Archives of Gynecology and Obstetrics

, Volume 274, Issue 2, pp 84–87

Emergency peripartum hysterectomy: a 9-year review


  • Oguz Yucel
    • Department of Obstetrics and Gynecology, Duzce School of MedicineAbant Izzet Baysal University
    • Department of Obstetrics and Gynecology, Duzce School of MedicineAbant Izzet Baysal University
  • Nese Yucel
    • Department of Obstetrics and GynecologyGoztepe Educational Hospital of Social Insurance Association
  • Aslı Somunkiran
    • Department of Obstetrics and Gynecology, Duzce School of MedicineAbant Izzet Baysal University
Original Article

DOI: 10.1007/s00404-006-0124-4

Cite this article as:
Yucel, O., Ozdemir, I., Yucel, N. et al. Arch Gynecol Obstet (2006) 274: 84. doi:10.1007/s00404-006-0124-4


Objective: To determine the incidence, indications, risk factors, and complications of emergency peripartum hysterectomy. Study design: A retrospective study of the patients requiring an emergency peripartum hysterectomy of a 9-year period was conducted. Emergency peripartum hysterectomy was defined as one performed for hemorrhage unresponsive to other treatment less than 24 h after delivery. Demographic and clinical variables were obtained from the maternal records. Results: There were 34 emergency peripartum hysterectomies out of 117,095 deliveries for a rate of 0.29 per 1,000. Of the 16 cases that were delivered by cesarean section, seven had a previous cesarean section and 18 cases were delivered vaginally, including two using vacuum extraction. Total hysterectomy was performed in 24 patients, and subtotal hysterectomy in ten patients. The indications for hysterectomy were uterine rupture (n=12), placenta accreta (n=10), uterine atony (n=7), and hemorrhage (n=5). There were two maternal deaths, six stillbirths, and two early neonatal deaths. Conclusion: This study identified surgical deliveries, uterine rupture, placenta accreta, and uterine atony as risk factors for emergency peripartum hysterectomy. The most common reason for abnormal placental adherence was a previous cesarean section. Multiparity and oxytocin use for uterine stimulation were among the risk factors for uterine atony that necessitated emergency peripartum hysterectomy.


Peripartum hysterectomyMorbidityMortalityEtiologic factors


Obstetrics is a “bloody business.” Hemorrhage can occur antepartum or intrapartum, but mostly develops postpartum. In the past, the most common indications for emergency peripartum hysterectomy were uterine atony and uterine rupture [1, 2]. In the recent years, however, abnormal placentation has become the most common indication due to the greater number of pregnant women who delivered previously by cesarean section [36].

Emergency peripartum hysterectomy, the most dramatic operation in the modern obstetrics, is generally performed when all conservative measures have failed to achieve homeostasis in the setting of life-threatening hemorrhage [3, 7]. The operation is considered one of the most major complications in modern obstetrics and carries a high maternal mortality and morbidity risk [1, 3]. In certain obstetric situations, there is no alternative to emergency hysterectomy and delaying intervention for even a few minutes may prove fatal. It is unplanned and must be performed expeditiously. In addition, the patients are generally in less than ideal condition due to the acute blood loss.

This study reviewed emergency peripartum hysterectomies for a 9-year period and compared the results with published reports worldwide, in order to examine the possible changes in trends and discuss the place of postpartum hysterectomy in modern obstetrics practice.

Materials and methods

We retrospectively reviewed the obstetrics and anesthetic patient files of 34 peripartum hysterectomies performed at two centres from 1995 to 2003. Emergency peripartum hysterectomy was defined as a hysterectomy for a life-saving indication performed for hemorrhage within 24 h of a delivery. The indication for peripartum hysterectomy was an insufficient response to more conservative treatment approaches involving both medical and surgical interventions such as vigorous fundal massage, bimanual uterine compression, use of blood, blood products and replacement fluids, administration oxytocics and prostaglandins, curetting of the placental bed, sutures on the placental bed, and hypogastric artery ligation. Peripartum hysterectomies for cancer and other medical indications or deliveries before 24 weeks of gestation were excluded from the study.

Maternal characteristics such as age, parity, gestational age, previous cesarean delivery, and mode of delivery were recorded. The indication for surgery, type of hysterectomy, additional procedures, operating time, intra- and postoperative complications, blood loss, need for blood transfusion, and days of postoperative hospitalization were obtained. Febrile morbidity was defined as a temperature >38 °C measured at least 24 h after the hysterectomy and repeated at least once. The surgical procedures were performed by resident physicians under the supervision of consultant surgeons.

The Mann–Whitney U test was used for the statistical analysis. P<0.05 was considered significant.


During the study period, 117,095 deliveries were performed with 34 emergency peripartum hysterectomies identified (rate of 0.29 per 1,000 deliveries) and the rate of cesarean section was 19%. The emergency peripartum hysterectomy rate was higher after cesarean section (0.07%) than after vaginal delivery (0.02%).

