Thesis defense Dacia Henriquez

On  1 December 2020 Dacia Henriquez defended her thesis “Recognition and management of persistent postpartum haemorrhage; time to take timing seriously”

Promotores:
Prof JG van der Bom MD PhD and Prof KWM Bloemenkamp MD PhD

Venue:
Leiden University Academy Building and online

Summary

The three central questions of this thesis concerned the recognition and management of women with severe postpartum haemorrhage. We specifically addressed the timing of recognition of women with high risk of adverse outcome, and the timing of obstetric and haemostatic interventions to stop bleeding in these women.

1. Women with postpartum haemorrhage: what issues to resolve?

Within the International Network for Obstetric Surveillance Systems, we performed a cross-country comparison of management and outcomes of women with severe postpartum haemorrhage in 6 high-resource countries. This comparison showed similar causes of postpartum haemorrhage among women with at least 8 units of packed red blood cells transfused, but revealed striking differences in both obstetric and haemostatic management. Case fatality rate between the countries did not differ, but there was up to 40% difference in hysterectomy rates and a 5-fold variation in the incidence of massive transfusion. In our review on how to improve maternal outcomes, we identified four opportunities for improvement of the management of women with postpartum haemorrhage: 1) definitions of severe postpartum haemorrhage and their use to recognise clinical deterioration in women with severe postpartum haemorrhage along with timely recognition of women at risk of adverse outcomes, 2) the timing of obstetric interventions to stop bleeding, 3) the timing of switch from fluid resuscitation with crystalloids and colloids to transfusion of packed red blood cells, and 4) the diagnosis of coagulopathy during ongoing postpartum haemorrhage and the timing of haemostatic interventions to stop bleeding.

 

2. II. Women with postpartum haemorrhage: who is at risk of adverse outcome?

Within the TeMpOH-1 study population, we compared the clinical characteristics and outcomes of women captured by the definition persistent postpartum haemorrhage with those of women captured by the most common definitions of severe postpartum haemorrhage based on volumes of blood loss and transfusion criteria. This definition identified women at an early stage of haemorrhage, comparable with definitions based on volume of blood loss. It also captured a large majority of adverse maternal outcomes, contrary to definitions based on transfusion criteria. Because women with hypertensive disorders of pregnancy have a high risk of postpartum haemorrhage compared with women without hypertensive disorders, we wondered whether 192 outcomes of women with persistent postpartum haemorrhage and concurrent hypertensive disorders of pregnancy would be worse than the outcomes of women with persistent postpartum haemorrhage who did not have concurrent hypertension. Women with concurrent preeclampsia experienced more haemorrhage-related adverse outcomes than women without concurrent hypertensive disorders of pregnancy, odds ratio 1.8 (95% confidence interval 1.1-3.0).

 

3. Women with postpartum haemorrhage: when and what to transfuse?

Finally, in this thesis we also addressed the timing of switch from fluid resuscitation with crystalloids and colloids to transfusion of packed red blood cells and the timing of initiation of plasma transfusion in women with persistent postpartum haemorrhage. Fluid resuscitation with more than 4L crystalloids and colloids was associated with more adverse maternal outcomes as compared with fluid resuscitation with ≤2L clear fluids, after adjustment for severity of haemorrhage. In our time-dependent propensity scorematched analysis, plasma transfusion within 60 minutes after diagnosing persistent postpartum haemorrhage was not associated with better maternal outcomes as compared with no or later start of plasma transfusion, adjusted odds ratio 1.09 (95% confidence interval 0.57-2.08). This cohort study was specifically designed to address the confounding-by-indication inherently associated with research questions on timing of initiation of treatment. Addressing this risk of bias remains challenging in the field of haemostatic interventions to manage postpartum haemorrhage, as outlined in our commentary on a secondary analysis of a recent large randomised controlled trial (the WOMAN trial) evaluating the timing of administration of tranexamic acid in women with postpartum haemorrhage. In this commentary we describe the timedependent confounding that the authors did not adjust for, and as a consequence, the conclusion that delay in administration of tranexamic acid in bleeding patients reduces its effectiveness was not justified.