Journal of Obstetrics and Gynaecology, 2014; Early Online: 1–4 © 2014 Informa UK, Ltd. ISSN 0144-3615 print/ISSN 1364-6893 online DOI: 10.3109/01443615.2014.961907
Peripartum hysterectomy in a tertiary hospital in Western Sydney A. Shamsa, A. Harris & A. Anpalagan
J Obstet Gynaecol Downloaded from informahealthcare.com by Nyu Medical Center on 05/02/15 For personal use only.
Department of Obstetrics and Gynaecology, Westmead Hospital, Westmead, NSW, Australia
The aim of this study was to review the incidence, indication, management and complications of peripartum hysterectomy (PH) in a tertiary level hospital over a decade. A retrospective review of all cases of PH performed at Westmead Hospital, Western Sydney, 2003–2012, was undertaken. PH was defined as hysterectomy performed after 20 weeks’ gestation or any time after delivery but within 6 weeks’ postpartum (Awan et al. 2011). There were 56 cases of peripartum hysterectomy of 46,177 births, a rate of 1.22 per 1,000 births. The most common indication for PH was morbid adherence of the placenta (58.2%) followed by uterine atony. Having a history of both caesarean section and placenta praevia is highly associated with a morbidly adherent placenta in the index pregnancy (p ⴝ 0.002). The most common complication was coagulopathy followed by febrile illness and urinary tract injury. Our data showed previous caesarean section and placenta praevia to be associated with abnormal placentation, the leading indication for PH. Since there is an association between a planned caesarean hysterectomy and reduced amount of estimated blood loss and blood transfused, the knowledge of placentation and adequate preoperative planning and consideration for elective hysterectomy could be beneficial. The morbidity associated with PH is considerable. Keywords: Abnormal placentation, critical care obstetrics, obstetric haemorrhage, peripartum hysterectomy
Introduction According to the most recent triennial report of maternal deaths in Australia, severe obstetric haemorrhage was shown to be the fourth leading cause of maternal mortality in Australia (Sullivan 2008). Peripartum hysterectomy (PH), a surgical procedure of last resort, is performed at the time of delivery or postpartum, when a sequence of conservative measures has failed to control uterine haemorrhage. Apart from the obvious effect on future fertility and the emotional impact on the survivors, there is considerable morbidity associated with peripartum hysterectomy (de la Cruz et al. 2013; Wright et al. 2010a, b). Studies on peripartum hysterectomy have shown that the incidence is increasing (Haynes 2004). Furthermore, primary and repeat caesarean sections as well as vaginal births after caesarean section (VBAC) have been associated with peripartum hysterectomy (Knight et al. 2008; Bodelon et al. 2009; Wu et al. 2005). Literature suggests that the national caesarean section rate is increasing (Li 2013), therefore, there is an urgent need to explore the epidemiology and management of peripartum hysterectomy in Australia. The incidence of PH varies from 0.2 to 8.9 per 1,000 births and has been reported to be 0.85 per 1,000 births in a tertiary
hospital in Sydney (Awan et al. 2011; Machado 2011). Similar results have been reported in the USA, Canada and Europe (Bodelon et al. 2009; Zwart et al. 2010). Although rare, emergency peripartum hysterectomy (EPH) is increasing in the developed world (Haynes 2004; Bodelon et al. 2009; Whiteman et al. 2006; Yoong et al. 2006; Muench et al. 2008). The main causes of the uncontrollable haemorrhage necessitating an EPH have changed since the 1980s (Kwee et al. 2006). Abnormal placentation has been reported to be the main cause in recent literature (Jones et al. 2013; Tadesse et al. 2011). This is due to an actual increase in the incidence of the morbidly adherent placenta (Wu et al. 2005), as well as reduced incidence of other causes including uterine atony and uterine rupture, as a result of advances in conservative measures. The objective of this study was to review the incidence, indication, management and complications of peripartum hysterectomy in a tertiary level hospital over a decade.
Materials and methods A retrospective review of all cases of peripartum hysterectomy performed at Westmead Hospital, a tertiary hospital, Western Sydney, NSW in the decade between January 2003 and December 2012, was undertaken. The ICD-10 procedure coding system was used to identify cases which were also verified by cross-reference with the ObstetriX database, operating theatre database and histopathology records. Further details were obtained by reviewing the electronic medical records and patients’ medical record. Peripartum hysterectomy was defined as hysterectomy performed after 20 weeks’ gestation or any time after delivery but within 6 weeks postpartum. A total of 56 women underwent peripartum hysterectomy over the 10-year period. Of the 56, 10 patients underwent a planned caesarean hysterectomy. The remaining 46 had an emergency caesarean hysterectomy for uncontrollable uterine bleeding not responsive to conservative measures. Operation details of one patient were not obtained due to missing patient’s medical records. Information abstracted from the obstetric database and patient medical records included: demographic data; past obstetric, medical and surgical history; details of the pregnancy and delivery; details of haemorrhage management and hysterectomy; complications and outcomes. Statistical analysis was performed using SPSS 21.0 (SPSS Inc., Chicago, IL). Means and standard deviations were reported for continuous variables unless skewed, in which case medians were used. The χ2-test or Fisher’s exact test, as appropriate, were used to test for association between categorical variables. Mann–Whitney was used to test for differences in the distribution of continuous or ordinal categorical variables by outcome. Two-tailed tests with
Correspondence: Department of Obstetrics and Gynaecology, Westmead Hospital, Darcy Road, Westmead, 2145 Australia. E-mail: [email protected]
2
A. Shamsa et al.
a significance level of 5% were used throughout. Statistical significance was taken at p ⬍ 0.05. Ethical approval for this study was granted by the Western Sydney Area Health Service Committee.
