Hemorragia postparto asociado a césareas by medios comunes medios comunes

Best Practice & Research Clinical Obstetrics and Gynaecology xxx (2012) 1–17

Contents lists available at SciVerse ScienceDirect

Best Practice & Research Clinical Obstetrics and Gynaecology journal homepage: www.elsevier.com/locate/bpobgyn

Postpartum haemorhage associated with caesarean section and caesarean hysterectomy Sue Fawcus, MA (Oxon), MB BCh (London), FRCOG a, Jagidesa Moodley, FCOG, FRCOG, Professor Emeritus b, * a

Department of Obstetrics and Gynaecology, University of Cape Town, and Head, Obstetric Services, Mowbray Maternity Hospital, Cape Town Women’s Health and HIV Research Group, Nelson R Mandela School of Medicine, University of KwaZulu-Natal, Private bag 7, Congella 4013, South Africa b

Keywords: caesarean section postpartum haemorrhage clinical and surgical management

Excessive haemorrhage associated with caesarean section, commonly defined as blood loss in excess of 1000 ml, is frequently underestimated, but is documented as occurring in more than 5–10% of caesarean sections. Common causes are uterine atony, abnormal placentation, uterine trauma and sepsis. It is a major cause of maternal morbidity globally and of maternal mortality in low- and middleincome countries; however, many reports do not disaggregate it from postpartum haemorrhage in general. In this chapter, we outline preventive measures, including uterotonic agents, and provide treatment algorithms for managing excessive haemorrhage during and after caesarean section. Several management options, including uterotonic therapy, uterine compression sutures, balloon tamponade, blood-vessel ligation and uterine artery embolisation are described; each has a role for treating the different causes of caesarean section bleeding in different contexts. Caesarean hysterectomy is indicated when medical and conservative surgical measures are unsuccessful, and as first-line surgery for extensive uterine rupture and bleeding from morbidly adherent placentae. It has an incidence ranging from 1–4 per 1000 caesarean sections, significantly greater than that for vaginal delivery. Although it is a life-saving procedure, it is associated with significant morbidity, including massive blood transfusion and intensive care (10–48%), urological injury (8%) and the need for relook laparotomy (8–18%). ! 2012 Elsevier Ltd. All rights reserved.

* Corresponding author. Tel./Fax: þ27 031 2604241. E-mail address: jmog@ukzn.ac.za (J. Moodley). 1521-6934/$ – see front matter ! 2012 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.bpobgyn.2012.08.018

Please cite this article in press as: Fawcus S, Moodley J, Postpartum haemorhage associated with caesarean section and caesarean hysterectomy, Best Practice & Research Clinical Obstetrics and Gynaecology (2012), http://dx.doi.org/10.1016/j.bpobgyn.2012.08.018

S. Fawcus, J. Moodley / Best Practice & Research Clinical Obstetrics and Gynaecology xxx (2012) 1–17

Introduction Excessive haemorrhage during and after caesarean section is a serious adverse event, accounting for major morbidity and associated with maternal mortality in well-resourced and poorly resourced settings. Caesarean hysterectomy is a major surgical procedure most commonly carried out as an emergency life-saving procedure for intractable obstetric haemorrhage. Less commonly, it may be carried out as a semi-elective procedure when operable cervical carcinoma is diagnosed during pregnancy. Excessive haemorrhage associated with caesarean section Postpartum haemorrhage (PPH) at caesarean section is commonly defined as blood loss of 1000 ml or more.1,2 The Australian College of Obstetricians define PPH at caesarean section as blood loss greater than 750 ml, and other definitions use 500 ml irrespective of mode of delivery.3 Blood loss greater than 1500 ml is described as severe PPH, with blood loss greater than 2500 ml constituting massive blood loss.1 Some definitions of PPH at caesarean section include a change in haematocrit of more than 10%, the need for blood transfusion, further surgical procedures, or both. Definitions of PPH at caesarean section need to be standardised.4 Excessive blood loss can be intra-operative, but may also occur postoperatively through vaginal bleeding or concealed intra-peritoneal bleeding. The latter problem is frequently unrecognised by attending health workers. Accuracy of blood loss estimation Estimation of total blood loss during and after caesarean section is problematic owing to difficulties in accurate collection. Blood may be collected in suction bottles where it is mixed with liquor, and blood-loss estimation from soaked swabs and theatre linen is usually by visual inspection only. Blood loss in the few hours after caesarean section is estimated by inspecting vaginal pads and bed linen. Visual estimation of blood loss during and after caesarean section or vaginal delivery is notoriously inaccurate, tending to overestimate at lower blood loss and underestimate at higher blood loss. Fifty years ago, Pritchard et al.2 estimated average blood loss at caesarean section to be 930 ml using a technique involving chromium labelled red blood cells. Such techniques are too complex to use in clinical practice. Several studies prefer to use haematocrit changes, the need for blood transfusion, or both, as more reliable indicators of excess bleeding. Stafford et al.5 used a formula that calculates blood loss based on haematocrit changes before and after caesarean section, maternal weight and height, and found that visual estimation significantly underestimated blood loss compared with calculated blood loss, particularly for blood loss in excess of 1000 ml and 1500 ml. Incidence of excessive blood loss associated with caesarean section The incidence of excessive blood loss associated with caesarean section is context specific, depending on the obstetric case mix at the particular level of care studied, and varies from 3–10% of caesarean section deliveries. A study of 4836 caesarean sections by Magann et al.6 in Perth, Australia, where blood loss was measured from collection drapes and weighing soaked swabs, showed the postpartum haemorrhage (PPH) rate for caesarean section (blood loss > 1000 ml) to be 4.84% for elective caesarean section and 6.75% for emergency caesarean section (P ¼ 0.007).6 The PPH rate for blood loss greater than 1500 ml was 3.04% for emergency caesarean section. The same study is one of the few to evaluate risk factors for PPH at caesarean section. It found that previous retained placenta, blood disorders, antepartum transfusion, preterm birth and general anaesthesia were risk factors for PPH after emergency caesarean section. Leiomyomata, placenta praevia, antepartum bleeding, preterm birth and general anaesthesia were risk factors for PPH after elective caesarean section. Combs et al.7 indirectly estimated blood loss using haematocrit changes, the need for blood transfusion, or both, and gave a caesarean section PPH rate of 5.9%. The investigators showed that general Please cite this article in press as: Fawcus S, Moodley J, Postpartum haemorhage associated with caesarean section and caesarean hysterectomy, Best Practice & Research Clinical Obstetrics and Gynaecology (2012), http://dx.doi.org/10.1016/j.bpobgyn.2012.08.018

