Date: March 11th, 2021

Guest Skeptic: Dr. Robert Edmonds is an emergency physician in the US Air Force in Ohio.


Reference: Post et al. Ultra-early tranexamic acid after subarachnoid haemorrhage (ULTRA): a randomised controlled trial. Lancet 2021

Case: You’re working a busy evening shift in your community emergency department (ED) when a 58 year old female presents with a rapid onset terrible intensity headache.  She has no significant headache history and you are concerned for subarachnoid hemorrhage so you order a head CT which confirms your suspicions.  You page neurosurgery at the bigger ED in town, and while you wait for the page back, you wonder if giving tranexamic acid (TXA) could help improve the patient’s chances for a good outcome given its effects in other bleeding processes.

Background: In the case presented, the woman would qualify using the Ottawa SAH Rule because of the rapid onset of an intense headache and her age. We have discussed the incredible work done by Dr. Jeff Perry and his group in the development of the Ottawa SAH Rule. Jeff was actually on the SGEM as the guest skeptic discussing this clinical decision instrument way back in 2013 (SGEM#48).

The Ottawa SAH Rule is to be applied to alert patients older than 15 years of age with new severe non-traumatic headache reaching maximum intensity within one hour.

It is not meant for patients with new neurologic deficits, previous aneurysms, SAH, brain tumor, or who have a history of recurrent headaches. This is defined as at least three or more episodes over the course of at least six months.

Our SGEM Bottom Line eight years ago was that the Ottawa SAH “Tool” was not ready for prime time to rule out low risk patients from investigations.

Fast forward to 2018 and the validation of the Ottawa SAH Rule by Dr. Perry and his group. The results of this prospective observational study were that the clinical decision instrument was 100% sensitive (missed no SAH patients) and 13.6% specific.

Dr. Chris Carpenter

The guest skeptic for SGEM#201 was Dr. Chris Carpenter, who literally wrote the book on diagnostic accuracy of clinical decision instruments in the ED with Dr. Jesse Pines. The SGEM bottom line from the episode was that the Ottawa SAH Rule needs external validation, a meaningful impact analysis performed, and patient acceptability of incorporating this rule into a shared decision-making instrument before being widely adopted.

Dr. Perry did publish a prospective implementation of the Ottawa SAH Rule (Stroke 2019). This article was covered on the SGEM with EM Nerd Dr. Rory Spiegel. The results demonstrated that the Ottawa SAH Rule is highly sensitive (100%) but has very poor specificity (13%). It is unclear how it performs against unstructured clinical judgement or in non-urban tertiary care teaching hospitals (SGEM#283).

Another issue the case brings up is whether a CT scan is good enough to rule out a SAH. The debate has historically been about whether or not you need to also get a lumbar puncture on these patients after a non-contrast CT head.

Dr. Jeff Perry

This brings us back to more work done by Jeff Perry. His team published a prospective cohort study that suggested if you got the CT scan done within six hours of headache onset, it was a third generation CT scanner, and it was read by a neuroradiologist, then you did not need to get an LP to rule out a SAH (BMJ 2011).

There were some limitations to this observational study. Another study was done in the UK that tried to address this issue of LP post normal CT. This was covered on SGEM#134. It found that the NNTap (number needed to Tap) to diagnose one aneurysm not identified by CT scan was 250.

The final issue the case identified was the use of TXA in treating patients with a SAH. We are skeptical given the previous review we did on the topic (SGEM#236). This was a structured critical review of the 2018 TICH trial published in the Lancet. The primary outcome showed no superiority of TXA compared to placebo for the mRS at 90 days. The SGEM bottom line was that TXA does not currently have evidence of improving outcomes in hemorrhagic stroke and routine administration cannot be recommended at this time.

However, as good healthy skeptics our positions are tentative and will change when presented with convincing evidence. Just because TXA was not demonstrated to “work” in one RCT does not mean we can claim TXA does not work. The burden of proof is on those making the claim of efficacy. Patients deserve the best care, based on the best evidence.

