Identifying risk factors for peripartum cesarean hysterectomy. A population-based study

To determine the incidence of, and obstetric risk factors for, emergency peripartum hysterectomy.

A population-based study comparing all singleton deliveries between the years 1988 and 1999 that were complicated with peripartum hysterectomy to deliveries without this complication. Statistical analysis was performed with multiple logistic regression analysis.

Emergency peripartum hysterectomy complicated 0.048% (n = 56) of deliveries in the study (n = 117,685). Independent risk factors for emergency peripartum hysterectomy from a backward, stepwise, multivariable logistic regression model were: uterine rupture (OR = 521.4, 95% CI 197.1-1379.7), placenta previa (OR = 8.2, 95% CI 2.2-31.0), postpartum hemorrhage (OR = 33.3, 95% CI 12.6-88.1), cervical tears (OR = 18.0, 95% CI 6.2-52.4), placenta accreta (OR = 13.2, 95% CI 3.5-50.0), second-trimester bleeding (OR = 9.5, 95% CI 2.3-40.1), previous cesarean section (OR = 6.9, 95% CI 3.7-12.8) and grand multiparity (> 5 deliveries) (OR = 3.4, 95% CI 1.8-6.3). Newborns delivered after peripartum hysterectomy had lower Apgar scores (< 7) at 1 and 5 minutes than did others (OR = 11.5, 95% CI 6.2-20.9 and OR = 27.4, 95% CI 11.2-67.4, respectively). In addition, higher rates of perinatal mortality were noted in the uterine hysterectomy vs. the comparison group (OR = 15.9, 95% CI 7.5-32.6). Affected women were more likely than the controls to receive packed-cell transfusions (OR = 457.7, 95% CI 199.2-1105.8) and had lower hemoglobin levels at discharge from the hospital (9.9 +/- 1.3 vs. 12.8 +/- 5.7, P < .001).

Cesarean deliveries in patients with suspected placenta accreta, specifically those performed due to placenta previa in women with a previous uterine scar, should involve specially trained obstetricians. In addition, detailed informed consent about the possibility of emergency peripartum hysterectomy and its associated morbidity should be obtained.