S. Fawcus, J. Moodley / Best Practice & Research Clinical Obstetrics and Gynaecology xxx (2012) 1–17

anaesthesia, pre-eclampsia, prolonged first and second stage of labour, chorioamnionitis, preeclampsia and Hispanic ethnicity, all to be risk factors for PPH at caesarean section. Naef et al.8 described a caesarean section PPH rate of 7.9%, based on estimation of blood loss greater than 1500 ml, haematocrit changes, the need for blood transfusion, or all three. This study showed that Native American ethnicity, prolonged active labour, and obesity were all risk factors. It would be useful to evaluate risk factors in other settings, such as rural districts in Africa, or other poorly resourced settings, where prolonged obstructed labour is more common and anaemia more prevalent. Lower-segment caesarean section is associated with less blood loss than classical or upper segment caesarean section.7 In clinical practice, knowledge of risk factors for caesarean section PPH enables adequate planning for level of care, availability of blood products, and availability of skilled surgical assistance. Causes of excessive blood loss associated with caesarean section Excessive bleeding at caesarean section may be caused by uterine atony, trauma, placental site bleeding, abruptio placenta, and adhesions. These are discussed below. Uterine atony Uterine atony is particularly common after prolonged labour and in the presence of chorioamnionitis, but also occurs with caesarean section for multiple pregnancy, and polyhydramnios. Trauma Trauma refers to uterine tears and lacerations, including lateral tears into the uterine vessels in the broad ligament and vertical tears down the lower segment. Trauma may be caused by a difficult delivery of the baby; an impacted fetal head after prolonged second-stage of labour, or obstructed labour being particularly high risk for excessive bleeding. In addition, caesarean section for transverse lie and large baby may be associated with uterine tears and surgical extension of the incision into the upper segment. Faulty surgical technique with the uterine incision being made too low on the lower segment or too lateral on one side of the uterus may cause traumatic bleeding, as can too rapid a delivery or excessive force used to deliver the fetal head. Placental site bleeding Placental side bleeding most commonly occurs with placenta praevia and morbidly adherent placenta, but may also occur with an excessively large placental site (i.e. for multiple pregnancies and diabetic mothers). With increasing rates of morbidly adherent placenta associated with higher caesarean section rates, this cause is becoming more frequent. Abruptio placentae Caesarean section is recommended for abruptio placentae when the fetus is alive. If complicated by a large retroplacental clot and couvelaire uterus, however, adequate uterine contraction does not occur, resulting in severe haemorrhage. Adhesions Caesarean section carried out in the presence of extensive adhesions usually from a previous caesarean section, requires sharp dissection and causes more intra-operative haemorrhage. The causes of haemorrhage at caesarean section include all those from the commonly used ‘four Ts’ classification (tissue, trauma, tone, thrombin) but, in addition, include aspects of surgical injury to the uterus or other soft tissues.9 Please cite this article in press as: Fawcus S, Moodley J, Postpartum haemorhage associated with caesarean section and caesarean hysterectomy, Best Practice & Research Clinical Obstetrics and Gynaecology (2012), http://dx.doi.org/10.1016/j.bpobgyn.2012.08.018

S. Fawcus, J. Moodley / Best Practice & Research Clinical Obstetrics and Gynaecology xxx (2012) 1–17

Mortality and morbidity caused by excessive haemorrhage associated with caesarean section Excessive haemorrhage is a major cause of severe acute maternal morbidity in well-resourced settings. For example, in Scotland, which has an ongoing audit of severe acute maternal morbidity, massive obstetric haemorrhage (>2500 ml) accounted for 50% of all severe morbidity between 2000 and 2002.10,11 The rate of massive obstetric haemorrhage was 3.7 per 1000 births, with 41% having had an emergency caesarean section and 17.2% an elective caesarean section. This compares with a rate of 15% emergency caesarean section for all Scottish births. Data from some well-resourced settings indicate that PPH-related morbidity is increasing, and this correlates with rising caesarean section rates.12 In poorly resourced countries, PPH is a major cause of maternal morbidity and is also the most common cause of maternal death.13 Not many country reports disaggregate the proportion of haemorrhage deaths specifically related to bleeding associated with caesarean section. It can be postulated, however, that, in poorly resourced settings where cephalopelvic disproportion is prevalent, prolonged labour and sepsis would contribute to bleeding associated with caesarean section. Shortage of surgical skills may mean that less experienced personnel carry out caesarean sections, with minimal support. In South Africa, which has an established national confidential enquiry into maternal deaths, bleeding associated with caesarean section was the most common cause of maternal deaths from obstetric haemorrhage between 2008 and 2010, accounting for 180 (26.2%) of the 688 maternal deaths caused by obstetric haemorrhage.14 Related morbidities included obstructed labour, previous caesarean section, abruptio placentae and placenta praevia. Most of these deaths were clearly avoidable. Lack of surgical skill to achieve haemostasis at the initial caesarean section, lack of skill to carry out the additional surgical measures required to arrest haemorrhage, and poor post-caesarean section monitoring were all avoidable factors.14,15 Such problems may occur in other poorly resourced countries where a serious lack of skills is evident, especially in more remote rural hospitals. The most recent confidential enquiry into maternal deaths in the UK shows a sustained and marked reduction of maternal mortality caused by obstetric haemorrhage.16 Prevention and early detection of excessive haemorrhage associated with caesarean section The following steps should be taken to minimise haemorrhage associated with caesarean section and to ensure early detection when it does occur.

Correct management of labour Correct management of labour using the partogram is necessary so that prolonged labour is diagnosed timeously, the appropriate interventions made, and emergency caesarean section for this indication is carried out before labour becomes obstructed. Unnecessary surgery Unnecessary caesarean section should be avoided. Experience and training A caesarean section, which is high risk for haemorrhage (i.e. major placenta praevia or suspected placenta accreta) should be carried out at the appropriate level of care by the most experienced surgeon. In many low-income countries, this may be difﬁcult because of problems in making a diagnosis in advance, and junior doctors could be faced with this type of scenario on their own in a remote rural hospital setting. Prior surgical training of this cadre of doctors, and the availability of a functional telephonic link to a specialist in a referral hospital, would be helpful in this situation. Please cite this article in press as: Fawcus S, Moodley J, Postpartum haemorhage associated with caesarean section and caesarean hysterectomy, Best Practice & Research Clinical Obstetrics and Gynaecology (2012), http://dx.doi.org/10.1016/j.bpobgyn.2012.08.018

S. Fawcus, J. Moodley / Best Practice & Research Clinical Obstetrics and Gynaecology xxx (2012) 1–17

Optimising haemoglobin Haemoglobin optimisation before delivery by prevention, early detection and treatment of anaemia can minimise haemorrhage associated with caesarean section. Surgical techniques Surgical techniques that have been shown, by available evidence, to reduce blood loss should be used. The use of blunt dissection rather than sharp dissection for abdominal entry and controlled cord traction rather than manual removal to deliver the placenta have both been shown to reduce blood loss at caesarean section.17 As caesarean section is one of the most frequently performed surgical procedures worldwide, there is a tendency among surgeons to ‘break the 10 minute’ barrier and compete to do the shortest duration caesarean section. In such a situation, due care to ensure correct uterine closure, particularly of the angles of the surgical incision, and careful checking for haemostasis, may be neglected. Prophylactic uterotonics and medication at caesarean section A uterotonic agent needs to be given after delivery of the baby to cause placental separation and expulsion, and to ensure uterine contraction, as with active management of third stage of labour. This has been shown to reduce the risk of PPH after vaginal delivery by 60%. The uterotonic regimen for which there is most evidence of efficacy, with minimal side-effect profile, is 10 iu oxytocin intramuscularly.18–20 Debate has taken place between anaesthetists and obstetricians on the appropriate choice of uterotonic agent at caesarean section, its dosage, and route of administration.21 The aim, however, is to ensure optimal uterine contraction, and the anaesthetist is also concerned about possible intraoperative side-effects. Hypotension has been shown to follow rapid intravenous boluses of oxytocin in doses greater than 5 iu, and hypertension, vomiting, or both, may be a side-effect of ergometrine; both of these agents are effective in preventing PPH. Shivering and pyrexia may follow misoprostol administration, but this agent has been shown to be inferior to other agents in the prevention of PPH. Current evidence recommends that the lowest intravenous bolus of oxytocin to achieve adequate uterine contraction at caesarean section is 5 iu given over a few minutes; and it has the least side-effect profile.22,23 An alternative suggested by some anaesthetic guidelines is to administer a 2.5 iu intravenous bolus of oxytocin slowly, and concurrently administer a prophylactic oxytocin infusion of 10 iu in 1 L fluid.21 One randomised-controlled trial has been conducted on the use of carbetocin, a longer-acting oxytocin, which has similar efficacy and reduces the need for the infusion. It is expensive, however, and not readily available in most settings.24 In one prospective audit of introducing prophylactic intramuscular ergometrine at caesarean section in addition to intravenous oxytocin 5 iu, a reduction in delayed PPH was reported, but an increase in side-effects of nausea and vomiting was found.25 There seems to be no role for prophylactic use of prostaglandin preparations at caesarean section.26 Recently, attention has focused on the use of tranexamic acid (TXA) to reduce blood loss if given prophylactically at caesarean section. This is not a uterotonic agent; TXA is an anti-fibrinolytic agent better known to gynaecologists for oral use as treatment of menorrhagia, and to trauma surgeons where it has been shown to reduce blood loss.27 Preliminary trials show that, given prophylactically at caesarean section, TXA does reduce blood loss and the need for additional uterotonics.28–30 Concern remains about the potential of TXA to promote venous thrombosis (VTE) in the context of a physiological state when VTE risk is already increased. This has not been confirmed in any of the studies, although most have not been powered for this outcome; Vigilant post-caesarean section monitoring Early detection of ongoing vaginal bleeding can be detected by deteriorating vital signs measured in the recovery area. It is important that both anaesthetist and recovery nurse ensure that ongoing bleeding is not occurring before transfer to a postnatal ward. Women at high risk should be monitored in a high-care area. Frequent observations need to be continued in the postnatal ward, and action taken on abnormal findings. This can be a problem in low-resource settings where staffing may be Please cite this article in press as: Fawcus S, Moodley J, Postpartum haemorhage associated with caesarean section and caesarean hysterectomy, Best Practice & Research Clinical Obstetrics and Gynaecology (2012), http://dx.doi.org/10.1016/j.bpobgyn.2012.08.018

