Emergency peripartum hysterectomy, that occurring after vaginal delivery or at the time of cesarean birth, is usually reserved for situations where conservative measures do not control hemorrhage. In the past, the most common indications for emergency peripartum hysterectomy were uterine atony and uterine rupture.1,2 More recent reports list placenta accreta as the most common indication and is most likely related to the increase in the number of cesarean deliveries observed over the past two decades.3–7
The purpose of this study was to estimate the incidence, indications, risk factors, and complications associated with emergency peripartum hysterectomies performed at a community-based academic medical center.
MATERIALS AND METHODS
We conducted a retrospective analysis of all cases of emergency peripartum hysterectomy performed at Winthrop-University Hospital between January 1, 1991, and December 31, 1997. Emergency peripartum hysterectomy was defined as a hysterectomy performed for hemorrhage unresponsive to other treatment within 24 hours of a delivery. Forty-eight hysterectomies were performed; 47 records were available for analysis.
Maternal characteristics such as age, parity, gestational age, previous cesarean delivery, previous uterine curettage, history of antepartum bleeding, and mode of delivery were recorded. The indication for surgery, type of hysterectomy, additional procedures, operating time, pre- and postoperative hemoglobin values, need for blood transfusion, postoperative complications, and postoperative hospitalization days were obtained. The study population was subdivided based on parity, comparing multiparous with primiparous women. In addition, the study group was subdivided and compared based on type of hysterectomy.
Statistical analysis was performed using SAS 6.12 (SAS System, Cary, NC). Wilcoxon rank sum tests were used to compare differences in categoric variables. Dichotomous variables were analyzed using Fisher exact test (two-tail). Cochran-Armitage exact trend test was used to detect trends in incidence of placenta accreta for increasing number of cesarean deliveries and curettages. A P < .05 was considered statistically significant.
During the 7-year study period, 34,241 deliveries were performed with 48 peripartum hysterectomies identified (rate of 1.4 per 1000 deliveries). This rate was similar to published rates of other institutions in the United States. Chestnut et al1 reported 44 hysterectomies in 36,561 deliveries between 1963 and 1983 for a rate of 1.2 per 1000. Clark et al2 identified 70 hysterectomies in 68,653 deliveries between 1978 and 1982 for a rate of 1.02 per 1000. Stanco et al3 studied the same population from 1985 to 1990 and reported 123 hysterectomies in 94,689 deliveries for a rate of 1.3 per 1000. Zelop et al4 described 117 hysterectomies in 75,650 deliveries between 1983 and 1991 for a rate of 1.55 per 1000.
The operative notes and the pathology reports of the uterus and placenta were used to determine the final indication for the procedure. The most common indications were placenta accreta (48.9%), uterine atony only (29.8%), previa without accreta (8.5%), and uterine laceration (4.3%) (Table 1).
The mean maternal age of the study group was 32.3 ± 4.8 years (range 22–43 years). The mean gestational age was 37.8 ± 3.9 weeks (range 20–42 weeks) with a mean birth weight of 3377 ± 1197 g (range 170–5107 g). The mean number of postoperative hospitalization days was 6.8 ± 2.4 (range 4–14 days), the median operating time was 125 minutes (range 62–230 minutes), and the median number of packed red blood cell units transfused was 4 (range 0–84). Forty-four (93.6%) women were delivered by cesarean, 24 (51.1%) women had a previous cesarean delivery, and 22 (46.8%) had a previous curettage. Nineteen (40.4%) women had second- or third-trimester vaginal bleeding, whereas 16 (34%) were diagnosed with placenta previa.
There were 34 (72.3%) multiparous and 13 (27.7%) primiparous women, and the multiparous women were significantly older and of lesser gestational age. Uterine atony was the most common indication for hysterectomy in primiparas, whereas placenta accreta was the most common in multiparas. There were no differences in the number of women receiving blood transfusions, operating time, or history of previous curettage (Table 2).
The most common indication for emergency peripartum hysterectomy was placenta accreta with 23 of 47 (48.9%) patients pathologically diagnosed. Eight of 23 (34.8%) women with no history of cesarean delivery had a placenta accreta, whereas nine of 16 (56%) women with one cesarean and six of eight (75%) women with two cesareans had placenta accreta (P = .052). A similar observation was noted in women with a previous curettage. Nine of 25 (36%) women with no curettage had an accreta compared with seven of 12 (58%) women with one curettage and seven of ten (70%) women with two curettages (P = .071). Twenty-two of 23 (95.6%) women with placenta accreta had a history of previous cesarean delivery or curettage. The only case of a placenta accreta with no previous uterine surgery was in a grand multipara with a placenta previa. Placenta accreta was diagnosed in 12 of 16 (75%) women with preoperative diagnosis of placenta previa compared with 11 of 31 (35.5%) women without placenta previa (P < .02). Four of seven (57.1%) women with placenta previa and no history of cesarean delivery had placenta accreta, whereas four of five (80%) with placenta previa and one cesarean delivery and all four (100%) with two or more cesarean deliveries had placenta accreta.
