Aetiology and treatment of severe postpartum haemorrhage.

This thesis is comprised of three studies focusing on severe postpartum haemorrhage (PPH). PPH is a major cause of maternal morbidity and mortality worldwide. Risk factors include retained placenta, prolonged duration of the third stage of labour, previous caesarean section, and operative vaginal delivery. Occurrence and development of PPH are, however, unpredictable and can sometimes give rise to massive haemorrhage or even hysterectomy and maternal death. Severe haemorrhage can lead to coagulopathy causing further haemorrhage and requiring substitution with blood transfusions. The aim of this thesis was to investigate causes of severe PPH and investigate methods of early prevention. 
The first study was a randomised controlled double-blinded trial investigating the effect of treatment with pre-emptive fibrinogen on women with severe PPH. The primary outcome was the need for red blood cell transfusion at 6 weeks postpartum. A total of 249 women were randomised to either 2 grams of fibrinogen or placebo. The mean concentration of fibrinogen increased significantly in the intervention group compared to the placebo group (0.40 g/l, confidence interval: 0.15-0.65), but there was no difference in the need for postpartum blood transfusions (relative risk 0.95, confidence interval: 0.15-1.54). No thromboembolic complications were detected.
The second study was a population-based observational study including 245 women receiving ≥10 red blood cell transfusion due to PPH. The cohort was identified by combining data from The Danish Transfusion Database with The Danish Medical Birth Registry, with further data extraction and validation through review of patient charts. The main causes of massive postpartum transfusion were atony (38%) and abnormal invasive placenta (25%). Two of the women in the cohort died, an additional six had a cardiac arrest, and a total of 128 women (52%) required a hysterectomy. Hysterectomy was associated with increased blood loss, increased number of blood transfusions, a higher fresh frozen plasma to red blood cell ratio (p=0.010), and an increased number of red blood cells before first platelet transfusion (p=0.023). Hysterectomy led to haemostasis in only 70% of cases.
The third study was a register-based cohort study, includ-ing 43,357 vaginal deliveries from two large Danish maternity units. Different cut-offs were used to define PPH. There was a difference in distribution of causes depending on the cut-off used, with atony playing a decreasing role and a retained placenta an increasing role the higher the cut-off used. In a multivariate linear regression model retained placenta was identified as a strong predictor of quantity of blood loss. The duration of the third stage of labour was a very weak predictor after adjusting for the influence of a retained placenta. 
In conclusion, an improved diagnosis of the causes of PPH especially retained placenta, together with an early recognition and treatment of coagulopathy, seem to be important in reducing severe PPH in an aim to minimize associated maternal morbidity.