J Obstet Gynaecol Downloaded from informahealthcare.com by Nyu Medical Center on 05/02/15 For personal use only.
Results The results reported are based on an analysis of the 56 case records available, excluding the operation details of one case. During the 10-year period, January 2003 to December 2012, there were a total of 46,177 deliveries and 56 peripartum hysterectomies at Westmead Hospital. Of these, there were 10 elective caesarean hysterectomies. The caesarean section rate rose from 26.6% in 2003 to 30.1% in 2012. The incidence of peripartum hysterectomy was 1.21 per 1,000 births. However, the incidence of emergency peripartum hysterectomy was 0.99 per 1,000 births. The total number of caesarean sections during the study period was 13,801, giving a rate of 29.8%. Among those delivered by caesarean section, the rate of emergency peripartum hysterectomy was 3.3 per 1,000 deliveries. The mean maternal age of the study group was 34.6 ⫾ 4.7; mean gestational age was 37 ⫾ 3.6; there were nine primiparous and 47 multiparous women in the group (Table I). The most common indication for peripartum hysterectomy in our study population was morbidly adherent placenta (58.2%) followed by uterine atony (Table II). Table III presents medical and surgical managements attempted prior to hysterectomy. Among women who had a peripartum hysterectomy, the rate of abnormal placentation changed significantly over the time period from 33.3% in 2003 to 77.8% in 2012 (p ⫽ 0.012). (Figure 1) There is a significant association between the history of a previous caesarean section and the number of previous caesarean sections with having a morbidly adherent placenta in the index pregnancy (p ⫽ 0.001). Despite our analysis showing no association between previous caesarean delivery and increased risk of placenta praevia, having a history of both caesarean and placenta praevia is highly associated with morbidly adherent placenta in an index pregnancy (p ⫽ 0.002). There were 15 cases where women had a history of caesarean section and placenta praevia in the index pregnancy, which resulted in emergency peripartum hysterectomy. Using logistic regression to examine the joint effect
Table I. Demography of women and details of deliveries that required hysterectomy. Clinical history Maternal age (years, mean ⫾ SD) Gravidity (median, IQR) Parity (median, IQR) Previous caesarean section (%) ⱖ 2 previous caesarean section (%) Previous uterine curettage (%) Previous uterine surgery (including caesarean section, uterine curettage and myomectomy) (%) Delivery Gestational age (weeks) (median, IQR) Laboured (%) Vaginal delivery (%) Elective caesarean delivery (%) Emergency caesarean delivery (%) Classical uterine incision (%) Midline skin incision (%) Elective (planned) hysterectomy (%)
34.6 ⫾ 4.7 4 (2–5) 2 (1–3) 66.1 28.6 43.1 78.6
37 ⫹ 6 (36 ⫹ 1 to 39 ⫹ 2) 39.3 16.1 55.3 28.6 14.3 30.9 17.9
Table II. Indications for peripartum hysterectomy. Cause of haemorrhage
n
(%)
Morbid adherence of the placenta∗ Placenta accrete Placenta increta Placenta percreta Placenta praevia Uterine atony† Extension of incision or tear at caesarean section Uterine rupture Cervical laceration Cervical cancer
32 10 13 9 3 9 5 4 2 1
57.1
5.4 16.1 8.9 7.1 3.6 1.8
∗24 cases of morbidly adherent placenta also had placenta praevia. †In a total of 30, uterine atony was a contributing cause of haemorrhage.
of placenta praevia and previous caesarean section on the occurrence of abnormal placentation, there is no statistically significant interaction between the effects of these variables (p ⫽ 0.287). The odds of abnormal placentation are 23 times higher in those with placenta praevia than in those without, after adjusting for the effects of previous caesarean section. After adjusting for the presence of placenta praevia, the odds are 4 times higher when there is a history of previous caesarean section. There was no significant association noted in our study group, between morbid adherence of the placenta and previous history of curettage. However, in the study population, curettage was associated with a higher rate of placenta praevia (p ⫽ 0.02). Mode of delivery, vaginal vs caesarean and emergency vs elective caesarean, did not influence the amount of blood loss or units of blood transfused. The total estimated blood loss included intrapartum, postpartum, intraoperative and postoperative periods. The Mean estimated blood loss was 4,427.8 ⫾ 2,931.5.8 ml, with a median of 3,500 ml. Four women did not require any blood transfusion, all of whom had an elective caesarean hysterectomy. A total of 23 (41%) women were transfused more than 10 units of packed red blood cells. The mean units of packed red blood cells transfused were 9.1 ⫾ 8.1 (range of 0–35) units. The Mean estimated blood loss in a planned caesarean hysterectomy was 2,000 ml compared with 4,000 ml in an emergency hysterectomy. When comparing the morbidity of planned peripartum hysterectomy to that of emergency cases, there is an association between a planned caesarean hysterectomy and reduced amount of blood transfused (p ⫽ 0.001). In women who underwent peripartum hysterectomy, having a morbidly adherent placenta or placenta praevia was not associated with higher estimated blood loss or quantity of blood transfused. Table III. Interventions made prior to hysterectomy. Intervention
n
(%)
Oxytocins Ergometrine/Syntometrine use Prostaglandins Haemostatic sutures Pressure (uterine pack of balloon tamponade) Ligation of uterine arteries Ligation of internal iliac arteries Recombinant activated factor VII Placenta left in utero Examination under anaesthesia Uterine artery embolisation B-Lynch suture
46 15 28 8 17 5 3 8 11 9 1 1
82.1 27.3 50 14.5 30.9 9.1 5.5 15.1 20 17 1.9 1.9
Peripartum hysterectomy in a tertiary hospital in Western Sydney 120 Rate of CS among all deliveries (%)
100
Rate of abnormal placentation in PH group (%)
80
60
40
20
0 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012
J Obstet Gynaecol Downloaded from informahealthcare.com by Nyu Medical Center on 05/02/15 For personal use only.
