Title: Reducing maternal mortality in the 21st century
Abstract: To critically examine the causes and potential prevention of maternal death in the United States since 2000. We examined medical records from all maternal deaths among 1.24 million births occurring between from 2000-2005 in the nation's largest healthcare system. We assessed the causal relationship between route of delivery and death and determined whether such deaths were potentially preventable. 78 maternal deaths occurred in 1,239,385 deliveries (6.3/100,000 births.) For women undergoing cesarean delivery, the rate was 13.2/100,000 cesareans and for those undergoing vaginal birth, 1.5/100,000 vaginal births. Leading causes of death were: Amniotic fluid embolism −15%, complications of preeclampsia −15%, pulmonary thromboembolism −13%, cardiac disease −11%, obstetric hemorrhage −9%, non-OB infection −8%, accident/suicide 7%. In only 4 cases of cesarean section was death caused by complications of the operation itself rather than by the indication for the cesarean. All of these cases were deemed potentially preventable. Cesarean section may have contributed to death in 9 additional cases of pulmonary thromboembolism. There was only 1 fatal pulmonary embolism among the women delivering vaginally. In 10 cases, (13%) death was attributable to medical provider error. In 10 cases, (13%) death was attributable to the actions of the patient or another non-medical provider. Most maternal deaths are not currently preventable. Although cesarean section is associated with a higher mortality rate than vaginal delivery, it is only uncommonly causative. Deaths causally linked to cesarean were, in our series, all potentially preventable. Given the diversity of causes of maternal death, significant reductions in the maternal mortality rate in the United States will only come from improved understanding of the nature and prevention of preeclampsia and amniotic fluid embolism or universal thromboprophylaxis for all women undergoing cesarean section.