Background Hemorrhage is a critical contributor to maternal morbidity but estimation of blood loss at delivery is frequently inaccurate. Due to this inaccuracy we sought to examine blood transfusion as a surrogate marker for morbidity in a large non-population based consecutive cohort. Methods A retrospective analysis of prospectively gathered data was carried out at two university institutions serving a heterogeneous urban obstetric population from January to December 2016. Data were analyzed to determine whether individual characteristics were associated with perinatal transfusion. Hematological indices and requirement for other blood products were also characterized. Results A total of 16,581 deliveries were recorded during the study and 1.7% (289/16,581) of the cohort required red cell transfusion. Those who received transfusion were more likely to be nulliparous, and to deliver <37 weeks' or >42 weeks' gestation. They were also more likely to have a macrosomic infant (birthweight >4 kg) and to have had a multiple pregnancy. Characteristics not associated with risk of transfusion included obesity [18% (52/289) vs. 15% (2445/16,292); P=0.18], and maternal age ≥35 years [28% (82/289) vs. 33% (5537/16,292); P=0.05]. Additional blood products were necessary in a small number of patients who received red cells. Conclusion The rate of transfusion in a contemporary Irish cohort has risen compared with previous data. Several variables associated with transfusion are consistent with older studies but importantly; maternal obesity and advanced maternal age are not associated with transfusion. These data may encourage the investment of resources in a population previously considered low-risk and, following future studies, to improve strategies aimed at limiting blood transfusion.
Keywords: hemorrhage; morbidity; obstetrics; red cells; transfusion.