Figure 1. Trends of caesarean section (CS) and abnormal placentation in the study group, 2003–2012.
The complications related to peripartum hysterectomy include blood transfusions, febrile morbidity and operative complications. There were seven cases of urinary tract injury, all of whom had a previous caesarean section except for one. Our results did not show any significant difference in the occurrence of urinary tract injury between planned caesarean hysterectomies and emergency procedures. Other complications included fever and disseminated intravascular coagulopathy (Table IV). In total, 26 (46.4%) women were transferred to intensive care unit. There were no maternal deaths in our study group.
Discussion The rare nature of peripartum hysterectomy has limited the number and types of studies that have been possible. Moreover, due to the absence of a national database, there is a lack of Australian data on PH. The incidence of emergency peripartum hysterectomy in our tertiary centre, 0.99 per 1,000 births, is comparable with other tertiary centres in NSW (Awan et al. 2011), the USA (Whiteman et al. 2006) and the UK (Jones et al. 2013), however it was higher than that reported from population-based studies of the UK, USA and Ireland (Knight et al. 2008; Flood et al. 2009; Bodelon et al. 2009). This variation might be due to the different periods of study, considering the changing trend of PH (Knight et al. 2008; Flood et al. 2009; Haynes 2004), and the definition of peripartum hysterectomy across studies. The main discrepancy has been in the definition of the postpartum period, which has varied between the immediate postpartum period (Ehtisham 2011), Table IV. Complications associated with hysterectomy. Complication Coagulopathy Fever Urinary tract injury Wound infection Oophorectomy Salpingectomy Pulmonary oedema Retro-peritoneal haematoma Paralytic ileus Return to operating theatre Cardiac arrest Maternal death
n
(%)
12 10 7 2 2 1 1 1 1 3 1 0
22.2 18.9 13.5 3.6 3.6 1.9 1.9 1.9 1.9 5.3 1.9
3
during the same hospitalisation (Whiteman et al. 2006); within 24–72 h (Francois et al. 2005; Sahin et al. 2014; Umezurike et al. 2008; Tadesse 2011); 30 days postpartum (Bodelon et al. 2009) and 6 weeks postpartum (Awan et al. 2011). Furthermore, some studies have excluded planned hysterectomies and those due to malignancies or structural abnormalities (Haynes 2004); hence it is difficult to compare the incidence reported in different studies. More than half of the patients in this study had evidence of morbidly adherent placenta, which would be considered the leading indication for hysterectomy and is similar to other studies (Machado 2011; Flood et al. 2009). Moreover, 50% had placenta praevia, provided we include praevia associated with accreta, and the frequency of these indications is in accordance with recent literature (Ehtisham 2011). There is a high rate of caesarean section as the mode of delivery in our PH cases, 83.1%. Furthermore, our data shows a previous caesarean section to be associated with abnormal placentation, the leading indication for PH. Considering the increasing rate of caesarean section in New South Wales, from 26.1% to 30.1% during a decade, there is concern about the increasing rate of PH and its complications. Clinicians can contribute to reducing this event by preventing unnecessary first caesarean sections. This may be facilitated by adhering to appropriate definitions for failed induction and arrest of labour progress, training and experience in operative vaginal delivery and including effects on future reproductive health when discussing the first caesarean delivery with patients (Spong et al. 2012). Whether encouraging vaginal birth after caesarean section (VBAC) can affect the rate of PH is unknown, as a case–control study, comparing PH following VBAC and elective caesarean section, would be required to show this effect. However, in theory, a greater number of previous caesarean sections can lead to higher incidence of abnormal placentation – the main cause of peripartum hysterectomy reported in many recent studies. Our study has limitations, the main one being that it was conducted in a single tertiary centre, which can potentially limit its generalisability. We also acknowledge the limitations due to the descriptive nature of the study. It is notable that extraction of data from operative reports and patient records may improve its accuracy as compared with population-based studies where data is obtained from a national electronic database. Also, the effectiveness of the B-Lynch suture, activate factor VII and uterine artery embolisation prior to hysterectomy was not evaluated. Our study highlights the significance of knowledge of placentation and adequate preoperative planning and consideration for elective hysterectomy in specific cases, i.e. placenta accreta. As this has shown to significantly reduce the amount of blood loss and blood products required which may impact patient morbidity. Preoperative planning would include having senior clinicians present, close liaison with blood bank and timing of operation to meet adequate resource availability. Furthermore, the higher rate of elective caesarean sections as compared with emergency caesarean sections is attributed to a high rate of planned caesarean hysterectomies in our institution. We would recommend a cohort study, considering a randomised control trail would not be feasible, to be performed to compare the conservative measures utilised in severe obstetric haemorrhages that led to hysterectomy with those which succeeded in controlling the haemorrhage without this surgical intervention. Also, a national birth register would be of paramount value to improve information on the characteristics of Australian women who undergo a peripartum hysterectomy. This would provide an estimate of the national incidence of peripartum hysterectomy, describe the trend and risk factors and guide clinicians
4
A. Shamsa et al.
in decision-making and appropriate counselling of patients at an increased risk. Declaration of interest: The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.
