OBJECTIVE: The aim was to find clinically useful risk factors for postpartum transfusion and to assess the joint predictive value in a population of women with a first and second delivery.

METHODS: All Danish women with a first and second delivery from January 2001 to September 2009 who gave birth in a hospital that reported transfusion of red blood cells to a national database: A total of 96 545 women were included.

RESULTS: Retained placental tissue explained more than all other risk factors in vaginal deliveries. Retained placental tissue at first delivery was associated with postpartum transfusion at a second vaginal delivery, and may also be used as an early predictor in parallel with a history of either placental abruption, postpartum transfusion or caesarean delivery. The positive predictive values of having more than one risk factor was low (2.2%-2.7%).

CONCLUSIONS: Prediction of postpartum transfusion is difficult. Retained placental tissue is the strongest predictor of postpartum blood transfusion in vaginal deliveries. Retained placental tissue is usually diagnosed for the first time when the bleeding starts, which limits the clinical value of prediction. We need tools for an early diagnosis of retained placenta to intervene early before transfusion is needed.