One caveat is that the subjective inclusion criteria may have introduced selection bias into the study. Clinical diagnosis of PPH was based on the ability to estimate blood loss (500 mL after vaginal delivery or 1,000 mL C-section) which can vary between clinicians. Hemodynamic instability was also an inclusion criteria but objective measurements to define this were not discussed. Additionally, patients were enrolled only if clinicians were uncertain about giving TXA leaving out subgroup of patients with more severe pathology who were given TXA as standard of care.
Excluding these patients, especially those with hemodynamic instability from severe hemorrhage may have lead to an underestimation of the TXA’s true efficacy.
Another caveat to consider is that the sample size was significantly increased from 15,000 to 20,000 after investigators discovered that the decision to perform hysterectomy was commonly made at the time of randomization and thus could not be impacted by intervention. This runs into the risk of overpowering which can interpret seemingly small and unremarkable differences between treatment and control arms as statistically significant.
Although we only reported individual outcomes in our review, the original study selected a composite outcome (death from all-causes or hysterectomy within 42 days of giving birth) as their primary outcome. The authors interpreted their results as positive despite the lack of statistical difference in this primary composite outcome. This was partly from shifting focus from all-cause mortality to cause-specific mortality specifically death due to bleeding. Although there was significantly less death from bleeding in the TXA group, the calculated fragility index was zero pointing to the lack of robustness of this dataset. We also do not consider disease-specific mortality a true efficacy endpoint.
In light of the results of this trial, WHO recommends administration of tranexamic acid to all patients with postpartum hemorrhage. This recommendation is not supported by the existing evidence with regards to all-cause mortality. Therefore we have assigned an NNT color recommendation of yellow (Unclear if Benefits).
However, considering the safety and low cost of tranexamic acid ($45-$55 per vial), and the potentially devastating consequences of postpartum hemorrhage, we believe TXA should be considered as an attractive adjunctive therapy when other modalities (e.g. treating uterine atony) fail to control the hemorrhage.
The original manuscript was published in
Medicine by the Numbers, American Family Physician as part of the partnership between TheNNT.com and AFP.