Trends in emergency peripartum hysterectomy at a single institution: 1988-2009.

OBJECTIVE: To evaluate trends of emergency peripartum hysterectomy over two decades. STUDY DESIGN: This was a retrospective cohort study of peripartum hysterectomies at one institution from 1988 to 2009. Medical records were reviewed and data from the first 11 years were compared with data from the second 11 years. RESULTS: During the study period, 558 emergent peripartum hysterectomies were performed with full records available for 553 to review: 280 in the first period, 273 in the second (p = 0.19). In the second period, prior cesarean deliveries, length of surgery, estimated blood loss, blood transfusions, and the number of units transfused were all increased (p < 0.001). Also in the second period, there were twice as many hysterectomies for suspected placental invasion, and an almost threefold increase in pathologically confirmed placental invasion (both p < 0.001). CONCLUSION: Although emergent peripartum hysterectomies are not increasing in frequency, the risk factors and morbidities have changed.

Calculated blood loss in severe obstetric hemorrhage and its relation to body mass index.

OBJECTIVE: To evaluate total calculated blood loss at the time of severe obstetric hemorrhage. STUDY DESIGN: This is a prospective observational study of women with obstetric hemorrhage. Women who received a blood transfusion for hypovolemia and those in which a body mass index (BMI) could be calculated were included. Total blood volume lost was calculated. Blood loss was analyzed in relation to maternal size as reflected in the BMI. RESULTS: Fourteen hundred forty-three women meeting inclusion criteria delivered at our hospital between March 2002 and June 2006. The median calculated volume of blood lost was 3529 mL, and 93% of women sustained losses ≥3000 mL. The blood loss sufficient to provoke signs and symptoms of hypovolemia was proportional to the woman's BMI. CONCLUSION: Women who develop hypovolemia during childbirth have suffered very large losses of blood, and infusion of blood products is required to restore circulation and prevent further morbidity.

Hydramnios in twin gestations.

OBJECTIVE: To estimate the effect of hydramnios on pregnancy outcomes in dichorionic and monochorionic twins. METHODS: This is a retrospective cohort study of women with twin pregnancies who underwent ultrasound evaluation between 1997 and 2010 and delivered liveborn neonates or stillborn fetuses at 24 weeks of gestation or more at a single institution. Hydramnios was defined as a single deepest pocket of amniotic fluid of at least 8 cm, and it was further categorized as mild (8-9.9 cm), moderate (10-11.9 cm), or severe (12 cm or more). The greatest degree of hydramnios identified during pregnancy was used for analysis. Monoamniotic pregnancies and pregnancies complicated by twin-twin transfusion syndrome were excluded. Anomalous neonates and stillborn fetuses were analyzed separately. RESULTS: Of 1,951 twin pregnancies, 1,311 were dichorionic (67%) and 640 were monochorionic (33%). Hydramnios was identified in 348 pregnancies (18%). Major anomalies were more common with increasing hydramnios in both dichorionic and monochorionic twins (P<.001), with anomaly prevalence nearly 20% in cases of severe hydramnios. Severe hydramnios was significantly associated with stillbirth in monochorionic gestations (3 of 11, 27%, P<.001). Hydramnios was not associated with preterm delivery, fetal growth restriction, neonatal intensive care unit admission, or neonatal death in either dichorionic or monochrorionic pregnancies. CONCLUSION: Hydramnios is common in twins, occurring in one of six dichorionic and monochorionic pregnancies. Anomaly prevalence increased with degree of hydramnios; in monochorionic gestations, severe hydramnios was associated with risk of stillbirth. Despite this, adverse outcomes do not appear to be more frequent in the setting of hydramnios in twin getstaions. LEVEL OF EVIDENCE: II.

Placental and uterine pathology in women undergoing peripartum hysterectomy.

OBJECTIVE: To estimate if peripartum hysterectomies performed for intractable uterine atony have pathologic findings consistent with infection more often than those hysterectomies performed for other indications. METHODS: This is a retrospective cohort study of all consecutive peripartum hysterectomies at our institution from 1988 to 2009. Scheduled cesarean hysterectomies were excluded. Maternal, fetal, and pathologic data were obtained by medical record review. Pathologic evaluation was performed for each specimen per a standardized protocol. Patients undergoing hysterectomy for uterine atony were compared with those requiring hysterectomy for another indication. Pearson's χ and Student's t test were used for analysis. RESULTS: Of 324,654 deliveries during the study period, 558 (1.7%) women underwent emergent peripartum hysterectomies; 190 (34%) were for intractable uterine atony. Those requiring hysterectomy for uterine atony were more likely to be at term (87% compared with 62%), have clinical chorioamnionitis (19% compared with 6%), and have longer labors (8 hours compared with 2.5 hours). Certain placental pathologic findings were significantly more common in the atony group, including chorioamnionitis, umbilical vasculitis, chorionic plate vasculitis, and funisitis. Acute endometritis and cervicitis were also more common in the atony group. Conversely, abnormal placental implantation (37% compared with 8%) and leiomyomas (21% compared with 8%) were significantly more common in the group requiring hysterectomy for other indications. CONCLUSION: Patients requiring emergent peripartum hysterectomies as a result of intractable uterine atony are more likely to have clinical and pathologic findings consistent with acute inflammation and infection. LEVEL OF EVIDENCE: II.