Neurology & Psychiatry (May 2015)
Review: Edlow et al. Aneurysmal Subarachnoid Hemorrhage: Update for Emergency Physicians. J Emerg Med 2008.
Original Research: Blok et al. CT within 6 hours of headache onset to rule out SAH. Am Acad Neurology 2015
Remember: look on the Asynchronous Learning page to find which questions you need to answer for each of the above articles.
Additionally, this month’s corresponding AIR module will be Psychiatry.
13 comments
Blok et al.
1. What are some strengths of the study? Included nonacademic/community hospitals in the study.
2. What are the limitations? Study not from the USA, the reported false-positives actually may not have been false (given possible lost to follow-up and/or no gold standard for CSF spectrophotometry). Also, +/- the study used spectrophotometry which not all hospitals use.
3. What are the main outcomes of the study? Presence of intracranial hemorrhage/SAH in patients whose initial head CT was reported as negative.
4. Does reading this article change your practice? If so, how? No but I look forward to more studies on this topic.
I totally agree Andrew.
It’s odd that the patients underwent CT within 6 hours, then LP > 12 hours after sx onset.
Overall, it’s retrospective and done in the Netherlands, where the pre-test probability of SAH in this population is lower than that of the BMJ article a couple of years back.
I don’t understand why Blok et al or the Dutch do not use RBC’s in the CSF.
In the Edlow review of SAH, 99% of US hospitals do NOT Use spectrophotometry but use visual inspection instead, which is less reliable.
Despite the limitations, this paper does add to the pre-existing literature negative the need for LP if CT is negative within 6 hours of presentation. Though I’m not sure this seals the deal on this topic, it is nice to see that community hospitals can show similar results.
Since there’s still a small degree of uncertainty, I still like the approach of discussing the benefits and risks with the appropriate patient. Tell them that their risk after a CT within 6 hours is extremely low, likely <1%, and see if they want to go through the LP. And document the heck out of that conversation.
Other thoughts?
Edlow et al.
1. How has reading the article changed your practice?
Changed my index of suspicion for SAH (it is more common than I previously thought).
2. What information had you believed in previously that were debunked by reading this article?
I did not know that SAH headaches can resolve with migraine medication.
3. What new information did you learn from reading this article?
Intracranial blood in anemic patients can appear isodense and be more difficult to see on CT. Unruptured aneurysms can be symptomatic.
4. What are current areas of uncertainty on this topic that can be potential areas for research?
Ruling out SAH by obtaining head CT in <6 hours. Ideal blood pressure parameters. Best anti-HTN agents.
Blok et al.
1. What are some strengths of the study? Good evidence for our community physicians to use for a common clinical complaint where the “worse-case” diagnosis we are looking for can be catastrophic if missed. They used 13 different nonacademic hospitals. They also reviewed all radiology reads for accuracy.
2. What are the limitations? Generalizability, retrospective, done in the Netherlands, original radiology read missed 1 perimesenphalic non-aneurysmal bleed (1/760), possibly incomplete follow-up (since patients may have gone to non-study hospitals with possible SAHs).
3. What are the main outcomes of the study? Whether staff radiologists in nonacademic hospitals can reliably rule out a SAH on head CT within 6 hrs after headache onset.
4. Does reading this article change your practice? If so, how? Definitely. I think this gives mores support to the no LP. LPs are time-consuming, uncomfortable for the patient, and have their own consequences..not too mention if it’s falsely positive (51/52 patients in this paper) you will undergo an even larger, costlier work up. There is ALWAYS room for shared decision making but I feel comfortable saying to patients that in this scenario, their risk of SAH is very very low.
Edlow et al
How has reading the article changed your practice?
I agree with andrew – this article has increased my index of suspicion for SAH. It’s more common than I thought. 10% of people with sudden HA different from prior have SAH
What information had you believed in previously that were debunked by reading this article?
I used to think that xanthochromia came from RBCs, but it’s actually the breakdown products (bilirubin, etc) that cause the color change. Hence, you shouldn’t see xanthochromia until a few hours after the onset of the SAH (maybe this is more of a reason NOT to LP someone with a clear HA history that started, say 1 hour ago with a negative head CT?)
What new information did you learn from reading this article?
As mentioned before, high incidence of SAH (!% of all HA patients), high rate of misdiagnosis (25%).
What are current areas of uncertainty on this topic that can be potential areas for research?
The article glances over seizure and vasospasm prophylaxis. Is there evidence for starting these in the ED (as opposed to in the ICU)? I sometimes find that consultants ask for keppra, but I don’t think I have seen nimodipine given in the ED (probably because onset of vasospasm is later).
1. What are some strengths of the study?
community EDs (everyday practice), large sample size, good follow-up
2. What are the limitations?
retrospective, single country (?generalizability to other countries), all reads were final attending radiologist reads (unable to measure how well resident reads held up). Why did they exclude patients with LP < 12 hour from symptom onset?
3. What are the main outcomes of the study?
99.9% NPV for SAH from a negative head CT read from an attending community radiologist (missed 1 out of 760 head CTs read in this study)
Does reading this article change your practice? If so, how?
Makes me lean toward not doing LP if symptom onset is clearly < 6h from CT scan, AND CT is read by an attending radiologist. At the very least, I think having informed discussion with the patient about the very low likelihood of SAH is warranted
Edlow et al.
