Revisit of risk factors for major obstetric hemorrhage: insights from a large medical center

To revisit risk factors of major obstetric hemorrhage in a large obstetric center. A retrospective case control study was conducted based on institutional electronic database and blood bank registry of a single center, 2005-2014. The major obstetric hemorrhage event was defined as transfusion of a parts per thousand yen5 red blood cells units within 48 h of birth and compared to matched group (ratio 1:4) based on the time of birth. Multivariable stepwise backward logistic regression models were fitted to determine risk factors for major obstetric hemorrhage. Odds ratio (OR), further evaluated by standard measures of the predictive accuracy of the logistic regression models, C statistics, and associated neonatal adverse outcome are reported. 113,342 women delivered during the study; 122 (0.1 %) women experienced major obstetric hemorrhage. There was one major obstetric hemorrhage fatality (0.8 %). Compared to the controls, we identified historical as well as significant current modifiable risk factors for major obstetric hemorrhage: multifetal pregnancy (OR 3.92; 95 % CI 1.34-11.52; p = 0.013), induction of labor (OR 2.81; 95 % CI 1.22-7.05; p = 0.027), cesarean section (OR 25.56; 95 % CI 12.88-50.75; p < 0.001), and instrumental delivery (OR 6.58; 95 % CI 2.36-18.3; p < 0.001). C statistics of the model for major obstetric hemorrhage prediction was 0.919 (95 % CI 0.890-0.948, p < 0.001). Major obstetric hemorrhage is a rare event with potentially modifiable risk factors which represent a platform of interventions for lessening obstetric morbidity.