Four of the 18 hysterectomies performed for uterine atony were found to have placenta accreta. Placental examination in the other 14 revealed histologic chorio-amnionitis (eight), multiple gestation (two), and no pathologic abnormality (four). In four cases where the placental pathology was normal, one woman delivered a large-for-gestational-age neonate, one underwent repeat cesarean delivery at term, one had a septate uterus and underwent cesarean delivery, and one had a primary cesarean delivery for arrest of descent.
There were nine total abdominal hysterectomies and 38 subtotal hysterectomies. A subtotal hysterectomy was done for all 14 cases of uterine atony. Three of the four women with placenta previa without accreta underwent total hysterectomy. No difference in type of hysterectomy was noted when the indication was placenta accreta. There were no differences with regard to the other characteristics studied (Table 3).
Uterine artery ligation was performed in 12 (25.5%) patients (four for uterine atony and eight for placenta accreta). Hypogastric artery ligation was done in three (6.4%) patients. One woman had uterine atony, another had placenta accreta, and the third had placenta previa. Additional procedures performed were three oophorectomies and three cystotomies. Febrile morbidity occurred in 16 (34%) women. Additional complications included severe disseminated intravascular coagulopathy in eight women, one transfusion reaction, one cardiac ischemia, one acute tubular necrosis, one acute respiratory distress syndrome, one pulmonary embolism, one deep vein thrombosis, and one ileus paralyticus.
Early studies on peripartum hysterectomy included hysterectomies done for nonemergent conditions, and between 1950 and the late 1970s cesarean hysterectomy was most commonly used for sterilization, defective uterine scar, myoma, and other gynecologic disorders.1,8–12 Since the 1980s, indications for peripartum hysterectomy have been restricted to emergent situations.2–4
We found placenta accreta to be the most common indication for an emergency peripartum hysterectomy. Chestnut et al1 found that the major indication for the procedure was uterine rupture followed by uterine atony and placenta accreta. Clark et al2 reported uterine atony (43%) to be the most common cause of emergency peripartum hysterectomy followed by placenta accreta (30%) from 1978 to 1982. However, Stanco et al3 studied the same population from 1985 to 1990 and found that placenta accreta (50%) had become the most frequent cause with uterine atony accounting for 21% of cases. Similarly, Zelop et al4 found placenta accreta (64%) and uterine atony (20%) the most common reasons for emergency peripartum hysterectomy. Why has placenta accreta become the most common cause for an emergency peripartum hysterectomy? Firstly, it may be attributed to the increase in cesarean births and uterine curettages over the past two decades.6 Secondly, it may be a result of better treatment of uterine atony with prostaglandin preparations decreasing the need for hysterectomy.
According to Clark et al,2 placenta accreta occurred in 5% of women with placenta previa and an unscarred uterus, and increased to 67% in women with previa and four previous cesarean deliveries.6 Recently, Miller et al7 reported a general incidence of 10% of placenta accreta in women with placenta previa. The incidence of placenta accreta was 2.1% in women less than 35 years old and no previous cesarean delivery, whereas it was 38% in women with two or more cesarean deliveries and complete placenta previa.7
Twenty-two of the 47 (46.8%) women who underwent emergency peripartum hysterectomy had previous curettages. In a review of postpartum hemorrhage, Zahn and Yeomans13 listed history of curettage as a risk factor associated with placenta accreta. Although there is conclusive evidence demonstrating the increased risk of placenta previa after a cesarean delivery,6,7,14 an association between placenta previa and previous curettage has not been clearly shown.15–17 Recently, Ananth et al18 found a strong association between a history of abortion and the subsequent development of placenta previa.
Sonography, color flow Doppler, and magnetic resonance imaging are useful in identifying placenta accreta/percreta. Women with placenta previa and a previous cesarean delivery or uterine surgical procedure (curettage, myomectomy) may benefit from sonographic and color Doppler evaluations. Magnetic resonance imaging is useful in assessing bladder involvement and posterior and lateral uterine wall involvement when the placenta is posterior. If the combination of risk factors and imaging findings is highly suggestive of placenta accreta, then, a cesarean hysterectomy should be planned, as there is reduced maternal morbidity and mortality when done electively.19
In 1985, Clark et al20 reviewed the hospital records of 19 cases of bilateral hypogastric ligation performed after uterine artery ligation and pharmacologic methods were attempted. Ligation was only 42% effective at hemostasis, an increase in blood loss and operating time was noted, as well as an increase in the number of complications (ureteral injury and cardiac arrest). These observations led them to recommend hypogastric ligation only for the hemodynamically stable patient of low parity.20 During the study period at our institution, hypogastric artery ligation was done in three patients, whereas uterine artery ligation was performed in 12. Newer techniques such as hemostatic sutures and uterine or hypogastric artery embolization were not attempted but are an option in attempting uterine conservation.21–24
Should a subtotal hysterectomy be performed for emergency peripartum hemorrhage? We found no differences in preoperative and postoperative hemoglobin, operating time, and blood transfusions given when total and subtotal emergency peripartum hysterectomies were compared. However, there was a trend for more surgical intensive care unit admissions and postoperative complications in the total abdominal hysterectomy group. The earlier literature supports the performance of a total hysterectomy for reduction in potential cervical stump malignancy, need for regular cytology, and other problems such as bleeding or discharge.8,12 In 1963, Tervila reported cancer rates in the retained cervical stump ranging from 0.39% to 1.9%.25 With the advent of cytologic screening, there has been a dramatic decrease in the incidence of cervical cancer. At the present time, the incidence of cervical cancer is reported as 0.1% to 0.15%; the unscreened women are most likely to be affected. Although subtotal hysterectomies were uncommonly done in the studies by Chestnut et al1 and Zelop et al4 (9% and 21%, respectively), Clark et al2 and Stanco et al3 reported 53% of their hysterectomies as subtotal. Eighty percent of the hysterectomies in our report were subtotal.