S. Fawcus, J. Moodley / Best Practice & Research Clinical Obstetrics and Gynaecology xxx (2012) 1–17

inadequate to monitor at the correct frequency, meaning that post-caesarean section bleeding is detected when it as at an advanced stage and the woman already in hypovolaemic shock.14,15 In addition, the attendant health workers need to have a high index of suspicion for ongoing intraperitoneal bleeding when vital signs are deteriorating in the absence of vaginal bleeding. The colour coded Maternity Early Warning monitoring charts pioneered in the UK may be a useful addition to monitoring by assisting the interpretation of abnormal measurements.31 Resuscitation for excessive haemorrhage at caesarean section Resuscitation for excessive haemorrhage at caesarean section will be carried out by the anaesthetist; good communication between anaesthetist and surgeon is essential to ensure that the former is aware of ongoing excessive bleeding and the latter is informed about the haemodynamic status of the patient.32 Most caesarean sections are carried out under spinal anaesthesia, but the anaesthetist will need to convert to a general anaesthetic in the event of severe ongoing haemorrhage. Blood loss of 1000 ml requires blood samples to be taken for haemoglobin, coagulations studies and emergency cross-matching. Fluid resuscitation is aimed at restoring intravascular volume and maintaining cardiac output. It requires two large bore intravenous cannulae and initially involves the rapid infusion of two litres of crystalloid solution. If bleeding is not controlled and the systolic blood pressure remains less than 100 mm Hg and the pulse greater than 110 beats per minute, 500 ml of a colloid should be given and repeated if necessary. A further litre of crystalloid can be given if no colloid is available. All fluids should be warmed. Further resuscitation for ongoing hypovolaemia requires blood transfusion. This is preferably crossmatched blood, but if no on-site blood bank is available, the ‘emergency blood’ stored in the maternity area fridge can be used. Facilities carrying out caesarean section should have some form of blood products available on site. In countries with limited resources, these are usually refrigerated O Negative and O Positive units at district hospitals, with ‘blood banks’ at regional and tertiary hospitals. Fresh frozen plasma is required after two units of red cell concentrate have been given, with a formula of one fresh frozen plasma for one unit red cell concentrate. Platelets will be needed if the haematology results show the count to be less than 50,000. Failure to achieve a response to resuscitation implies continued bleeding, which must be identified and treated by the surgeon while resuscitation continues. Cell-salvage techniques are of potential value in situations in which urgent availability of blood is limited, and in some women who are of the Jehovah’s Witness faith. This has been proven to be safe and effective for ectopic pregnancy. At caesarean section, however, the blood suctioned from the amniotic cavity is often mixed with amniotic fluid, and therefore could cause severe adverse reactions if transfused back into the patient. Nevertheless, in some case studies, this has been carried out with the use of special techniques to ‘clean’ the blood with success and minimal side-effects. Randomisedcontrolled trials are indicated before this technique is accepted as routine practice.33 Management of excessive bleeding at caesarean section It is useful to display in theatre algorithms for the management of PPH at caesarean section, as well as to have diagrams of useful techniques that may be used, such as uterine compression sutures. The algorithms presented in this chapter are currently being used in South Africa and have been published in The monograph of the management of postpartum haemorrhage,34 a pocket manual on PPH for healthcare workers based on current evidence-based guidelines and expert opinions. The surgeon must diagnose the following: cause of the bleeding; uterine atony, tears and lacerations; placental site bleeding, abruptio placentae or bleeding from adhesions; and whether there is concurrent coagulopathy. The surgeon must also monitor the response to the various measures used to arrest haemorrhage. This may be by inspecting the bleeding areas visually or inspecting vaginal blood loss. The latter is particularly important for monitoring ongoing bleeding from uterine atony or the placental bed after uterine closure at caesarean section, and can be facilitated by having the patient in the Lloyd–Davies position. Please cite this article in press as: Fawcus S, Moodley J, Postpartum haemorhage associated with caesarean section and caesarean hysterectomy, Best Practice & Research Clinical Obstetrics and Gynaecology (2012), http://dx.doi.org/10.1016/j.bpobgyn.2012.08.018

S. Fawcus, J. Moodley / Best Practice & Research Clinical Obstetrics and Gynaecology xxx (2012) 1–17

The following treatment modalities can be used to arrest bleeding at caesarean section from the different causes. Atonic uterus Medical treatment with stepwise use of uterotonic agents should be the first treatment option. Oxytocin infusion 20 iu/L at 120–240 ml/h should be commenced. If the woman responds poorly and is not cardiac or hypertensive, ergometrine 0.2 mgm intravenously can be given and repeated once. Current research is evaluating the safety and efficacy of lower doses of ergometrine in women who are hypertensive. The third-line uterotonic medication is a prostaglandin. This can be in the form of sublingual misoprostol with doses of 400–600 mcg. The latter dose should not be exceeded. If the woman is shocked or ventilated, the rectal route is recommended. If prostaglandin preparations, are available, carboprost can be given 0.25 mgm intramuscularly or prostaglandin F2 alpha can be given by intramyometrial injection. Care must be taken not to exceed the optimal dose, as bronchoconstriction and severe hypertension are untoward side-effects. The 5 mgm vial should be diluted with sterile water in a 20 ml syringe, and given as an intramyometrial injection of 2 ml, equivalent to 0.5 mg active ingredient. This can be repeated until a maximum of 2 mgm (8 ml of solution) has been given.19,20 If medical treatment does not correct the uterine atony, surgical measures must be instituted and, if a junior doctor is operating, senior assistance should be sought. The B lynch uterine compression suture is the first surgical measure that should be carried out, and can be easily learned by all grades of doctor or clinical officer who carry out caesarean sections. An absorbable suture should be used.35 The B Lynch is carried out when the uterine cavity is open, but there are alternatives, such as the Hayman suture, when the uterine incision has been closed.36 Successful outcome has been shown in case studies, defined as arresting the bleeding without the need to proceed to hysterectomy, and minimal immediate postoperative morbidity have been identified.37–39 Long-term morbidity has been less well assessed; cases of successful pregnancies after insertion have been reported.40 A few cases of uterine necrosis after insertion have also been reported.41–43 No major long-term side-effects have been observed when compression sutures have been combined with balloon tamponade; however, an increased risk of uterine ischaemia and necrosis has been reported when compression sutures are combined with vessel ligation. It is recommended that registers are kept of women in whom uterine compression sutures are placed to enable long-term events to be ascertained.44 Trauma: lateral tears into broad ligament or tears into lower or upper segment of uterus If the caesarean section incision has been inadvertently extended too far laterally or has torn laterally into the broad ligament during delivery of the baby; arterial bleeding from vessels of the uterine artery will occur. Haemostatic sutures placed in the broad ligament may control the bleeding but, if poorly placed, can cause a broad ligament haematoma. It is preferable to do a uterine artery ligation procedure such as the mass uterine artery ligation technique first described by O0 Leary45 and developed into the stepwise uterine artery devascularisation approach suggested by Abdrabbo.46 The suture should be placed at the level of the uterine incision; the medial placement will be in the lateral border of the uterus, and the lateral placement will be in the avascular window of broad ligament peritoneum between the uterine arteries medially and the infundibulo-pelvic ligament laterally. Some practitioners, especially gynaecologists, prefer to open the broad ligament and identify the uterine artery before ligating it. Tears down the lower segment of uterus must be carefully identified and the apex of the tear secured. It must be sutured with a continuous absorbable suture from the apex up to the uterine incision. Then, the uterine incision must be closed. Lower-segment tears that are long and are directed laterally will lie close to the ureteric path, and some case studies have shown the ureter having been occluded by sutures placed for haemostasis. The path of the ureter at this level needs to be identified. Tears or extension of the uterine incision into the upper uterine segment require closure as for a classical caesarean section, and must be documented in the notes. The woman will need to be managed as though she had a classical caesarean section in future pregnancies. Please cite this article in press as: Fawcus S, Moodley J, Postpartum haemorhage associated with caesarean section and caesarean hysterectomy, Best Practice & Research Clinical Obstetrics and Gynaecology (2012), http://dx.doi.org/10.1016/j.bpobgyn.2012.08.018