1. How has reading the article changed your practice?
Like Andrew, this made me more vigilant for SAH. I also was unaware of how high the prevalence of intracranial aneurysms are in the general population. Beyond that, this paper is a little old (2008) to change much of my practice
2. What information had you believed in previously that were debunked by reading this article?
HALF of all SAH patients have atypical symptoms; I thought it was a lower amount of pts.
I though it took >12h for Xanthrochromia to show up visually on LP but apparently it can show up earlier.
3. What new information did you learn from reading this article?
Lots – One in 100 patients with h/a have an SAH! Misdiagnosis occurs in 1/4 of patients (demonstrating the importance of good d/c instructions), Traumatic taps occur 10-15% of the time (more when I’m procedure resident). In vasospasm there is no evidence for “triple-H” therapy, which I got pimped on every time there was an SAH in the SICU.
4. What are current areas of uncertainty on this topic that can be potential areas for research?
HA <6h — CT only pathway needs more evidence
Evidence for Triple H therapy
Better scanning techniques (higher generation scanners) to see more sentinel bleeds
Does u/s guided LP decrease amount of traumatic taps?
Agree with Andrew – bp control and anti-HTN agents in SAH.
Edlow et al
How has reading the article changed your practice? Not much. I think I have a high suspicion for SAH already and knew of these numbers before hand. It is a great review article.
What information had you believed in previously that were debunked by reading this article? Nimodipine in the ED. Of the 2 SAH I have seen NSGY asked me to get it to the patient right away even though I was pretty sure that it shows up later and also the effect of nimodipine is only when it shows up
What new information did you learn from reading this article?
As mentioned above – its a great review article and was interesting to read some of the management guidelines after diagnosis. Also leaned about H&H score. I thought they summarized reasons for missing diagnosis well. Data for CTA and LP first strategy was also new to me.
What are current areas of uncertainty on this topic that can be potential areas for research? Would love to see imaging that can essentially safely rule out SAH in any comers tot he ED and eliminate LPs (which are easy to do in academic centers but a pain in the butt at community places that are busy)
What are some strengths of the study? Asks a great question and done in community hospitals where I feel this is more pertinent
What are the limitations? Retrospective design, one country, LP at 12 hours
What are the main outcomes of the study? Presence of SAH if CT within 6 hours
Does reading this article change your practice? If so, how? Yes; it will make me talk to patients regarding further work up and the chance of an actual SAH. I think a lot of patients are very appreciative of this and do not want to go through LPs if not necessary. Obviously documentation is important
Blok et al.
1. What are some strengths of the study? They had a fairly large sample size and used minimal exclusion criteria, so it does apply to most patients that would be seen in the ED for headache.
2. What are the limitations? The main limitations are with regard to the ability to generalize the data to our population. This study was done in the Netherlands, with CTs done 12 hrs after symptom onset. This large time gap just would not be feasible in our population, and it is not addressed how this may have affected the data.
3. What are the main outcomes of the study? Presence of an intracranial hemorrhage in patients whose initial head CT was reported as negative, as detected by spectrophotometry of the CSF.
4. Does reading this article change your practice? If so, how? As most have already said, It would not necessarily change my practice but would make me feel more comfortable entering into shared decision making with the patient when I know the post-test probability for a SAH after a negative CT is extraordinarily low.
Edlow et al.
1. How has reading the article changed your practice?
As everyone else has already said, the main takeaway from this article was that I got more nervous about working up patients with headaches because so many of them turned out to have SAH (even when they only present with atypical symptoms).
2. What information had you believed in previously that were debunked by reading this article?
I had previously believed that without neurologic deficits, the pretest probability for a SAH was extremely low, but this article made me more vigilant of who should be getting CTs.
3. What new information did you learn from reading this article?
I did not realize that it took so long for the LP to result to show xanthrochromia. This makes me feel like it is really unrealistic in our environment, and basically requires you to admit the patient for observation.
4. What are current areas of uncertainty on this topic that can be potential areas for research?
They didn’t touch on sodium level control, even though that is definitely something that was really pushed hard when I was in the SICU with these patients, they were all put on 3% at some point, which I was told was to prevent vasospasm, but I would like to know some more on the evidence and guidelines behind that.
How has reading the article changed your practice?
– Reading the article and learning about the high rate of misdiagnosis of SAH has made me keep SAH high up on differential for any HA patient.
– I don’t think I would typically measure opening pressure on LP for r/o SAH, but according to this article, 2/3 of SAH pts have elevated opening pressure, so I might begin to measure it in future LPs.
What information had you believed in previously that were debunked by reading this article?
– I didn’t realize the high rate of SAH patients who present with atypical or mild features.
– I also didn’t know previously that intracranial blood in anemic pts (HCT <30) is more difficult to see/identify on CT.
What new information did you learn from reading this article?
– the points I mentioned above, as well as the high rate of rebleeding over initial 2 weeks.
What are current areas of uncertainty on this topic that can be potential areas for research?
– role and timing of use of hypertonic saline, anti-epileptics
– Strengths of study
decent number of community EDs
reasonably good sample size
– limitations
generalizability
non-US setting
LP at 12h
– Main outcomes
99.9% NPV for SAH from a negative head CT read (attending community radiologist)
– Does this study change your practice
I think, as others have said, it gives great incentive to have a thorough discussion of risks/benefits of LP w/ patients before they consent to having a procedure which may not change their outcomes.