Subtotal hysterectomy should be a reasonable alternative to total abdominal hysterectomy when an emergent peripartum hysterectomy is being performed.
1. Chestnut DH, Eden RD, Gall SA, Parker RT. Peripartum hysterectomy: A review of cesarean and postpartum hysterectomy. Obstet Gynecol 1985;65:365–70.
2. Clark SL, Yeh SY, Phelan JP, Bruce S, Paul RH. Emergency hysterectomy for obstetric hemorrhage. Obstet Gynecol 1984;64:376–80.
3. Stanco LM, Schrimmer DB, Paul RH, Mishell DR. Emergency peripartum hysterectomy and associated risk factors. Am J Obstet Gynecol 1993;168:879–83.
4. Zelop CM, Harlow BL, Frigoletto FD, Safon LE, Saltzman DH. Emergency peripartum hysterectomy. Am J Obstet Gynecol 1993;168:1443–8.
5. Rates of cesarean delivery—United States, 1991. MMWR 1993;42:285–9.
6. Clark SL, Koonings PP, Phelan JP. Placenta previa/accreta and prior cesarean section. Obstet Gynecol 1985;66:89–92.
7. Miller DA, Chollet JA, Goodwin TM. Clinical risk factors for placenta previa-placenta accreta. Am J Obstet Gynecol 1997;177:210–4.
8. Barclay DL. Cesarean hysterectomy. Obstet Gynecol 1970;35:120–31.
9. Haynes DM, Marton BJ. Cesarean hysterectomy: A twenty-five year review. Am J Obstet Gynecol 1979;134:393–6.
10. Plauche WC, Grunch FG, Bourgeouis MO. Hysterectomy at the time of cesarean section: Analysis of 108 cases. Obstet Gynecol 1981;58:459–64.
11. Sturdee DW, Rushton DI. Caesarean and postpartum hysterectomy 1968–1983. Br J Obstet Gynecol 1986;93:270–4.
12. Thonet RGN. Obstetric hysterectomy—an 11-year experience. Br J Obstet Gynaecol 1986;93:794–8.
13. Zahn CM, Yeomans ER. Postpartum hemorrhage: Placenta accreta, uterine inversion, and puerperal hematomas. Clin Obstet Gynecol 1990;33:422–31.
14. Taylor VM, Kramer MD, Vaughan JL, Peacock S. Placenta previa and prior cesarean delivery: How strong is the association? Obstet Gynecol 1994;84:55–7.
15. Barrett JM, Boehm FH, Killam AP. Induced abortion: A risk factor for placenta previa. Am J Obstet Gynecol 1981; 141:769–72.
16. Grimes DA, Techman T. Legal abortion and placenta previa. Am J Obstet Gynecol 1984;149:501–4.
17. Taylor VM, Kramer MD, Vaughn TL, Peacock S. Placenta previa in relation to induced and spontaneous abortion: A population-based study. Obstet Gynecol 1993;82:88–91.
18. Ananth CV, Smulian JC, Vintzileos AM. The association of placenta previa with history of cesarean delivery and abortion: A metaanalysis. Am J Obstet Gynecol 1997;177:1071–8.
19. Hudon L, Belfort MA, Broome DR. Diagnosis and management of placenta percreta: A review. Obstet Gynecol Surv 1998;53:509–17.
20. Clark SL, Phelan JP, Yeh SY, Bruce SR, Paul RH. Hypogastric artery ligation for obstetric hemorrhage. Obstet Gynecol 1985;66:353–6.
21. Ferguson JE, Bourgeois FJ, Underwood PB. B-Lynch suture for postpartum hemorrhage. Obstet Gynecol 2000; 95:1020–2.
22. Cho JH, Jun HS, Lee CN. Hemostatic suturing technique for uterine bleeding during cesarean delivery. Obstet Gynecol 2000;96:121–31.
23. Deux JF, Bazot M, Le Blanche AF, Tasart M, Khalil A, Berkane N, et al. Is selective embolization of uterine arteries a safe alternative to hysterectomy in patients with postpartum hemorrhage? AJR 2001;177:145–9.
24. Pelage JP, Le Dief O, Jacob D, Soyer P, Herbreteau D, Rymer R. Selective arterial embolization of the uterine arteries in the management of intractable postpartum hemorrhage. Acta Obstet Gynecol Scan 1999;78:698–703.
25. Tervila L. Carcinoma of the cervical stump. Acta Obstet Gynecol Scand 1963;42:200–10.