J Obstet Gynaecol Downloaded from informahealthcare.com by Nyu Medical Center on 05/02/15 For personal use only.
References Awan N, Bennett MJ, Walters WA. 2011. Emergency peripartum hysterectomy: a 10-year review at the Royal Hospital for Women, Sydney. Australian and New Zealand Journal of Obstetrics and Gynaecology 51:210–215. Bodelon C, Bernabe-Ortiz A, Schiff MA, Reed SD. 2009. Factors associated with peripartum hysterectomy. Obstetrics and Gynecology 114:115–123. de la Cruz CZ, Coulter ML, O’Rourke K, Amina Alio P, Daley EM, Mahan CS. 2013. Women’s experiences, emotional responses, and perceptions of care after emergency peripartum hysterectomy: a qualitative survey of women from 6 months to 3 years postpartum. Birth 40:256–263. Ehtisham S. 2011. Emergency peripartum hysterectomy. Pakistan Journal of Surgery 27:288–291. Flood KM, Said S, Geary M, Robson M, Fitzpatrick C, Malone FD. 2009. Changing trends in peripartum hysterectomy over the last 4 decades. American Journal of Obstetrics and Gynecology 200:632.e1–e6. Francois K, Ortiz J, Harris C, Foley MR, Elliott JP. 2005. Is peripartum hysterectomy more common in multiple gestations? Obstetrics and Gynecology 105:1369–1372. Haynes K, Stone C, King J. 2004. Major morbidities associated with childbirth in victoria: obstetric haemorrhage and associated hysterectomy. Melbourne: Public Health Group, Department of Human Services. Jones B, Zhang E, Alzouebi A, Robbins T, Paterson-Brown S, Prior T, Kumar S. 2013. Maternal and perinatal outcomes following peripartum hysterectomy from a single tertiary centre. Australian and New Zealand Journal of Obstetrics and Gynaecology 53:561–565. Knight M, Kurinczuk JJ, Spark P, Brocklehurst P; United Kingdom Obstetric Surveillance System Steering Committee. 2008. Cesarean delivery and peripartum hysterectomy. Obstetrics and Gynecology 111:97–105. Kwee A, Bots ML, Visser GH, Bruinse HW. 2006. Emergency peripartum hysterectomy: A prospective study in The Netherlands. European Journal of Obstetrics, Gynecology, and Reproductive Biology 124:187–192.
Li Z, Zeki R, Hilder L, Sullivan EA. 2013. Australia’s mothers and babies 2011. Perinatal statistics, Series No. 28, Cat. No. PER 59. Canberra: AIHW National Perinatal Epidemiology and Statistics Unit. Machado LS. 2011. Emergency peripartum hysterectomy: Incidence, indications, risk factors and outcome. North American Journal of Medical Sciences 3:358–361. Muench MV, Baschat AA, Oyelese Y, Kush ML, Mighty HE, Malinow AM. 2008. Gravid hysterectomy: a decade of experience at an academic referral center. Journal of Reproductive Medicine 53:271–278. Sahin S, Guzin K, Eroğlu M, Kayabasoglu F, Yaşartekin MS. 2014. Emergency peripartum hysterectomy: our 12-year experience. Archives of Gynecology and Obstetrics 289:953–958. Spong CY, Berghella V, Wenstrom KD, Mercer BM, Saade GR. 2012. Preventing the first cesarean delivery: summary of a joint Eunice Kennedy Shriver National Institute of Child Health and Human Development, Society for Maternal-Fetal Medicine, and American College of Obstetricians and Gynecologists Workshop. Obstetrics and Gynecology 120:1181–1193. Sullivan E, Hall B, King JF. 2008. Maternal deaths in Australia 2003–2005. Sydney: AIHW National Perinatal Statistics Unit, Series No. 3. Cat. No. PER 42. Tadesse W, Farah N, Hogan J, d’Arcy T, Kennelly M, Turner MJ. 2011. Peripartum hysterectomy in the first decade of the 21st century. Journal of Obstetrics and Gynaecology 31:320–321. Umezurike CC, Feyi-Waboso PA, Adisa CA. 2008. Peripartum hysterectomy in Aba southeastern Nigeria. Australian and New Zealand Journal of Obstetrics and Gynaecology 48:580–582. Whiteman MK, Kuklina E, Hillis SD, Jamieson DJ, Meikle SF, Posner SF, Marchbanks PA. 2006. Incidence and determinants of peripartum hysterectomy. Obstetrics and Gynecology 108:1486–1492. Wright JD, Bonanno C, Shah M, Gaddipati S, Devine P. 2010a. Peripartum hysterectomy. Obstetrics and Gynecology 116:429–434. Wright JD, Devine P, Shah M, Gaddipati S, Lewin SN, Simpson LL et al. 2010b. Morbidity and mortality of peripartum hysterectomy. Obstetrics and Gynecology 115:1187–1193. Wu S, Kocherginsky M, Hibbard JU. 2005. Abnormal placentation: twentyyear analysis. American Journal of Obstetrics and Gynecology 192: 1458–1461. Yoong W, Massiah N, Oluwu A. 2006. Obstetric hysterectomy: changing trends over 20 years in a multiethnic high risk population. Archives of Gynecology and Obstetrics 274:37–40. Zwart JJ, Dijk PD, van Roosmalen J. 2010. Peripartum hysterectomy and arterial embolization for major obstetric hemorrhage: a 2-year nationwide cohort study in the Netherlands. American Journal of Obstetrics and Gynecology 202:150.e1–e7.