S. Fawcus, J. Moodley / Best Practice & Research Clinical Obstetrics and Gynaecology xxx (2012) 1–17

The surgeon should always explore the posterior surface of the uterus before closing the abdomen because posterior uterine rupture has been missed, especially after obstructed labour. Bleeding from the placental bed Bleeding from the placental bed is mostly associated with placenta praevia and a morbidly adherent placenta in any part of the uterus. Uterotonic medical treatment should be commenced, but is frequently insufficient to reduce this type of bleeding. It is best managed surgically by additional haemostatic ‘figure of eight’ sutures. Excessive bleeding of the lower segment can be successfully managed in many cases by balloon tamponade.47,48 A purpose-designed tamponade system, such as the Cook-Bakri, or Rusch catheter, can be used or a ‘home made’ device using a condom or sterile glove as the balloon attached to a Foley’s catheter34,49,50 This would be inserted with the balloon compressing the lower segment and the catheter passing out through the cervix into the vagina. Antibiotic cover is required. The balloon can be deflated and removed after 8–12 h. The ‘uterine sandwich’ technique describes a combination of surgical modalities, where a uterine compression suture is first placed followed by insertion of a balloon tamponade sytem.51 This is particularly effective for placental site bleeding but could also be used for intractable cases of uterine atony. The use of medical treatment other than uterotonics requires more research. Infiltration of the lower segment bleeding area with vasopressin has been described in individual cases, and the use of tranexamic acid, which can be given as 1 gm intravenously for this problem, is being investigated. If placenta accrete is present, uterine curettage with the largest uterine curette can be a useful way of scraping off the adherent placenta. Placenta increta or percreta usually require direct recourse to hysterectomy. With a placenta increta that fails to separate from the uterine wall after administration of oxytocin and in the absence of any bleeding, it is best not to attempt placental removal at all, but rather to leave it in situ and await spontaneous expulsion. Technical details of procedures such as B Lynch uterine compression suture, balloon tamponade and uterine artery ligation can be found in various texts; they should be displayed as posters in theatre and taught to all practitioners who carry out caesarean sections. Management of ongoing bleeding from all causes after above medical and surgical conservative measures have been shown to be unsuccessful Aortic compression Aortic compression can be applied by an assistant as a temporising measure while help is called. Recombinant factor V11 Recombinant factor V11 has been found to be effective in individual case reports, but is expensive and has a strong association with thrombo-embolic complications; therefore, current evidence does not strongly support its use.52–54 Uterine tourniquet Some practitioners have described the use of a Foley’s catheter or feeding tube tied as a tourniquet around the lower part of the uterus in a similar way to its use at myomectomy. This compresses the uterine vessels and reduces blood loss while awaiting help or during transfer of a patient to a level of care with more expertise. This technique has not been properly evaluated, but individual case reports from South Africa have shown some women transferred from a district hospital after bleeding at caesarean section with a uterine tourniquet in situ, who have had successful outcome. It is not known as for how long such a tourniquet can be placed in situ without causing irreversible ischaemic changes to the uterus. Hysterectomy In cases of irreparable uterine rupture or placenta increta or percreta, it is usually necessary to proceed immediately to hysterectomy without attempting conservative measures. Hysterectomy will also be necessary if all other modalities of treatment have been unsuccessful. Please cite this article in press as: Fawcus S, Moodley J, Postpartum haemorhage associated with caesarean section and caesarean hysterectomy, Best Practice & Research Clinical Obstetrics and Gynaecology (2012), http://dx.doi.org/10.1016/j.bpobgyn.2012.08.018

S. Fawcus, J. Moodley / Best Practice & Research Clinical Obstetrics and Gynaecology xxx (2012) 1–17

Several research studies have shown less blood loss and better outcomes if there is a shorter delay in carrying out the hysterectomy. This still remains a difficult area of clinical judgement, however, especially in the case of a young primipara. In many institutions, it is thought that two experienced obstetricians or doctors must make the decision, and the Head of the hospital be informed if prior consent has not been obtained. The second section of this chapter provides more detail on caesarean hysterectomy. Internal iliac artery ligation Internal iliac artery ligation could be considered if fertility needs to be preserved and a hysterectomy is indicated.55,56 The success rate, however, is only 50% and requires significant surgical expertise; in well-resourced countries, it is mainly carried out by gynaecological oncologists. One study57 showed higher success rates when this technique was used for controlling bleeding at caesarean section, rather than for PPH after vaginal delivery. In addition to being carried out as a uterine preservation procedure, it can also be carried out after hysterectomy if bleeding continues from pelvic vessels. It is probably beyond the competence of generalist doctors carrying out caesarean sections in rural hospitals in lessresourced settings where other procedures described above, intra-abdominal packing after hysterectomy, or both, may be more feasible. Uterine artery embolisation Some institutions may have the radiological equipment and skills to carry out uterine artery embolisation.58,59 This interventional radiological technique will be the procedure of choice if available, but requires a well-resuscitated patient. These facilities tend to be only available in tertiary settings and seem to be used more frequently in some countries (e.g. the Netherlands) compared with others. This procedure has been used with some success for bleeding after caesarean section, but perhaps has a greater role when severe haemorrhage is predicted pre-operatively, such as for placenta praevia increta or percreta where the facilities can be prepared in advance (Figs. 1 and 2). Bleeding after caesarean section Bleeding after caesarean section is a problem that should be minimised by the preventive measures described previously in this chapter. Excessive bleeding per vaginam is more easily detected and most commonly is caused by ongoing uterine atony, but also can be caused by placental site bleeding, especially in placenta praevia and accreta. Occasionally, bleeding vessels in the angle of the uterine incision, which were nor adequately ligated, point inwards to the uterine cavity and cause postoperative vaginal bleeding. Intra-abdominal bleeding is more difficult to detect because the bleeding is concealed. Deteriorating vital signs (e.g. tachycardia above 110 beats/min and blood pressure systolic less than 100 mms Hg with pallor) are clues. The health worker should not diagnose hypotension as being a side-effect of a spinal anaesthetic unless haemorrhage has been excluded. Intrabdominal bleeding is usually caused by bleeding from the angle of the uterine incision, which may cause free blood in the peritoneal cavity or a broad ligament haematoma. Management is shown in the algorithm that follows. It involves volume resuscitation and blood transfusion as appropriate. It is important to diagnose the cause of the bleeding. Uterine atony can initially be treated medically. Placental site bleeding, if anticipated because it follows a placenta praevia or placenta accreta, could be treated with balloon tamponade. If a woman responds poorly to these measures, bleeding is massive, or both, a relook laparotomy must be carried out without delay. Occasionally, bleeding may be unrelated to the uterus and arise from perforating vessels under the rectus muscle, which were injured during abdominal entry; these should be ligated. A Hayman uterine compression suture is indicated for uterine atony, which is not responsive to medical therapy. The ‘uterine sandwich’ technique can be used in an attempt to conserve the uterus. Haemostatic sutures can be placed for bleeding at the uterine angle together with a unilateral uterine artery ligation, especially if the bleeding is thought to be from vessels in the broad ligament. If there is vaginal bleeding in association with a well-contracted uterus, and no risk for placental site bleeding, it is advisable to reopen the uterine incision and look for arterial bleeders at the angles, which can then be ligated. If all these measure fail, it is necessary to proceed to hysterectomy. Please cite this article in press as: Fawcus S, Moodley J, Postpartum haemorhage associated with caesarean section and caesarean hysterectomy, Best Practice & Research Clinical Obstetrics and Gynaecology (2012), http://dx.doi.org/10.1016/j.bpobgyn.2012.08.018

S. Fawcus, J. Moodley / Best Practice & Research Clinical Obstetrics and Gynaecology xxx (2012) 1–17

Management

Prevention 2.5 iu oxytocin iv over 30 s after delivery of baby, followed by oxytocin infusion. Delivery placenta by cord traction. Good surgical technique.