TXA has been discussed on the SGEM numerous times for treating a variety of conditions including: Trauma, isolated traumatic brain injury, gastrointestinal bleeding, post-partum hemorrhage and epistaxis.  Although some EM docs would like to believe TXA is one of the universal duct tapes of the ED, it has mixed results.

Patients suffering an aneurysmal SAH are at risk of rebleeding, which could worsens their long term clinical outcome and chance of death.  A hypothesis raised in the previous Sprigg et al RCT was that perhaps by treating these people earlier with TXA they would have a patient-oriented benefit.

Clinical Question: Does rapid administration of tranexamic acid in patients with CT confirmed SAH improve clinical outcome?

Reference:  Post et al. Ultra-early tranexamic acid after subarachnoid haemorrhage (ULTRA): a randomised controlled trial. Lancet 2021

  • Population: Adults 18 years and older with signs and symptoms for less than 24 hours indicating subarachnoid hemorrhage (SAH) AND who have a non-contrast CT confirming SAH.
    • Exclusions: Perimesencephalic bleed in combination with a GSC score of 13–15, and without loss of consciousness directly after ictus or focal neurological deficit on admission; traumatic SAH; ongoing treatment for VTE (DVT/PE); a history of a hypercoagulability disorder; pregnancy; severe renal failure, or imminent death within 24 h.
  • Intervention: TXA 1g IV bolus as soon as possible after a non-contrast CT proven diagnosis of SAH. This was followed by 1g infusion every 8 hours up to 24 hours (4g total) or until endovascular or surgical treatment, whichever came first.
  • Comparison: Usual care
  • Outcome:
    • Primary Outcome: Neurologic outcome dichotomized to good (mRS of 0-3) or poor (mRS 4-6) at six months
    • Secondary Outcomes: Excellent clinical outcome (mRS 0-2), ordinal shift of the mRS score, all-cause mortality (30 days and 6 months) and serious adverse events.

Authors’ Conclusions: In patients with CT-proven subarachnoid haemorrhage, presumably caused by a ruptured aneurysm, ultra-early, short-term tranexamic acid treatment did not improve clinical outcome at 6 months, as measured by the modified Rankin Scale.

Quality Checklist for Randomized Clinical Trials:

  1. The study population included or focused on those in the emergency department. Unsure
  2. The patients were adequately randomized. Yes
  3. The randomization process was concealed. Yes
  4. The patients were analyzed in the groups to which they were randomized. Yes
  5. The study patients were recruited consecutively (i.e. no selection bias). Unsure
  6. The patients in both groups were similar with respect to prognostic factors. Yes
  7. All participants (patients, clinicians, outcome assessors) were unaware of group allocation. No
  8. All groups were treated equally except for the intervention. Yes
  9. Follow-up was complete (i.e. at least 80% for both groups). Yes
  10. All patient-important outcomes were considered. Yes
  11. The treatment effect was large enough and precise enough to be clinically significant. No

Results: The cohort consisted of 955 patients with a mean age of 58 years and two-thirds were female. The median time from ictus to CT scan was 93 minutes and the median time from signs and symptoms suggestive of SAH and TXA was 185 minutes.

 Key Result: No superiority of TXA compared to usual care for patients with a subarachnoid hemorrhage.

  • Primary Outcome: Good clinical outcome (mRS 0-3)
    • 60% TXA vs 64% control
    • Adjusted Odds Ratio (aOR) 0.86 ( 95% CI; 0.66 to 1.12)
  • Secondary Outcomes: 
    • Excellent clinical outcome (mRS 0-2): 48% TXA vs 56% control, aOR 73 (95% CI; 0.57 to 0.95)
    • Ordinal shift of the mRS: No statistical difference
    • All-cause mortality at 30 days (22%) and 6 months (26%): No statistical difference
    • Adverse events: No statistical difference

1. Open label: The patient, investigators, and clinicians treating the patient were ALL aware of the patient’s inclusion in the treatment arm of the trial. This opens the study to significant biases. It’s unknown why the authors elected to pursue an open label study instead of providing the control group with a placebo, as this would have reduced the chances for bias significantly.