Peripartum hysterectomy in a tertiary hospital in Western Sydney A. Shamsa, A. Harris & A. Anpalagan
J Obstet Gynaecol Downloaded from informahealthcare.com by Nyu Medical Center on 05/02/15 For personal use only.
Department of Obstetrics and Gynaecology, Westmead Hospital, Westmead, NSW, Australia
The aim of this study was to review the incidence, indication, management and complications of peripartum hysterectomy (PH) in a tertiary level hospital over a decade. A retrospective review of all cases of PH performed at Westmead Hospital, Western Sydney, 2003–2012, was undertaken. PH was defined as hysterectomy performed after 20 weeks’ gestation or any time after delivery but within 6 weeks’ postpartum (Awan et al. 2011). There were 56 cases of peripartum hysterectomy of 46,177 births, a rate of 1.22 per 1,000 births. The most common indication for PH was morbid adherence of the placenta (58.2%) followed by uterine atony. Having a history of both caesarean section and placenta praevia is highly associated with a morbidly adherent placenta in the index pregnancy (p ⴝ 0.002). The most common complication was coagulopathy followed by febrile illness and urinary tract injury. Our data showed previous caesarean section and placenta praevia to be associated with abnormal placentation, the leading indication for PH. Since there is an association between a planned caesarean hysterectomy and reduced amount of estimated blood loss and blood transfused, the knowledge of placentation and adequate preoperative planning and consideration for elective hysterectomy could be beneficial. The morbidity associated with PH is considerable. Keywords: Abnormal placentation, critical care obstetrics, obstetric haemorrhage, peripartum hysterectomy
Introduction According to the most recent triennial report of maternal deaths in Australia, severe obstetric haemorrhage was shown to be the fourth leading cause of maternal mortality in Australia (Sullivan 2008). Peripartum hysterectomy (PH), a surgical procedure of last resort, is performed at the time of delivery or postpartum, when a sequence of conservative measures has failed to control uterine haemorrhage. Apart from the obvious effect on future fertility and the emotional impact on the survivors, there is considerable morbidity associated with peripartum hysterectomy (de la Cruz et al. 2013; Wright et al. 2010a, b). Studies on peripartum hysterectomy have shown that the incidence is increasing (Haynes 2004). Furthermore, primary and repeat caesarean sections as well as vaginal births after caesarean section (VBAC) have been associated with peripartum hysterectomy (Knight et al. 2008; Bodelon et al. 2009; Wu et al. 2005). Literature suggests that the national caesarean section rate is increasing (Li 2013), therefore, there is an urgent need to explore the epidemiology and management of peripartum hysterectomy in Australia. The incidence of PH varies from 0.2 to 8.9 per 1,000 births and has been reported to be 0.85 per 1,000 births in a tertiary
hospital in Sydney (Awan et al. 2011; Machado 2011). Similar results have been reported in the USA, Canada and Europe (Bodelon et al. 2009; Zwart et al. 2010). Although rare, emergency peripartum hysterectomy (EPH) is increasing in the developed world (Haynes 2004; Bodelon et al. 2009; Whiteman et al. 2006; Yoong et al. 2006; Muench et al. 2008). The main causes of the uncontrollable haemorrhage necessitating an EPH have changed since the 1980s (Kwee et al. 2006). Abnormal placentation has been reported to be the main cause in recent literature (Jones et al. 2013; Tadesse et al. 2011). This is due to an actual increase in the incidence of the morbidly adherent placenta (Wu et al. 2005), as well as reduced incidence of other causes including uterine atony and uterine rupture, as a result of advances in conservative measures. The objective of this study was to review the incidence, indication, management and complications of peripartum hysterectomy in a tertiary level hospital over a decade.
Materials and methods A retrospective review of all cases of peripartum hysterectomy performed at Westmead Hospital, a tertiary hospital, Western Sydney, NSW in the decade between January 2003 and December 2012, was undertaken. The ICD-10 procedure coding system was used to identify cases which were also verified by cross-reference with the ObstetriX database, operating theatre database and histopathology records. Further details were obtained by reviewing the electronic medical records and patients’ medical record. Peripartum hysterectomy was defined as hysterectomy performed after 20 weeks’ gestation or any time after delivery but within 6 weeks postpartum. A total of 56 women underwent peripartum hysterectomy over the 10-year period. Of the 56, 10 patients underwent a planned caesarean hysterectomy. The remaining 46 had an emergency caesarean hysterectomy for uncontrollable uterine bleeding not responsive to conservative measures. Operation details of one patient were not obtained due to missing patient’s medical records. Information abstracted from the obstetric database and patient medical records included: demographic data; past obstetric, medical and surgical history; details of the pregnancy and delivery; details of haemorrhage management and hysterectomy; complications and outcomes. Statistical analysis was performed using SPSS 21.0 (SPSS Inc., Chicago, IL). Means and standard deviations were reported for continuous variables unless skewed, in which case medians were used. The χ2-test or Fisher’s exact test, as appropriate, were used to test for association between categorical variables. Mann–Whitney was used to test for differences in the distribution of continuous or ordinal categorical variables by outcome. Two-tailed tests with
Correspondence: Department of Obstetrics and Gynaecology, Westmead Hospital, Darcy Road, Westmead, 2145 Australia. E-mail: [email protected]
2
A. Shamsa et al.
a significance level of 5% were used throughout. Statistical significance was taken at p ⬍ 0.05. Ethical approval for this study was granted by the Western Sydney Area Health Service Committee.