There were 27 multiparous and seven nulliparous women. The mean age of the patients was 31 years (range 15–44) and the median parity was two (range 0–6). Sixteen cases were delivered by Cesarean section and 18 cases, including two vacuum extractions, were delivered vaginally. The indication for cesarean section was placental abruption in five cases, placenta previa in four, repeated cesarean section in three, fetal distress in two, and cephalo–pelvic disproportion and transverse lie in one case each. The indications for hysterectomy are shown in Table  1. Uterine rupture and placenta accreta were responsible in 64.7% of the patients.
Table 1.

Indications for emergency peripartum hysterectomy


n (%)

Uterine rupture

12 (35.3)

Placenta accreta

10 (29.4)

Uterine atony

7 (20.6)


5 (14.7)

Values presented as n is in percentage

Total and subtotal emergency peripartum hysterectomy is compared in Table  2. There were no statistically significant differences between these two groups in terms of duration of operation, blood loss, transfusion rate, and hospital stay. Blood transfusion was required in 88% of the patients. The mean operating time was 119.6 min, and the mean length of hospitalization was 9.6 days.
Table 2

Comparison of total vs. subtotal emergency peripartum hysterectomy


Total (n=24)

Subtotal (n=10)

All patients (n=34)


Operative time (min.) a

127.7 (60–300)

100 (45–130)



Blood loss (ml) a

1,481 (600–3,600)

2,056 (1,300–2,800)



Transfusions (units) a

4.4 (2–9)

5 (3–8)



Hospital stay (days) a

9.8 (6–21)

9.3 (7–13)



aMean (range)

Pharmacological and surgical maneuvers were first used to control hemorrhage in order to avoid surgery. Accordingly, oxytocin was administered in 16 patients, methergine in nine, and prostaglandin F in six. Curettage was performed on eight patients for suspected retained placental fragments, a cervix laceration was sutured in seven, and the placental bed was sutured in four. The hypogastric artery was ligated in three patients. The remaining patients had no intervention before proceeding to peripartum hysterectomy based on clinical judgment that a delay in hysterectomy would endanger the patient’s life.

There were 35 infants delivered to 34 women with a mean gestational age of 37.9 weeks at delivery (range 27–41). The mean weight at delivery was 2,963 g (range 1,000–4,800). The mean Apgar score was 6 at 1 min and 7 at 5 min. The perinatal mortality rate was 22.8% and consisted of five abruptio placenta and three birth asphyxia, including one premature. The maternal morbidity was high, with febrile morbidity in 26.5% and bladder injury in 8.8% (Table  3). There were two maternal deaths in the immediate postoperative period: one due to embolus and the other of hemorrhagic shock.
Table 3.

Complications associated with emergency peripartum hysterectomy


n (%)

Blood transfusion

30 (88)

Febrile morbidity

9 (26.5)

Perinatal death

8 (22.8)

Bladder injury

3 (8.8)

Maternal death

2 (5.8)

Wound infection

2 (5.8)

Disseminated intravascular coagulopathy

1 (2.9)


1 (2.9)

Vaginal cuff bleeding

1 (2.9)


1 (2.9)

Values presented as n is in percentage

Seven of the 34 cases (20.6%) had undergone previous cesarean section. The indications for repeat cesarean section were placenta previa in three, previous cesarean section in two, and placental abruption in one. One delivered vaginally.


The incidence of peripartum hysterectomy in the literature varies from 0.2 to 1.5 per 1,000 deliveries [15, 814]. In our study, its incidence was 0.29 per 1,000 deliveries, which is in good accord with reported values [2, 4, 1012]. It has been suggested that the high incidence of emergency peripartum hysterectomy can be attributed to the increasing number of cesarean sections, which in turn gives rise to an increased number of abnormal placentation, placenta previa, and scarred uterus [2, 3, 12]. The incidence of emergency peripartum hysterectomy was 2.7 in 1,000 deliveries in a study in USA with a cesarean section rate of 29.2% [15]. However, Zeteroglu et al. [16] reported that the incidence of peripartum hysterectomy in a teaching hospital in the eastern region of Turkey was 5.09 in 1,000 deliveries, which was higher than the previous studies. The authors stated that when referrals from others hospitals were excluded, rate was 2.75/1,000 deliveries. From the same country, involving two different time periods, Zorlu et al. [4] reported the incidence of hysterectomy during 1985–1989 was 0.40/1,000 deliveries and during 1985–1989 it was 0.24/1,000 deliveries with a declining ratio due to the availability of high standard obstetric care. In our study, the incidence was 0.29/1,000 with a cesarean section rate of 19% and this incidence was similar to that in Zorlu’s study, but was very lower to that in Zeteroglu’s study. This discrepancy between the two studies in two distinct parts of our country can be explained with socioeconomic and sociocultural status.