2. Patient Selection: It was not stated in the methods section that these were ED patients but it is likely that they were from the ED. They also did not explicitly say the patients were recruited consecutively 24/7/365. Twenty patients allocated to the TXA group did not receive treatment with nine of the twenty for “unknown” reasons. Patients were also excluded on the subjective assessment that death was imminent within 24 hours. These factors could have introduced some selection bias into the cohort of included patients.

3. Subgroup Analysis: The authors state that their findings show that routine use of TXA in spontaneous SAH cannot be recommended. However, the fact that their secondary outcome of excellent clinical outcome (mRS 0-2) was worse in the TXA group by a statistically significant amount is concerning.

We need to be careful not to over-interpret subgroup analyses. They should generally be considered hypothesis generating. If the subgroup had shown a benefit of TXA would there be a press release, infographic and appeal from authorities to implement a protocol based on a subgroup as was seen in the CRASH-3 trial and discussed on SGEM#270.

4. Adverse Events: It is reassuring that they did not identify a statistical increase in adverse events. There was a 4% absolute (50% relative) increase in seizures reported in the TXA treated group compared to the standard care group. This is most likely a spurious finding and represents noise in the data. The odds ratio was 1.52 (95% CI; 1.00 to 2.33) and this was only mentioned in a table and in the text of the result section. We wonder if this finding was reversed and TXA appeared protective against seizures would it have been highlighted?

5. External Validity: This was a large multicentred RCT performed in the Netherlands. The median time from onset of stroke signs/symptoms to CT scans was 93 minutes. There are many sites in Canada and the USA where it would not be possible to get a CT scan that quickly due to lack of local CT scanner and geography.

The median time to aneurysm treatment was 14 hours after symptom onset.  The authors posit that in a prior trial by Hillman looking at short term TXA use for SAH, the rate of rebleeds was lower than in their study, and in the Hillman study the first hospital contact was more than four hours after symptom onset.

Taken together, it challenges the generalizability of this study as other hospital systems in more rural/remote settings may have more substantial delays in patient presentation and time to aneurysm treatment. It remains unknown if in these other systems what role TXA might play, as it’s possible much of the benefit in rebleeding reduction isn’t seen when the aneurysm is treated in a timelier manner.

Comment on Authors’ Conclusion Compared to SGEM Conclusion: We generally agree with the authors’ conclusions.

SGEM Bottom Line: TXA use in subarachnoid hemorrhage does not presently have sufficient evidence of improving patient outcomes to recommend its routine use.

Case Resolution: You start nimodipine and monitor the patient’s blood pressure closely.  Neurosurgery accepts the patient and you transfer the patient to the larger hospital in town.

Clinical Application: No change in current practice, but it is helpful to know the limitations to the uses of TXA and to avoid falling for indication creep.

Dr. Robert Edmonds

What Do I Tell My Patient?  I’m glad you came in to the ED. Your headache is from an aneurysm. There is bleeding in your brain which is very painful. This can be treated. I’ve already spoken with the neurosurgeons. We are going to urgently transfer you to her hospital. She may recommend surgery to fix the problem. We will treat your pain and monitor your blood pressure closely. The ambulance should be here soon to transfer to the bigger hospital.

Keener Kontest: Last weeks’ winner was Sebastian Nemetz. He knew that the Kiesselbach’s plexus is the most common area for nose bleeds.

Listen to the podcast on iTunes for this weeks’ keener question. If you think you know the answer then send an email to with “keener” in the subject line. The first correct answer will receive a cool skeptical prize.

Remember to be skeptical of anything you learn, even if you heard it on the Skeptics