J Obstet Gynaecol Downloaded from informahealthcare.com by Nyu Medical Center on 05/02/15 For personal use only.
Results The results reported are based on an analysis of the 56 case records available, excluding the operation details of one case. During the 10-year period, January 2003 to December 2012, there were a total of 46,177 deliveries and 56 peripartum hysterectomies at Westmead Hospital. Of these, there were 10 elective caesarean hysterectomies. The caesarean section rate rose from 26.6% in 2003 to 30.1% in 2012. The incidence of peripartum hysterectomy was 1.21 per 1,000 births. However, the incidence of emergency peripartum hysterectomy was 0.99 per 1,000 births. The total number of caesarean sections during the study period was 13,801, giving a rate of 29.8%. Among those delivered by caesarean section, the rate of emergency peripartum hysterectomy was 3.3 per 1,000 deliveries. The mean maternal age of the study group was 34.6 ⫾ 4.7; mean gestational age was 37 ⫾ 3.6; there were nine primiparous and 47 multiparous women in the group (Table I). The most common indication for peripartum hysterectomy in our study population was morbidly adherent placenta (58.2%) followed by uterine atony (Table II). Table III presents medical and surgical managements attempted prior to hysterectomy. Among women who had a peripartum hysterectomy, the rate of abnormal placentation changed significantly over the time period from 33.3% in 2003 to 77.8% in 2012 (p ⫽ 0.012). (Figure 1) There is a significant association between the history of a previous caesarean section and the number of previous caesarean sections with having a morbidly adherent placenta in the index pregnancy (p ⫽ 0.001). Despite our analysis showing no association between previous caesarean delivery and increased risk of placenta praevia, having a history of both caesarean and placenta praevia is highly associated with morbidly adherent placenta in an index pregnancy (p ⫽ 0.002). There were 15 cases where women had a history of caesarean section and placenta praevia in the index pregnancy, which resulted in emergency peripartum hysterectomy. Using logistic regression to examine the joint effect
Table I. Demography of women and details of deliveries that required hysterectomy. Clinical history Maternal age (years, mean ⫾ SD) Gravidity (median, IQR) Parity (median, IQR) Previous caesarean section (%) ⱖ 2 previous caesarean section (%) Previous uterine curettage (%) Previous uterine surgery (including caesarean section, uterine curettage and myomectomy) (%) Delivery Gestational age (weeks) (median, IQR) Laboured (%) Vaginal delivery (%) Elective caesarean delivery (%) Emergency caesarean delivery (%) Classical uterine incision (%) Midline skin incision (%) Elective (planned) hysterectomy (%)
34.6 ⫾ 4.7 4 (2–5) 2 (1–3) 66.1 28.6 43.1 78.6
37 ⫹ 6 (36 ⫹ 1 to 39 ⫹ 2) 39.3 16.1 55.3 28.6 14.3 30.9 17.9
Table II. Indications for peripartum hysterectomy. Cause of haemorrhage
n
(%)
Morbid adherence of the placenta∗ Placenta accrete Placenta increta Placenta percreta Placenta praevia Uterine atony† Extension of incision or tear at caesarean section Uterine rupture Cervical laceration Cervical cancer
32 10 13 9 3 9 5 4 2 1
57.1
5.4 16.1 8.9 7.1 3.6 1.8
∗24 cases of morbidly adherent placenta also had placenta praevia. †In a total of 30, uterine atony was a contributing cause of haemorrhage.
of placenta praevia and previous caesarean section on the occurrence of abnormal placentation, there is no statistically significant interaction between the effects of these variables (p ⫽ 0.287). The odds of abnormal placentation are 23 times higher in those with placenta praevia than in those without, after adjusting for the effects of previous caesarean section. After adjusting for the presence of placenta praevia, the odds are 4 times higher when there is a history of previous caesarean section. There was no significant association noted in our study group, between morbid adherence of the placenta and previous history of curettage. However, in the study population, curettage was associated with a higher rate of placenta praevia (p ⫽ 0.02). Mode of delivery, vaginal vs caesarean and emergency vs elective caesarean, did not influence the amount of blood loss or units of blood transfused. The total estimated blood loss included intrapartum, postpartum, intraoperative and postoperative periods. The Mean estimated blood loss was 4,427.8 ⫾ 2,931.5.8 ml, with a median of 3,500 ml. Four women did not require any blood transfusion, all of whom had an elective caesarean hysterectomy. A total of 23 (41%) women were transfused more than 10 units of packed red blood cells. The mean units of packed red blood cells transfused were 9.1 ⫾ 8.1 (range of 0–35) units. The Mean estimated blood loss in a planned caesarean hysterectomy was 2,000 ml compared with 4,000 ml in an emergency hysterectomy. When comparing the morbidity of planned peripartum hysterectomy to that of emergency cases, there is an association between a planned caesarean hysterectomy and reduced amount of blood transfused (p ⫽ 0.001). In women who underwent peripartum hysterectomy, having a morbidly adherent placenta or placenta praevia was not associated with higher estimated blood loss or quantity of blood transfused. Table III. Interventions made prior to hysterectomy. Intervention
n
(%)
Oxytocins Ergometrine/Syntometrine use Prostaglandins Haemostatic sutures Pressure (uterine pack of balloon tamponade) Ligation of uterine arteries Ligation of internal iliac arteries Recombinant activated factor VII Placenta left in utero Examination under anaesthesia Uterine artery embolisation B-Lynch suture
46 15 28 8 17 5 3 8 11 9 1 1
82.1 27.3 50 14.5 30.9 9.1 5.5 15.1 20 17 1.9 1.9
Peripartum hysterectomy in a tertiary hospital in Western Sydney 120 Rate of CS among all deliveries (%)
100
Rate of abnormal placentation in PH group (%)
80
60
40
20
0 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012
J Obstet Gynaecol Downloaded from informahealthcare.com by Nyu Medical Center on 05/02/15 For personal use only.