In our seven repeat cesarean section cases, three (42.9%) underwent hysterectomy because of placenta accreta due to placenta previa. This agrees with studies reporting that previous cesarean section is a risk factor for abnormal placental adherence [3, 5]. In placenta previa with accreta, the incidence of peripartum hysterectomy was higher than in placenta previa only [17]. Therefore, in those patients, it is necessary for obstetricians to prepare for the possibility of peripartum hysterectomy due to unexpected massive bleeding during cesarean section. Ultrasound with color Doppler and magnetic resonance imaging (MRI) have been used to detect placenta accreta antenatally [1820]. Persistent blood flows after the latent phase is suspicious for placenta accreta [21]. However, the high cost of MRI and the extensive experience needed for the ultrasound limits their use.

Multiparity and oxytocin use for uterine stimulation are among the risk factors for uterine atony, which could necessitate emergency peripartum hysterectomy [2, 9]. Of the seven-atony cases in our study, six were multiparous, of which five had been given oxytocin for uterine stimulation. The uterine atony rate among our emergency peripartum hysterectomy cases was 20.6%. Clark et al. [2] reported a 43% atony rate in their emergency peripartum hysterectomy series in 1984. From the same institution, Stanco et al. [22] subsequently reported a 20.4% atony rate, which is very close to the 21% reported by Zelop et al. [3] in 1993. Newly developed pharmacological treatment strategies could explain the decrease in uterine atony requiring emergency peripartum hysterectomy [23]. As an indication for emergency peripartum hysterectomy, our placenta accreta rate of 28.4% agreed with the reported rates [4, 23]. In one study with 64% placenta accreta, a 29% cesarean section rate (79% repeat section) was reported [15]. In our study, the cesarean section and repeat cesarean section rates were 19% and 20.6%, respectively. During the study period, newly developed techniques, such as hemostatic sutures and uterine or hypogastric artery embolization, were not attempted, but are an option in attempting uterine conservation [24, 25].

At 35.3%, uterine rupture was the most common indication for emergency peripartum hysterectomy in our study. Rates of uterine rupture from 11.4 to 28% have been reported [4, 9, 15, 22, 26], and Gardeil et al. [8] reported a rate of 45.5%. The two institutions involved in our study are referral hospitals in their geographical areas, where patients are sent when complications arise or usually after the patients become hemodynamically unstable. Of the patients with uterine rupture, 58.3% were referred from other hospitals.

Should a subtotal hysterectomy be performed for emergency peripartum hysterectomy? In the past, total hysterectomy was the recommended surgical method in emergency peripartum hysterectomy due to the potential for cervical stump malignancy, need for regular cytology, and other problems, such as bleeding or discharge. Currently, the proportion of subtotal hysterectomies ranges from 53 [22] to 80% [6]. In most cases, a subtotal hysterectomy, which is associated with less blood loss, a reduced need for blood transfusion, and reduced intra-and postoperative complications, will be sufficient [9]. Both total and subtotal hysterectomy are associated with high maternal mortality [3, 9, 23, 27]. Total hysterectomy should be considered when active bleeding occurs from the lower segment or cervix. We recommend that the choice of subtotal versus total hysterectomy be individualized. All pedicles must be doubly ligated because of hyperemia and peripartum pelvic tissue tear. In our series, the subtotal hysterectomy cases (n=10) had a greater blood loss, necessitating more transfusions than in subjects undergoing total hysterectomy, although their operative times were shorter. Total hysterectomized patients stayed longer in hospital compared with that of the subtotal hysterectomized patients. Both groups were similar with regard to morbidity.

The maternal morbidity in emergency peripartum hysterectomy cases was high, as reported elsewhere [2, 3, 9]. The febrile morbidity rate of 26.5% agrees with other series [1, 4]. Bilateral adnexectomy is performed in a patient to control intraoperative bleeding. Bladder injury occurred in three (8.8%) patients. Our maternal mortality rate of 5.8% in this series compares with Zorlu’s 4.5% [4] and Aboelmagd’s 4% [26]. A study from Nigeria reported 12.5% maternal mortality associated with peripartum hysterectomy [27].

In conclusion, although no risk-assessment system can predict all instances where cesarean delivery will be needed, nevertheless, a significant percentage of the patients who are at high risk for severe hemorrhage and subsequent emergency cesarean hysterectomy can be identified before surgery. These preoperative risk factors could be used to facilitate consultation, referral or transfer of patients before surgery. The operation should be performed by an experienced obstetrician before the patient’s condition is extreme. Proper surgical interventions such as hemostatic sutures and uterine or hypogastric artery embolization are an option in attempting uterine conservation for the hemodynamically stable patients of low parity in whom future fertility is important.

Copyright information

© Springer-Verlag 2006