[display_podcast]

Date: April 13th, 2018

Reference: Effect of early tranexamic acid administration on mortality, hysterectomy, and other morbidities in women with post-partum haemorrhage (WOMAN): an international, randomised, double-blind, placebo-controlled trial. The Lancet 2017

Guest Skeptic: Dr. Nick Papalia completed his MD at Western University. He is currently completing his third year of Obstetrics and Gynecology residency at the University of Calgary.

Case: 37-year-old primiparous woman has a spontaneous vaginal delivery following an induction of labour at 39 weeks for gestational diabetes for which she is treated with insulin. She delivered a vigorous 3800g boy following a brief episode of shoulder dystocia. She is otherwise healthy but does have a remote history of asthma. Aside from insulin and prenatal vitamins, she takes no medications.

Approximately five minutes after delivery of the placenta, as you are evaluating her perineum, she begins to have a moderate amount of vaginal bleeding. Her nurse performs fundal massage which temporarily improves her flow.

Background: Utako and Shosuke Okamoto discovered TXA in the 1950’s while searching for a treatment for postpartum hemorrhage [1].  TXA is a synthetic analog of the amino acid lysine. The normal process is for plasminogen to breakdown fibrin and prevent blood clots. TXA acts as an antifibrinolytic agent. It binds to lysine receptor sites on plasminogen blocking its action on fibrin. The ultimate result is the fibrin matrix structure is maintained and bleeding is reduced.

Post-partum hemorrhage (PPH) is defined by the World Health Organization (WHO) as:

  • a cumulative blood loss of greater than or equal to 1,000 mL or blood loss accompanied by signs or symptoms of hypovolemia within 24 hours after the birth process”. PPH is one of the leading causes of maternal mortality around the world [2].

We have covered the use of TXA for trauma (CRASH-2) with Dr. Anand Swaminathan. This trial showed TXA is safe and effective treatment in patients with hemorrhagic shock from trauma in reducing mortality.  We also did a couple of podcasts on TXA being effective for epistaxis (SGEM#210 and SGEM#53).

The American College of Obstetricians and Gynecologists (ACOG) published guidelines for the management of postpartum hemorrhage including the use of TXA. They give TXA a Level B recommendation (based on limited or inconsistent scientific evidence): [3]

  • “Given the mortality reduction findings, tranexamic acid should be considered in the setting of obstetric hemorrhage when initial medical therapy fails.”

There was also a commentary published in the issue of Obstetrics and Gynecology by Pacheco et al. [4] They concluded that TXA appeared to be a safe and effective option in the treatment of post-partum hemorrhage. Their recommendation is as follows:

  • “In established postpartum hemorrhage, may administer 1 g intravenously within 3 hours of birth. If required, may administer a second dose within the next 24 hours. The half-life of TXA is 2 hours and its antifibrinolytic effect lasts for up to 7–8 hours in serum. A single dose should suffice in most cases.
”

We checked to see if the Society of Obstetricians and Gynaecologists of Canada (SOGC) and could not find any practice guidelines on TXA for PPH. The latest guidelines from the SOGC on PPH was published in 2009 [5].


Clinical Question: Does TXA improve survival in women with post-partum hemorrhage?


Reference: Effect of early tranexamic acid administration on mortality, hysterectomy, and other morbidities in women with post-partum haemorrhage (WOMAN): an international, randomised, double-blind, placebo-controlled trial. The Lancet 2017

  • Population: Women greater than 16 years of age with a clinical diagnosis of PPH after vaginal birth (> 500ml blood loss) or cesarean section (> 1,000ml blood loss) or blood loss causing haemodynamic instability and the clinician was uncertain whether or not to use TXA.
    • Exclusions: If the clinician felt that TXA would either clearly be beneficial or clearly would not be appropriate.
  • Intervention: 1g TXA via slow IV infusion in addition to usual management. Option of additional administration if bleeding continued for 30 minutes or more or stopped and restarted.
  • Comparison: Placebo
  • Outcomes:
    • Primary Outcome: All-cause mortality or hysterectomy within 42 days of randomisation
    • Secondary Outcomes:
      • Mortality due to bleeding
      • Thromboembolic events (deep-vein thrombosis, pulmonary embolism, myocardial infarction, and stroke)
      • Surgical interventions (intrauterine tamponade, embolisation, brace sutures, arterial ligation, hysterectomy, and laparotomies done after randomisation to control bleeding and achieve haemostasis)
      • Complications (renal failure, cardiac failure, respiratory failure, hepatic failure, sepsis, and seizures)
      • Adverse events (untoward medical events)
      • Quality of life measured using the EQ5D
      • Status of any thromboembolic events in breastfed babies

Authors’ Conclusions: Tranexamic acid reduces death due to bleeding in women with post-partum haemorrhage with no adverse effects. When used as a treatment for postpartum haemorrhage, tranexamic acid should be given as soon as possible after bleeding onset.”