Figure 1. Trends of caesarean section (CS) and abnormal placentation in the study group, 2003–2012.
The complications related to peripartum hysterectomy include blood transfusions, febrile morbidity and operative complications. There were seven cases of urinary tract injury, all of whom had a previous caesarean section except for one. Our results did not show any significant difference in the occurrence of urinary tract injury between planned caesarean hysterectomies and emergency procedures. Other complications included fever and disseminated intravascular coagulopathy (Table IV). In total, 26 (46.4%) women were transferred to intensive care unit. There were no maternal deaths in our study group.
Discussion The rare nature of peripartum hysterectomy has limited the number and types of studies that have been possible. Moreover, due to the absence of a national database, there is a lack of Australian data on PH. The incidence of emergency peripartum hysterectomy in our tertiary centre, 0.99 per 1,000 births, is comparable with other tertiary centres in NSW (Awan et al. 2011), the USA (Whiteman et al. 2006) and the UK (Jones et al. 2013), however it was higher than that reported from population-based studies of the UK, USA and Ireland (Knight et al. 2008; Flood et al. 2009; Bodelon et al. 2009). This variation might be due to the different periods of study, considering the changing trend of PH (Knight et al. 2008; Flood et al. 2009; Haynes 2004), and the definition of peripartum hysterectomy across studies. The main discrepancy has been in the definition of the postpartum period, which has varied between the immediate postpartum period (Ehtisham 2011), Table IV. Complications associated with hysterectomy. Complication Coagulopathy Fever Urinary tract injury Wound infection Oophorectomy Salpingectomy Pulmonary oedema Retro-peritoneal haematoma Paralytic ileus Return to operating theatre Cardiac arrest Maternal death
n
(%)
12 10 7 2 2 1 1 1 1 3 1 0
22.2 18.9 13.5 3.6 3.6 1.9 1.9 1.9 1.9 5.3 1.9
3
during the same hospitalisation (Whiteman et al. 2006); within 24–72 h (Francois et al. 2005; Sahin et al. 2014; Umezurike et al. 2008; Tadesse 2011); 30 days postpartum (Bodelon et al. 2009) and 6 weeks postpartum (Awan et al. 2011). Furthermore, some studies have excluded planned hysterectomies and those due to malignancies or structural abnormalities (Haynes 2004); hence it is difficult to compare the incidence reported in different studies. More than half of the patients in this study had evidence of morbidly adherent placenta, which would be considered the leading indication for hysterectomy and is similar to other studies (Machado 2011; Flood et al. 2009). Moreover, 50% had placenta praevia, provided we include praevia associated with accreta, and the frequency of these indications is in accordance with recent literature (Ehtisham 2011). There is a high rate of caesarean section as the mode of delivery in our PH cases, 83.1%. Furthermore, our data shows a previous caesarean section to be associated with abnormal placentation, the leading indication for PH. Considering the increasing rate of caesarean section in New South Wales, from 26.1% to 30.1% during a decade, there is concern about the increasing rate of PH and its complications. Clinicians can contribute to reducing this event by preventing unnecessary first caesarean sections. This may be facilitated by adhering to appropriate definitions for failed induction and arrest of labour progress, training and experience in operative vaginal delivery and including effects on future reproductive health when discussing the first caesarean delivery with patients (Spong et al. 2012). Whether encouraging vaginal birth after caesarean section (VBAC) can affect the rate of PH is unknown, as a case–control study, comparing PH following VBAC and elective caesarean section, would be required to show this effect. However, in theory, a greater number of previous caesarean sections can lead to higher incidence of abnormal placentation – the main cause of peripartum hysterectomy reported in many recent studies. Our study has limitations, the main one being that it was conducted in a single tertiary centre, which can potentially limit its generalisability. We also acknowledge the limitations due to the descriptive nature of the study. It is notable that extraction of data from operative reports and patient records may improve its accuracy as compared with population-based studies where data is obtained from a national electronic database. Also, the effectiveness of the B-Lynch suture, activate factor VII and uterine artery embolisation prior to hysterectomy was not evaluated. Our study highlights the significance of knowledge of placentation and adequate preoperative planning and consideration for elective hysterectomy in specific cases, i.e. placenta accreta. As this has shown to significantly reduce the amount of blood loss and blood products required which may impact patient morbidity. Preoperative planning would include having senior clinicians present, close liaison with blood bank and timing of operation to meet adequate resource availability. Furthermore, the higher rate of elective caesarean sections as compared with emergency caesarean sections is attributed to a high rate of planned caesarean hysterectomies in our institution. We would recommend a cohort study, considering a randomised control trail would not be feasible, to be performed to compare the conservative measures utilised in severe obstetric haemorrhages that led to hysterectomy with those which succeeded in controlling the haemorrhage without this surgical intervention. Also, a national birth register would be of paramount value to improve information on the characteristics of Australian women who undergo a peripartum hysterectomy. This would provide an estimate of the national incidence of peripartum hysterectomy, describe the trend and risk factors and guide clinicians
4
A. Shamsa et al.
in decision-making and appropriate counselling of patients at an increased risk. Declaration of interest: The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.
