using this technique to send people home sooner and without LPs. I have one caveat. The radiologist reading these CTs has to be good at it. I had one case a few months ago where I got the CT at 5.5 hours and the contract out of state night hawk radiologist read it as normal. The patient was still in a lot of pain and looked really sick even after the therapies so I did a CTA and then an LP. The nighthawk read the CTA as normal. The tap was 100K rbcs both tubes one and four (after letting it drip first). The Neurosurgeons pointed to the normal reads and refused to even see the pt. An hour later our regular radiologist overread the CT as SAH and the CTA as positive aneurysm. The NSG team sheepishly came down and clipped the pt. This has taught me not to trust these kinds of rules to those who are OK but not excellent at doing them.

    1. Post
      Rob Orman

      This is one of the challenges in fully committing to this idea. In the most recent study, a similar thing happened (to your situation) where a subarachnoid was missed but later seen by the independent reviewers.

  6. Hector Singson (@D0CT0REY)

    SAH with a negative initial CT, if no LP contemplated, at least give the patient the benefit of in-hospital neuro-observation and a repeat CT if and when situation calls for it. Maybe a challenge in terms of convincing different services in as far as referrals and admission, but sticking to your guns and gestalt and index of suspicion may be the one that will save the patient’s life and your life as well. This is a real challenge for us because tolerance for a missed diagnosis is almost zero. Great Podcast, you nailed it though. I hope that the journals stated herein are Open Access.

  7. Daniel McKinney

    Great Podcast Rob. What do our neurosurgical friends think of this practice? What is the role of CTA? If we have a positive tap or CT, but a negative CTA, is there any surgical intervention?

    1. Daniel McKinney

      I overlooked MP’s post, so my post is some what redundant. It still raises the question, if there is a negative CTA, what is there to do?
      Is there a study where negative CTs have been followed by CTAs as oppose to LPs? With so many of our patients on anticoagulants these days, I think this question needs to be answered.

      1. Ran

        There are no studies powered to answer this question. But I think the answer lies in what is done when there is a positive CT or LP for SAH. The next step for most neurosurgeons is a digital subtraction angiography, which is the gold standard for vessel imaging. If DSA is negative, it is repeated in 1-2 weeks.

        Wait a minute: why repeat a gold standard test?

        In the context of an acute bleed, vessel imaging tests become unreliable because there can be vasospasm of the aneurysm to sizes that cannot be detected by DSA, much less CTA. Some textbooks cite 22% false negative rate for DSA in context of acute bleed but I can’t find studies. For this reason, DSA is repeated 1-2 weeks later preferably when the patient is without headache and thus, presumed to no longer have vasospasm.

        Thus if DSA cannot exclude aneurysm in patients with WOHL, I do not think CTA can be relied upon either.

  8. Sid W (@brawny dirigible)

    I have been in practice almost 15 years and I recently started using the six hour rule and I’m quite comfortable with it. Obviously there will be outliers but I trust my clinical judgment will steer me off strict inherence to the rule.

    I do have to say though that last bit about a patient having six red cells on their LP and being in the 6 to 12 hour range could be a subarachnoid hemorrhage seems dubious. Recent presenters that I have heard including Billy Mallon, q I do have to say though that last bit about a patient having six red cells on their LP and being in the 6 to 12 hour range could be a subarachnoid hemorrhage seems dubious. Recent presenters that I have heard including Billy Mallon have used cut offs of 200. It would be a pretty hard sell to a consultant that a patient with six red cells has a subarachnoid hemorrhage and is not the result of the traumatic tap.

    Criticism aside I very much enjoyed the presentation. I have had just 2 negative ct/positive LPs in my career but I’m sure I have given post LP headaches to countless patients as I strive for zero misses. The six hour rule is a welcome addition to my practice.

  9. mjasumback

    I tend to discuss this with my patients. Basically quote the Perry article, advise them of the risks/benefits of LP/no LP. and have them decide. In my practice SAH is a transfer out as no one local does aneurysm anymore. Thus, this is a relatively high stakes game, independent of the risk of SAH complications.

    I would suggest, though, that we have become slaves of the medicolegal industry. Reading Joe and YoneAmuka’s comments is telling. We have been forced into believing that this is a “CAN’T MISS” diagnosis and have harmed many more patient’s than we have helped by slavishly doing CT/LP. Listen to Newman’s podcast on this.

    1. Aneurysmal SAH is rare (8% in the highest risk cohort, the Perry study)
    2. CT is extraordinarily good in the 12 hrs same as 6-12 (ct for other reasons and not EBM, more CYA/Pt Sat)

    My practice for traumatic tap or positive LP is CTA. I am told, that a negative CTA after the above, is good enough to DC pt home. Whether true or not, my local Neurosurgeons and consultant neurosurgeons at tertiary centers have both said this.

    When I was first in practice, in the days of the 4 vessel angio, I was told that a negative 4v angio= go home and get a repeat in 3 mos to rule out thrombosed aneurysm.

    All of this falls apart when you have the dreaded patient:
    Bad HA, CT neg, LP positive for RBCs who is one of the 5% of pts with an asymptomatic aneurysm.

    All this for a rare disease, because the lawyers (and Stuart Swadron) have agreed that this is “CAN’T MISS”

    Mike Jasumback

  10. Scott C

    Great podcast guys. Gotta agree with Sid though -6 RBCs is a negative tap. I don’t keep patients for > 12 hours waiting for xanthrochromia, a crappy test and an unnecessary admission. I tap and if I get a clean csf ( < 100-200 rbcs). -- I am done. Blood in csf needs further work up. My trick for getting clear taps is to use my lignocaine needle to suck out lp needle hub prior to collecting csf…

  11. Pingback: Post Lumbar Puncture Headaches - R.E.B.E.L. EM - Emergency Medicine Blog

Awesome article, I know - please share your erudite thoughts...