Quality Checklist for Randomized Clinical Trials:

  1. The study population included or focused on those in the emergency department. No.
  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. Yes
  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. Unsure

Results: There were 20,021 women included in the trial (10,036 TXA and 9,985 placebo). Maternal death occurred in 483 (2.4%) of all women within 24 hours and 43 (9%) of the deaths were within one hour after randomization


Key Result: No statistical difference in the primary composite outcome (all-cause mortality or hysterectomy)


  • Primary Outcome: All-cause mortality or hysterectomy within 42 days of randomisation
    • No difference 5.3% TXA vs. 5.5% Placebo
    • RR 0.97, 95% CI 0.87 to 1.09; p=0.65

  • Secondary Outcomes:
    • All-Cause Mortality: No difference 2.3% vs. 2.6%
      • RR 0.88, 95% CI 0.74–1.05; p=0.16
    • Hysterectomy Rate: No difference 3.6% vs. 3.5%
      • RR 1.02, 95% CI 0.88–1.07; p=0.84
    • Mortality Due to Bleeding: Significant 1.5% vs. 1.9%
      • RR 0.81, 95% CI 0.65-1.00; p=0.045
      • Adjusted RR 0.78, 95% CI 0.62–0.98; p=0.03
      • No difference in deaths from other causes (PE, organ failure, sepsis, eclampsia and other causes)
      • Subgroup less than three hours: Significant difference 2% vs. 1.7% RR 0.69, 95% CI 0.52–0.91; p=0.008
    • Thrombotic Events: No difference
    • Surgical Interventions: Women who had TXA were more likely to have had brace sutures and less likely to have had a laparotomy.
    • Transfusions: No difference
      • Of those who did get transfused, no difference in mean number of units between groups
    • Quality of Life: No difference
    • Adverse Events: No difference

1) Inclusion – Clinical diagnosis of PPH was based on estimated blood loss. This is a subjective parameter. Can physicians reliably tell who has lost more than 500ml after a vaginal delivery and 1,000ml post C-section? In addition, they used hemodynamic instability as another inclusion criteria but did not define what would constitute compromised stability. Was it based on an objective measure of blood pressure and/or heart rate?

It is even more complicated/vague because the inclusion was not just based on the presence of post-partum hemorrhage but the clinicians needed to be uncertain about giving TXA. While these inclusion criteria may be pragmatic they can also introduce selection bias. It is unclear if the bias would be in favor or against the efficacy of TXA.

You could also see how this vague inclusion criteria could lead to indication creep where all women with any bleeding post-partum are given TXA routinely.

2) Power –There were some complications with how they powered the study. It was originally powered for a 25% relative reduction of the composite outcome (all-cause mortality or hysterectomy). This relative reduction translates to a 1% absolute reduction (4% to 3%). Then they figured out the decision to do a hysterectomy was made usually at the same time as randomization.

Therefore, they recalculated the sample size for 25% all-cause mortality relative reduction. This decreases the absolute reduction to only 0.75% (3% to 2.25%).  As a result, the sample size increased from 15,000 to 20,000 due to the small effect size they were trying to determine.

Their sample size was achieved by recruiting just over 20,000 women to determine if TXA would decrease all-cause mortality by 0.75%. However, the observed all-cause mortality rate was a little lower (2.6%) than the 3% estimated.  Regardless, they found only a 0.4% absolute reduction in all-cause mortality and this was not statistically significant (RR 0·88, 95% CI 0·74–1·05; p=0·16)

3) Subgroup Analyses – They pre-planned a number of subgroup analyses for the primary composite outcome. One was to look at time to TXA after birth, another was type of birth (vaginal vs. C-section) and the third was primary cause of hemorrhage (uterine atony vs. all others).  The only statistical difference observed was in the time to TXA if given in less than three hours after birth. This positive result should be viewed with much skepticism. Subgroup analyses are considered hypothesis generating and should not be over-interpreted.

4) Fragility – This study can also be viewed using the fragility index [6]. This is a way to measure the robustness of the results obtained. It is calculated by converting one patient in the group (treatment or control) from “non-event”to an “event”. In this case, how many women would have to have a different outcome for the study not to be statistically significant (p ≥ 0.05)?

The primary composite outcome of all-cause mortality or hysterectomy was not statistically different. All-cause mortality was also not statistically different. However, there was a statistical difference in death due to bleeding. But the calculated fragility index was zero [7]. This result emphasizes the lack of robustness of the WOMAN trial.