Call for more senior help (if available or telephonic advice).

Diagnosis Visual estimation Blood loss in suction bottles >500ml ↓BP & ↑HR as detected by anaesthetist

Resuscitation (anaesthetist) Second intravenous infusion line. 20iu oxytocin in 1 litre as infusion. Maintain blood pressure with fluids and blood. Convert to general anaesthesia. Central line.

Atonic uterus Oxytocin infusion. Ergometrine 0.2 mg iv (not if hypertension or cardiac) ; repeat x1 Misoprostol 400 – 600 µgm per rectum Prostaglandin F2 alpha 1 mg intramyometrial (repeat x 1). B-Lynch compression suture. Subtotal abdominal hysterectomy.

Arrest Haemorrhage* (surgeon)

Uterine tears Lateral tears Uterine artery ligation Inferior tears Secure apex & suture (check ureters are lateral to tear) Rupture* Repair or Subtotal abdominal hysterectomy

Placental site bleeding Mattress suture. Compression sutures. Stepwise uterine devascularisation. Balloon tamponade. Subtotal abdominal hysterectomy.

Fig. 1. Bleeding at caesarean section. * Proceed immediately to subtotal abdominal hysterectomy if uterine rupture is irreparable or placenta increta or perceta occur. STAH, subtotal abdominal hysterectomy.

Current evidence and research gaps Limited, good-quality evidence is available to inform on the best practices for managing bleeding at caesarean section. Much of what is presented here stems from ‘expert opinion’ and descriptive case series. More robust research evidence involving randomised-controlled trials for the prevention and medical prophylaxis of PPH at caesarean section is available. The problem of research on surgical treatment of PPH at caesarean section is the ethical issues of informed consent for randomised-controlled trials in shocked patients; however, this could be addressed with methodologies as outlined in the CRASH trial.27 Randomised-controlled trials are currently in process on the efﬁcacy and side-effect proﬁle of tranexamic acid used for the prevention, treatment, or both, of PPH at caesarean section.60 Please cite this article in press as: Fawcus S, Moodley J, Postpartum haemorhage associated with caesarean section and caesarean hysterectomy, Best Practice & Research Clinical Obstetrics and Gynaecology (2012), http://dx.doi.org/10.1016/j.bpobgyn.2012.08.018

S. Fawcus, J. Moodley / Best Practice & Research Clinical Obstetrics and Gynaecology xxx (2012) 1–17

Management

Prevention and early detection Haemostasis at initial caesarean section. Regular post-operative monitoring. Monitoring of at-risk women who bled intra-operatively in high-care area (if available).

Resuscitate Second intraveneous infusion line Oxytocin 20 iu in 1 litre infusion. Maintain Blood pressure with fluids and blood

Uterus atonic Massage to remove clots. 20 iu oxytocin in 1 L as infusion. Ergometrine 0.2 mg iv (not if hypertension or cardiac) ; repeat x1 Misoprostol 400 – 600 µgm per rectum.

If ongoing bleeding

Bleeding from uterine incision

↓

Compression sutures

Single bleeding vessel

↓

STAH

Haemostatic sutures

↓

Stepwise uterine artery devascularisation

↓

STAH

Diagnosis Excessive vaginal bleeding. (revealed post partum haemorrhage) A decrease in blood pressure, an increase in heart rate, abdominal distension, and pallor (concealed bleeding).

Uterus wall contracted

↓

Atonic uterus

Bleeding along whole incision

↓

Open uterine incision; explore for bleeders and resuture

↓

Laparotomy

(Lloyd–Davies position)

Suspected placental site bleeding* Balloon tamponade

↓

Stepwise uterine artery devascularisation

↓

STAH

Stepwise uterine artery devascularisation

↓

STAH Fig. 2. Bleeding after caesarean section. * Proceed immediately to subtotal abdominal hysterectomy if the woman is very unstable. STAH, subtotal abdominal hysterectomy.

The so called ‘conservative’ surgical techniques, named as such because they obviate the need for hysterectomy, have been introduced in the past 20 years. They include balloon tamponade, uterine compression sutures, pelvic vessel ligation including uterine artery ligation and internal iliac artery ligation, and interventional radiological embolisation. Success rates of 78–84% have been described for balloon tamponade, 81–91.2% for uterine compression sutures, 42–84.6% for vessel ligations and 50–90% for radiological embolisation, in published case studies.11,44,61 Most of these studies do not distinguish all PPH from those only associated with caesarean section. In clinical practice, balloon tamponade is more frequently used for PPH after vaginal delivery and compression sutures for Please cite this article in press as: Fawcus S, Moodley J, Postpartum haemorhage associated with caesarean section and caesarean hysterectomy, Best Practice & Research Clinical Obstetrics and Gynaecology (2012), http://dx.doi.org/10.1016/j.bpobgyn.2012.08.018

S. Fawcus, J. Moodley / Best Practice & Research Clinical Obstetrics and Gynaecology xxx (2012) 1–17

bleeding associated with caesarean section. Research studies comparing the different treatment modalities, however, are difficult to design, implement and interpret. Also, different modalities of management may be of more value for the different causes of bleeding at caesarean section, so are not strictly comparable. Various case studies have shown that these treatment modalities reduce the need for hysterectomy. The Scottish confidential audit of maternal morbidity for 2008 showed a significant decrease in peripartum hysterectomy rate from 2003 to 2008, whereas, during the same time period, increased use of conservative measures were reported, especially balloon tamponade and compression sutures.62 In some of the case studies showing the ‘success’ of conservative surgical measures, it is possible that there is lower threshold for doing them than there would be for hysterectomy. An audit by the UK Obstetric Surveillance System (UKOSS) of contemporary practice for managing severe PPH, showed that balloon tamponade was a primary procedure in most severe PPH cases followed by compression sutures in 75%, pelvic vessel ligation in 36% and embolisation in 86% of cases, with 26% requiring hysterectomy.63 Another study conducted identified five factors retrospectively in women with severe ongoing haemorrhage who required advanced interventional procedures (e.g uterine artery embolisation, intra-abdominal packing, arterial ligation or hysterectomy). These were abnormal placentation, heart rate greater than 115 beats/min, prothrombin time less than 50%, fibrinogen less than 2 g/L, and detectable troponin.64 A score was devised from these factors that correlated with the need for advanced interventional procedures in a prospective cohort, thus validating the scoring system. The investigators suggest such a score would prompt health workers to be more aggressive earlier in the sequence of treatment, as many audits show that delays in instituting aggressive measures after failed medical therapy are the most common modifiable factor. Some of these measures, however, require accessible onsite laboratory facilities and these are beyond the scope of many hospitals in poorly resourced settings where heart rate and abnormal placentation would be the only ones of these five parameters that could be assessed. More research in the form of prospective registers for long-term follow up of women with uterine compression sutures and vessel ligations would be of value to review future fertility as well as side-effects. More research is also needed into the efficacy of these interventions and hysterectomy rates in developing countries, as most of the trials and case studies have been conducted in well-resourced countries. Level of care, health system and health worker training issues Uterine compression sutures, balloon tamponade, and uterine artery ligations are all procedures that need to be learned by any doctor trained to carry out caesarean section. These procedures can also be carried out in a district hospital with emergency blood available. The skill to carry out subtotal hysterectomy may not be available at this level, but should be available at all regional or tertiary hospitals which, in the ideal situation, are the more appropriate level of care to manage women with massive haemorrhage. Algorithms, posters and booklets are useful, and require distribution to relevant health workers; however, on their own, they are insufficient. It is important to emphasise the need for ‘hands on’ surgical training for all doctors carrying out caesarean section, as well as an approach to, and demonstration of, additional surgical skills to arrest excessive bleeding. The availability of experienced specialist assistance for difficult cases preferably on site, or easily available telephonic advice in the case of a remote rural hospital, is also important. Clinical outreach can help maintain skills and include surgical training. Clinical managers of maternity services should monitor bleeding associated with caesarean section as an indicator, ensure that guidelines are in place, that surgical training occurs, and that senior assistance is available. In addition, constant availability of emergency blood must be ensured, and adequate staffing and functional theatres are essential. Caesarean hysterectomy The first recorded caesarean hysterectomy in which both mother and infant survived was carried out in 1876 by Porro in Italy, as a life-saving procedure for a woman with uncontrollable bleeding. As Please cite this article in press as: Fawcus S, Moodley J, Postpartum haemorhage associated with caesarean section and caesarean hysterectomy, Best Practice & Research Clinical Obstetrics and Gynaecology (2012), http://dx.doi.org/10.1016/j.bpobgyn.2012.08.018