J Obstet Gynaecol Downloaded from informahealthcare.com by Nyu Medical Center on 05/02/15 For personal use only.
References Awan N, Bennett MJ, Walters WA. 2011. Emergency peripartum hysterectomy: a 10-year review at the Royal Hospital for Women, Sydney. Australian and New Zealand Journal of Obstetrics and Gynaecology 51:210–215. Bodelon C, Bernabe-Ortiz A, Schiff MA, Reed SD. 2009. Factors associated with peripartum hysterectomy. Obstetrics and Gynecology 114:115–123. de la Cruz CZ, Coulter ML, O’Rourke K, Amina Alio P, Daley EM, Mahan CS. 2013. Women’s experiences, emotional responses, and perceptions of care after emergency peripartum hysterectomy: a qualitative survey of women from 6 months to 3 years postpartum. Birth 40:256–263. Ehtisham S. 2011. Emergency peripartum hysterectomy. Pakistan Journal of Surgery 27:288–291. Flood KM, Said S, Geary M, Robson M, Fitzpatrick C, Malone FD. 2009. Changing trends in peripartum hysterectomy over the last 4 decades. American Journal of Obstetrics and Gynecology 200:632.e1–e6. Francois K, Ortiz J, Harris C, Foley MR, Elliott JP. 2005. Is peripartum hysterectomy more common in multiple gestations? Obstetrics and Gynecology 105:1369–1372. Haynes K, Stone C, King J. 2004. Major morbidities associated with childbirth in victoria: obstetric haemorrhage and associated hysterectomy. Melbourne: Public Health Group, Department of Human Services. Jones B, Zhang E, Alzouebi A, Robbins T, Paterson-Brown S, Prior T, Kumar S. 2013. Maternal and perinatal outcomes following peripartum hysterectomy from a single tertiary centre. Australian and New Zealand Journal of Obstetrics and Gynaecology 53:561–565. Knight M, Kurinczuk JJ, Spark P, Brocklehurst P; United Kingdom Obstetric Surveillance System Steering Committee. 2008. Cesarean delivery and peripartum hysterectomy. Obstetrics and Gynecology 111:97–105. Kwee A, Bots ML, Visser GH, Bruinse HW. 2006. Emergency peripartum hysterectomy: A prospective study in The Netherlands. European Journal of Obstetrics, Gynecology, and Reproductive Biology 124:187–192.
Li Z, Zeki R, Hilder L, Sullivan EA. 2013. Australia’s mothers and babies 2011. Perinatal statistics, Series No. 28, Cat. No. PER 59. Canberra: AIHW National Perinatal Epidemiology and Statistics Unit. Machado LS. 2011. Emergency peripartum hysterectomy: Incidence, indications, risk factors and outcome. North American Journal of Medical Sciences 3:358–361. Muench MV, Baschat AA, Oyelese Y, Kush ML, Mighty HE, Malinow AM. 2008. Gravid hysterectomy: a decade of experience at an academic referral center. Journal of Reproductive Medicine 53:271–278. Sahin S, Guzin K, Eroğlu M, Kayabasoglu F, Yaşartekin MS. 2014. Emergency peripartum hysterectomy: our 12-year experience. Archives of Gynecology and Obstetrics 289:953–958. Spong CY, Berghella V, Wenstrom KD, Mercer BM, Saade GR. 2012. Preventing the first cesarean delivery: summary of a joint Eunice Kennedy Shriver National Institute of Child Health and Human Development, Society for Maternal-Fetal Medicine, and American College of Obstetricians and Gynecologists Workshop. Obstetrics and Gynecology 120:1181–1193. Sullivan E, Hall B, King JF. 2008. Maternal deaths in Australia 2003–2005. Sydney: AIHW National Perinatal Statistics Unit, Series No. 3. Cat. No. PER 42. Tadesse W, Farah N, Hogan J, d’Arcy T, Kennelly M, Turner MJ. 2011. Peripartum hysterectomy in the first decade of the 21st century. Journal of Obstetrics and Gynaecology 31:320–321. Umezurike CC, Feyi-Waboso PA, Adisa CA. 2008. Peripartum hysterectomy in Aba southeastern Nigeria. Australian and New Zealand Journal of Obstetrics and Gynaecology 48:580–582. Whiteman MK, Kuklina E, Hillis SD, Jamieson DJ, Meikle SF, Posner SF, Marchbanks PA. 2006. Incidence and determinants of peripartum hysterectomy. Obstetrics and Gynecology 108:1486–1492. Wright JD, Bonanno C, Shah M, Gaddipati S, Devine P. 2010a. Peripartum hysterectomy. Obstetrics and Gynecology 116:429–434. Wright JD, Devine P, Shah M, Gaddipati S, Lewin SN, Simpson LL et al. 2010b. Morbidity and mortality of peripartum hysterectomy. Obstetrics and Gynecology 115:1187–1193. Wu S, Kocherginsky M, Hibbard JU. 2005. Abnormal placentation: twentyyear analysis. American Journal of Obstetrics and Gynecology 192: 1458–1461. Yoong W, Massiah N, Oluwu A. 2006. Obstetric hysterectomy: changing trends over 20 years in a multiethnic high risk population. Archives of Gynecology and Obstetrics 274:37–40. Zwart JJ, Dijk PD, van Roosmalen J. 2010. Peripartum hysterectomy and arterial embolization for major obstetric hemorrhage: a 2-year nationwide cohort study in the Netherlands. American Journal of Obstetrics and Gynecology 202:150.e1–e7.