If you calculate the fragility index for the subgroup analysis of the secondary outcome of bleeding mortality treated within three hours, it is nine. So, if nine women had their death attributed to something other than bleeding by the treating clinician than the results would no longer be considered statistically significant.

5) Spin – This study seems to have been spun as a positive study despite the lack of statistical difference in the primary composite outcome. This is evident in the first sentence of the discussion. They did not lead with the primary outcome but rather with a secondary outcome (death due to bleeding) and gave a relative number (nearly 1/3 reduction in death) rather than absolute number. The absolute reduction in death due to bleeding was 0.4% (NNT=250). But the fragility index was zero.

An infographic created from the WOMAN trial data presents a distorted picture. They highlight the relative reduction of the secondary outcome. They do not identify that there was no statistical difference in the primary outcome. They also do not report the absolute reduction in death due to bleeding was actually 0.4%.

Dr. Jerome Hoffman

6) Jerome Hoffman – I discussed this trial with Legend of EM Dr. Jerome Hoffman about this paper. He wanted to emphasize a few points:

  1. External validity from developing countries to developed countries. What resources did their health care system have to address post-partum hemorrhage?
  2. The over powering of studies. We often talk about under-powered studies but not over powered studies. Studies are usually designed to find a difference between two things and if you have a large enough study you will find a statistical difference because no two things are identical.
  3. Disease specific mortality can be hard to define. Overall mortality was not statistically different between the two groups but death due to bleeding was lower. Clinicians were requested to record the immediate cause of death rather than the underlying cause of death. At the end of the day, if we save more women from bleeding to death due to post-partum hemorrhage but the overall number of deaths is the same then what is the benefit?
  4. Coagulation in the body is finely tuned. When we tip the balance towards less bleeding it tends to increase clotting.
  5. The study was funded in part by Pfizer the maker of TXA. This should make us more skeptical of the results and how the results will be marketed by the sponsor.

Comment on Authors’ Conclusion Compared to SGEM Conclusion: We do not agree with the authors’ conclusions.


SGEM Bottom Line: Tranexamic acid MAY or MAY NOT improve survival in women with post-partum hemorrhage.


Case Resolution: You determine that your patient’s bleeding to be secondary to atony, the number one cause of post-partum hemorrhage. You perform a bimanual exam to evacuate any remaining tissue and clots while asking the bedside nurse to initiate an oxytocin infusion. You also request for the post-partum hemorrhage kit in the room which has essential medications. Her bleeding has improved and she’s no longer passing clots. You decide she needs a uterotonic beyond oxytocin but given her history of asthma, you elect for ergotomine which the nurse administers intra-muscularly.

Over next 15 minutes you repair her second degree laceration which has mostly been hemostatic. You assess her total estimated blood loss to be 1000mL (which is probably inaccurate) so you ask the nurse to administer 1g TXA by IV.

She has no further bleeding and is transferred to post-partum. On post-partum day one her hemoglobin is 83g/L (8.3g/dl) and she is discharged with a prescription for iron polysaccharide.

Dr. Nick Papalia

Clinical Application: We do not have good evidence that TXA provides a mortality benefit in post-partum hemorrhage. Until there is better evidence of efficacy it should not be incorporated into protocols for routine use but rather may be considered as suggested by the ACOG level B recommendation.

What Do I Tell the Patient? When you are having heavier bleeding after your vaginal delivery.  We have strong evidence for some medicines can improve bleeding. There is a newly purposed medicine called TXA that may or may not have a benefit. There is not enough evidence to suggest we should use TXA routinely in favour of the other medications that we know are effective.

Keener Kontest: Last weeks’ winner was Daniel Ritter a PGY1 at the University of Wisconsin in Madison, WI. He knew the trigeminal nerve affects the vagus nerve induce the mammalian diving reflex to terminate SVT.

Listen to the SGEM podcast on iTunes to hear this weeks’ question. If you know the answer send an email to TheSGEM@gmail.com with “keener” in the subject line. The first correct answer will receive a cool skeptical prize.

Other FOAMed:

 


Remember to be skeptical of anything you learn, even if you heard it on the Skeptics’ Guide to Emergency Medicine.


  1. TXA Central
  2. Say L, Chou D, Gemmill A, Tuncalp O, Moller AB, Daniels J, et al. Global causes of maternal death: a WHO systematic analysis. Lancet Glob Health 2014
  3. ACOG Practice Bulletin – Postpartum Hemorrhage 2017.
  4. Pacheco LD et al. Tranexamic Acid for the Management of Obstetric Hemorrhage. Obs and Gyn Oct 2017
  5. SOGC – Active Management of the Third Stage of Labour: Prevention and Treatment of Postpartum Hemorrhage 2009
  6. Life in the Fast Lane – Fragility Index
  7. ClinCalc – Fragility Index