S. Fawcus, J. Moodley / Best Practice & Research Clinical Obstetrics and Gynaecology xxx (2012) 1–17

the procedure became more commonly used for this type of scenario, a shift took place in the mid 20th century for it to be carried out electively as a form of sterilisation in selected women requiring caesarean section. The observed associated morbidity, however, halted this approach. Caesarean hysterectomy is currently viewed as a life-saving measure to be used when no other procedure can arrest the haemorrhage (e.g. extensive uterine rupture and bleeding placenta percreta), or after conservative medical or surgical measures have failed to control bleeding from other causes. A systematic review of obstetric hysterectomy shows that the incidence varies according to the obstetric context from 0.2 per 1000 deliveries in Norway to 5 per 1000 in Eastern Turkey; and risk factors include past or current caesarean section delivery, multiparity and abnormal placentation.65 A study in Ireland showed the hysterectomy rate for vaginal delivery was 0.08 per 1000 vaginal deliveries compared with to 1.6 per 1000 caesarean-section deliveries.66 This markedly increased risk of hysterectomy after caesarean section compared with vaginal delivery is a consistent feature in other studies.67–70 This difference, however, is more often explained by the condition that required the caesarean section to be carried out rather than the operative procedure itself. Kacmar et al.71 partially controlled for this by only including women for whom vaginal delivery was planned. The investigators still found caesarean section was associated with an increased risk for caesarean hysterectomy. The Obstetric Surveillance System in the UK conducted between 2005 and 2006 gives hysterectomy rates of one in 30,000 vaginal deliveries, one in 1700 women with current first caesarean section, one in 1300 for women with one previous caesarean section, and one in 220 for women with two or more previous caesarean section.72 Therefore, with rising rates of caesarean section worldwide, it may be anticipated that caesarean hysterectomy rates will increase, and this could possibly offset the contribution of conservative surgical measures for PPH in reducing the need for caesarean hysterectomy. Caesarean hysterectomy is more frequently carried out in women over 35 years and of higher parity, but all case studies do show a small percentage carried out in young primipara.72 Indications in well-resourced countries tend to be mainly for abnomal placentation and intractable uterine atony; however, in low- or middle-income countries, sequelae of obstructed labour such as uterine rupture, atony and sepsis, may be major underlying factors for the ongoing haemorrhage necessitating the hysterectomy.73 Less common causes are secondary PPH, fibroids and infection. Elective procedures are uncommonly carried out, but are indicated when caesarean section is carried out for carcinoma of the cervix. Hysterectomy is seen as a last resort and, although some practitioners view it as a failure of management, it can undoubtedly be a life-saving procedure. It is a major procedure and is commonly associated with massive blood transfusion and intensive care, and also can be complicated by coagulopathy and urological injuries. The challenges of caesarean section versus hysterectomy relate to the expertise required but also to the decision making to do it. The possibility of hysterectomy should form part of preoperative counselling of all women undergoing caesarean section, and, particularly, for conditions where the risk of hysterectomy is higher. Preferably more than one obstetrician or senior doctor should make the decision to do the hysterectomy. Hysterectomy is often carried out too late after all conservative measures have been attempted but failed, by which time the woman may be coagulopathic, thus adding to the morbidity of the procedure. This is where clinical judgement is important. It is recommended to proceed straight to hysterectomy in the presence of placenta percreta, or ruptured uterus which is irreparable, or when above conservative measures are unsuccessful. Bleeding to caesarean section hysterectomy time interval is longer in young primipara, presumably because of the surgeon persisting longer to conserve the uterus.11 A subtotal abdominal hysterectomy (STAH) is usually sufficient to control the bleeding. If tears go down into the cervix, a uterine rupture extends into the cervix or lower segment bleeding occurs after major placenta praevia, the cervix will also need to be removed by total abdominal hysterectomy. A STAH is a shorter procedure and tends to be associated with less blood transfusion and urological injuries; however, these differences have not been found to be statistically significant.65,67,74 The technique of STAH should be taught to all practitioners carrying out caesarean sections; in Tanzania and Malawi, clinical officers have learned these procedures with good outcomes.75,76 Please cite this article in press as: Fawcus S, Moodley J, Postpartum haemorhage associated with caesarean section and caesarean hysterectomy, Best Practice & Research Clinical Obstetrics and Gynaecology (2012), http://dx.doi.org/10.1016/j.bpobgyn.2012.08.018

S. Fawcus, J. Moodley / Best Practice & Research Clinical Obstetrics and Gynaecology xxx (2012) 1–17

Problems at surgery relate to ongoing bleeding, coagulopathy, the possibility of bladder and ureteric injury, complications of massive blood transfusion, such as transfusion related acute lung injury and postoperative bleeding. Different case studies give morbidity rates ranging from 10–48%.65–69,74 Average blood transfusion per individual ranges from six to 12 units. Urological injuries occurred in 8% (4% bladder and 4% ureter) in the Irish studies.66 This is similar to 7.5% of urological injuries observed in the Nigerian study.73 Ongoing bleeding and the need for re-laparotomy is also a known consequence of the procedure, being required in 8–18% of operations. The Nigerian study gave a high mortality rate of 12.5%, and the investigators suggested that mortality after caesarean hysterectomy would be greater in situations where there are shortages of blood products and non-availability of intensive care facilities. This would be the scenario in many rural hospitals in low-income countries. The Nigerian studies also showed caesarean hysterectomy to be associated with a high perinatal mortality rate of 250 per 1000.73 This is a result of conditions such as uterine rupture and obstructed labour causing fetal demise in addition to the massive bleeding. This is reﬂective of many low-income countries where prolonged labour cannot be managed timeously, resulting in perinatal loss and the mother losing her chance of future fertility. If haemostasis is not satisfactory after STAH, a suction drain can be left in situ. If coagulopathy is evident after STAH and total abdominal hysterectomy, then it is useful to carry out abdominal packing to tamponade the abdominal cavity. At least ﬁve paediatric swabs followed by abdominal swabs can be used. The woman will need to be kept ventilated and the packs removed after 48 h. The woman should be observed in a high-care area where vital signs, including urine output, can be monitored and blood products replaced as necessary. The need for adequate psychological counselling to such women and their families after surgery cannot be over emphasised. Follow up after discharge is required for clinical reasons, but also for further counselling, especially if the baby died. Conclusion Excessive bleeding at caesarean section causes severe morbidity and mortality. Because caesarean section rates are increasing worldwide and particularly in low and middle income countries, preventative measures such as timeous operative delivery for prolonged labour, and careful surgical techniques together with careful post-operative observations need to be emphasised. Finally, bleeding during and after caesarean section not responding to conservative medical and/or surgical measures, requires the consideration of caesarean hysterectomy to save the mother’s life.

Practice points # Excessive bleeding at caesarean section causes severe maternal morbidity and mortality. # Preventive measures, such as timeous caesarean section for prolonged labour, oxytocin and careful surgical technique, can reduce bleeding at caesarean section. # Vigilant postoperative monitoring can allow earlier detection and treatment. # Algorithms are available to guide management for the various causes of bleeding during and after caesarean section. # Management requires team work between surgeon, anaesthetist and midwives. # Surgeons who carry out caesarean section need to know the medical treatment of uterine atony and be able to carry out additional surgical procedures for arresting haemorrhage and preferably subtotal hysterectomy. # Bleeding after caesarean section may be concealed (intra-abdominal) as well as revealed (per vaginam). # Bleeding after caesarean section not responding to medical or conservative measures requires urgent re-laparotomy.

S. Fawcus, J. Moodley / Best Practice & Research Clinical Obstetrics and Gynaecology xxx (2012) 1–17

Research agenda # Prospective recording in registers of women undergoing conservative surgical methods, such as compression sutures and vessel ligations, so long-term complications and future fertility can be evaluated. # Morbidity surveillance for caesarean section in poorly resourced settings. # Evaluation of techniques that can be carried out by non-gynaecological surgeons at district hospitals to enable definitive treatment guidelines for this level of care to be promoted. # Evaluation of a Foley catheter uterine tourniquet as a temporising method during transit or awaiting senior help. # Role of tranexamic acid in the prophylaxis and management of bleeding at caesarean section. References 1. Coker A & Oliver R. Definitions and classifications. In: B-Lynch C, Keith LG, Lalonde AB et al, editors. Textbook of post partum haemorrhage. Dumfries: Sapiens Publishing, UK, 2006. 2. Pritchard JA, Baldwin RM, Dickey JC et al. Blood volume changes in pregnancy and the puerperium. Red blood cell loss and changes in apparent blood volume during and following vaginal delivery, caesarean section, and caesaraean section plus total hysterectomy. Am J Obstet Gynecol 1962; 84: 1271–1282. 3. National Centre for Classification in Health. Australian Coding Standards. The International Statistical Classification of Diseases and related health problems. Tenth revision, 7th ed. Australian modification (ICD-10-AM), Sydney, Australia, 2008. 4. Knight M, Callaghan W, Berg C et al. Trends in post partum haemorrhage in high resourced countries; a review and recommendations from the International PPH collaborative group. BMC Pregnancy Childbirth 2009; 9: 55. 5. Stafford I, Dildy GA, Clark SL et al. Visually estimated and calculated blood loss in vaginal and caesarean delivery. Am J Obstet Gynecol 2008; 199: 519.e1–519.e7. 6. Magann EF, Evans S, Hutchinson M et al. Postpartum hemorrhage after cesarean delivery: an analysis of risk factors. S Med J 2005; 98: 681–685. 7. Combs CA, Murphy EL & Laros R. Factors associated with haemorrhage in cesarean deliveries. Obstet Gynecol 1991; 77: 77– 82. 8. Naef RW, Chauhan SP, Chevalier SP et al. Prediction of hemorrhage at cesarean delivery. Obstet Gynecol 1994; 83: 923–926. 9. Schuurman N, Mackinnon C, Lane C et al. Prevention and management of postpartum haemorrhage. J Soc Obstet Gynaecol Can 2000; 22: 271–281. 10. Penney G & Brace V. ‘Near misses’ and severe maternal morbidity; the Scottish experience. In: Lewis G, editor. Why mothers die 2000–2002. London: RCOG Press, 2004, pp. 267–271. 11. Brace V, Kernaghan D & Penney G. Leaning from adverse clinical outcomes: major obstetric haemorrhage in Scotland, 2003–2005. BJOG 2007; 114: 1388–1396. 12. Zwart JJ, Richters JM, Ory F et al. Severe maternal morbidity during pregnancy, delivery and puerperium in the Netherlands: a nationwide population-based study of 371,000 pregnancies. BJOG 2008; 115: 842–850. *13. Ronsmans C. Severe acute maternal morbidity in low-income countries. Clin Obstet Gynaecol 2009; 23: 305–316. 14. Fawcus S, Mbombo N & Hofmeyr GJ. Saving mothers, chapter 4. The fifth report of the national confidential enquiry into maternal deaths in South Africa 2008–2010. A short report. Pretoria: Department of Health, 2012. 15. Fawcus S & Mbombo N. Deaths from haemorrhage associated with caesarean section in South Africa 2009. Abstract FIGO. Cape Town, South Africa. *16. Cantwell R, Clutton-Brock T, Cooper G et al. Saving Mothers’ lives: reviewing maternal deaths to make motherhood safer: 2006–2008. The eighth report of the confidential enquiries into maternal deaths in the United Kingdom. BJOG 2011; 118(Suppl. 1): 71–76. 17. Berghella V, Baxter JK & Chauhan SP. Evidence based surgery for caesarean delivery. Am J Obstet Gynecol 2005; 193: 1607. 18. CotterA Ness A & Tolosa J. Prophylactic oxytocin in the third stage of labour. Cochrane Database Syst Rev 2001; 4. CD001808. *19. WHO guidelines for the management of postpartum haemorrhage and retained placenta. Geneva: Geneva: World Health Organization, 2009. *20. Prevention and management of postpartum haemorrhage. RCOG Green-top Guideline 2009; 52. update April 2011. 21. Dyer RA, van Dyk D & Dresner A. The use of uterotonic drugs at caesarean section. Int J Obstet Anaethesia 2010; 19: 313–319. 22. Munn MB, Owen J, Vincent R et al. Comparison of two oxytocic regimens to prevent uterine atony at cesarean delivery; a randomized controlled trial. Obstet Gynecol 2001; 98: 386–390. *23. National Collaborating Centre for Women and Children’s Health. Caesarean section. Clinical Guideline. London: RCOG Press, 2004, www.rcog.org.uk/womenshealth/clinical guidance/caesarean section. 24. Dansereau J, Joshi A, Helewa M, Doran T, Lange I, Luther E, Farine D, Schulz M, Horbay G, Griffinn P & Wassenaar W. Double-blind comparison of carbetocin versus oxytocin in prevention of uterine atony after cesarean section. Am J Obstet Gynecol 1999; 180: 670–676. 25. Lourens R & Paterson-Brown S. Ergometrine given during caesarean section and the incidence of delayed postpartum haemorrhage due to uterine atony. J Obstet Gynaecol 2007; 27: 795–797.

S. Fawcus, J. Moodley / Best Practice & Research Clinical Obstetrics and Gynaecology xxx (2012) 1–17

*26. Gulmezoglu AM, Forna F, Villar J & Hofmeyr GJ. Prostaglandins for preventing postpartum haemorrhage. Cochrane Database Syst Rev 2007; 3. CD 000494. 27. CRASH-2 trial collaboration. Effects of tranexamic acid on death, vascular occlusive events, and blood transfusion in trauma patients with significant haemorrhage (CRASH-2): a randomized, placebo-controlled trial. Lancet 2010; 376: 23–32. 28. Gai M, Wu L, Su Q et al. Clinical observation of blood loss reduced by tranexamic acid during and after caesarean section; a multi-center randomized trial. Eur J Obstet Gynaecol Reprod Biol 2001; 112: 154–157. 29. Sekhavat L, Tabatabaii A, Dalili M et al. Efficacy of tranexamic acid in reducing blood loss after caesarean section. J Mat Fetal Neonat Med 2009; 22: 72–75. 30. Gungorduk K, Yildirim G, Asicioglu O et al. Efficacy of intravenous tranexamic acid in reducing blood loss after elective caesarean section; a prospective, randomized, double-blind, placebo- controlled study. Am J Perinataol 2011; 28: 233–240. 31. Royal College of Physicians. National Early Warning Score (MEWS): standardising the assessment of the acute-illness severity in the NHS. Report of a Working Party. London: RCP, 2012. 32. Upadhay K & Scholefield H. Risk management and medicolegal issues related to postpartum haemorrhage. Clin Obstet Gynaecol 2008; 22: 1149–1169. 33. Geoghegan J, Daniels JP, Morre PAS et al. Cell salvage at caesarean section: the need for an evidence based approach. BJOG 2009; 116: 743. 34. National Department of Health South Africa. A monograph of the management of postpartum haemorrhage. National Committee on Confidential Enquiries into Maternal Deaths in South Africa 2010. 35. B-Lynch Coker A, Lawal AH et al. The B-Lynch surgical technique for the control of massive postpartum haemorrhage: an alternative to hysterectomy? Five cases reported. BJOG 1997; 104: 372–375. *36. Steer P. The surgical approach to postpartum haemorrhage. Obstet Gynaecol 2009; 11: 231–238. *37. Allam MS & B-Lynch C. The B-Lynch and other uterine compression suture techniques. Int J Gynecol Obstet 2005; 89: 236–241. *38. Nelson GS & Birch C. Compression sutures for uterine atony and hemorrhage following cesarean delivery. Int J Gynecol Obstet 2006; 92: 248–250. 39. Baskett TF. Uterine compression sutures for postpartum haemorrhage. Obstet Gynecol 2007; 110: 68–71. 40. Fotopoulou C & Dudenhausen JW. Uterine compression sutures for preserving fertility rate in severe postpartum haemorrhage: an Overview 13 years after their first description. J Obstet Gynecol 2010; 30: 339–349. 41. Treloar EJ, Anderson RS, Andrews HS et al. Uterine necrosis following B-Lynch suture for primary postpartum haemorrhage. BJOG 2006; 113: 486–488. 42. Price N & B-Lynch C. Uterine necrosis following B-Lynch suture for primary postpartum haemorrhage. BJOG 2006; 113: 1341. 43. Saroja CSM, Nankani A & El-Hamamy E. Uterine compression sutures, an update: a review of efficacy, safety and complications of B-Lynch suture and other uterine compression techniques for postpartum haemorrhage. Arch Gynecol Obstet 2010; 281: 581–588. 44. Kayem G, Kurinczuk JJ, Alfirevic Z, et alon behalf of the U.K. Uterine compression sutures for the management of postpartum haemorrhage. Obstetric Surveillance System (UKOSS). Obstet Gynecol 2011; 117: 14–20. 45. O’Leary JA. Uterine artery ligation in the control of postcesarean hemorrhage. J Reprod Med 1995; 40: 189–193. 46. Abdrabbo SA. Stepwise uterine devascularisation: a novel technique for management of uncontrolled postpartum hemorrhage with preservation of the uterus. Am J Obstet Gynecol 1999; 171: 694–700. *47. Bakri JN, Amri A & Abdul Jabbar F. Tamponade-balloon for obstetrical bleeding. Int J Gynecol Obstet 2001; 74: 139–142. 48. Georgiou C. Balloon tamponade in the management of postpartum haemorrhage; a review. BJOG 2009; 116: 748–757. 49. Keriakos R & Mukhopadhyay A. The use of the Rusch balloon for management of severe postpartum haemorrhage. J Obstet Gynaecol 2006; 26: 335–338. 50. Akhter S, Begum MR & Kabir J. Condom hydrostatic tamponade for massive postpartum haemorrhage. Int J Gynaecol Obstet 2005; 90: 134–135. 51. Nelson WL & O’Brien J. The uterine sandwich for persistent uterine atony: combining the B-Lynch compression suture and an intrauterine Bakri balloon. Am J Obstet Gynaecol 2007; 196: e9–e10. 52. Ahonen J, Jokela R & Kortila K. An open non-randomised study of recombinant factor V11 in major postpartum haemorrhage. Acta Obstet Gynecol Scand 2007; 51: 929–936. 53. Hossain N, Shamsi T, Haider S et al. Use of recombinant activated factor V11 for massive postpartum haemorrhage. Acta Obstet Gynecol Scand 2007; 86: 1200–1206. 54. O’Connell KA et al. Thromboembolic adverse events after use of recombinant human coagulation factor V11. JAMA 2006; 295: 293–298. 55. Evans S & McShane P. The efficacy of internal iliac artery ligation in obstetric hemorrhage. J Reprod Med 1985; 40: 189–193. 56. Sl Clark, Phelan JP, Yeh SY et al. Hypogastric artery ligation for obstetric hemorrhage. Obstet Gynaecol 1985; 66: 353–356. 57. Joshi VM, Otiv SR, Majumder R et al. Internal iliac artery ligation for arresting postpartum haemorrhage. BJOG 2007; 114: 356–361. 58. Vedantham S, Goodwin SC, Mclucas B et al. Uterine artery embolisation: an underused method of controlling pelvic hemorrhage. Am J Obstet Gynecol 1997; 176: 938–948. 59. Hansch E, Chitkara U, McAlpine J et al. Pelvic arterial embolisation for control of obstetric hemorrhage: a five year experience. Am J Obstet Gynecol 1999; 180: 1454–1460. 60. Post–partum haemorrhage and the WOMAN trial. Int J Epidemiol 2010; 39: 949–950. 61. Doumouchtsis SK, Papageorghiou AT & Arulkumaran S. Systematic review of conservative management of postpartum hemorrhage: what to do when medical treatment fails. Obstet Gynecol Surv 2007; 62: 540–547. 62. NHS Quality Improvement Scotland. Scottish confidential audit of severe maternal morbidity. 6th Annual Report, 2008. NHS Quality Improvement Scotland, 2010. 63. Kayem GF, Kurinczuk JJ, Alfirevic Z et al. Specific second-line therapies for postpartum haemorrhage: a national cohort study. BJOG 2011; 118: 856–864. 64. Gayat E, Resche-Rigon M, Morel O et al. Predictive factors of advanced interventional procedures in a mulicentre severe postpartum haemorrhage study. Inten Care Med 2011; 37: 1816–1825.

S. Fawcus, J. Moodley / Best Practice & Research Clinical Obstetrics and Gynaecology xxx (2012) 1–17

65. Rossi AC, Lee RH & Chmait RH. Emergency postpartum hysterectomy for uncontrolled postpartum bleeding. A systematic review. Obstet Gynecol 2010; 115: 637–644. 66. Roopnarinesingh R, Fay L & Mckenna P. A 27-year review of obstetric hysterectomy. J Obstet Gynaecol 2003; 23: 252–254. 67. Chew S & Biswas A. Caesarean and postpartum hysterectomy. Singapore Med J 1998; 39: 9–13. 68. Baskett TF. Emergency obstetric hysterectomy. J Obstet Gynaecol 2003; 23: 353–355. 69. Forna F, Miles A & Jamieson J. Emergency peripartum hysterectomy: a comparison of cesarean section and postpartum hysterectomy. Am J Obstet Gynecol 2004; 190: 1440–1444. 70. Sheiner E, Levy A, Katz M et al. Identifying risk factors for peripartum hysterectomy. A population based study. J Reprod Med 2003; 48: 622–626. 71. Kacmar J, Bhimani L, Boyd M et al. Route of delivery as a risk factor for emergent peripartum hysterectomy: a case-control study. Obstet Gynecol 2003; 102: 141–145. 72. Knight M, Kurinczuk JJ, Spark P, et alfor UKOSS. Cesarean delivery and peripartum hysterectomy. Obstet Gynecol 2008; 111: 97–105. 73. Okogbenin SA, Gharoro EP, Otoide VO et al. Obstetric Hysterectomy: fifteen years experience in a Nigerian tertiary centre. J Obstet Gynaecol 2003; 23: 356–359. 74. Knight M, & on behalf of UKOSS. Peripartum hysterectomy in the UK: management and outcomes of the associated haemorrhage. BJOG 2007; 114: 1380–1387. 75. Pereira C, Mbaruku G, Nzabuhakwa C et al. Emergency obstetrical surgery by non-physician clinicians in Tanzania. Int J Gynaecol Obstet 2011; 114: 180–183. 76. Chilopora G, Pereira C, Kamwendo F et al. Postoperative outcome of caesarean sections and other major obstetric surgery by clinical officers and medical officers in Malawi. Hum Res Health 2007; 5: 17.