OBSTETRIC ANAESTHESIA

Obstetric haemorrhage Learning objectives

Rosanne Ching After reading this article, you should be able to:
Thomas Mount C define postpartum haemorrhage
C classify the aetiology of postpartum haemorrhage
Kirsty MacLennan
C list two ways to reduce risk of AFE when using intraoperative
cell salvage

Abstract
Obstetric haemorrhage remains a significant cause of maternal
morbidity and mortality. It is the leading obstetric cause for admission Definitions
to intensive care units. Knowledge of risk factors and early recognition Antepartum haemorrhage (APH) occurs prior to delivery in the
of haemorrhage enables rapid activation of a coordinated multidisci- period of 24 weeks’ gestation to full term of pregnancy. It is much
plinary team response. Clear unit protocols for the management of less common than postpartum haemorrhage. The common cau-
massive haemorrhage that are reinforced by team drills help to in- ses include placenta praevia, placental abruption, trauma and
crease awareness in the multidisciplinary team, improve performance uterine rupture.
and thus can improve patient outcome. Pharmacological agents and Postpartum haemorrhage is defined as primary if blood loss
surgical manoeuvres are reviewed in the article, as are blood conser- from the genital tract occurs within 24 hours of delivery or sec-
vation techniques. ondary if blood loss occurs over 24 hours following delivery up
Keywords Cell salvage; interventional radiology; obstetric haemor- to 6 weeks post-delivery. Primary postpartum haemorrhage
rhage; resuscitation; uterotonic drugs (PPH) is the most common form of major obstetric haemorrhage.
The Royal College of Obstetricians and Gynaecologists define
Royal College of Anaesthetists CPD Matrix: 1A02, 2A04, 2A05, 2A06, PPH as minor (with 500e1000 ml blood loss) or major (with over
2B05, 2B06
1000 ml blood loss). Major PPH can be further classified into
moderate (with between 1000 ml and 2000 ml blood loss) or
severe (with over 2000 ml blood loss).4
Background Other definitions include the need for a transfusion of greater
Obstetric haemorrhage is a significant cause of maternal than four units of packed red cells or a haemoglobin fall of
morbidity and mortality. The Mothers and Babies: Reducing Risk greater than 4 g/dl. Definitions based upon physiological changes
through Audits and Confidential Enquiries across the UK can be unreliable due to the physiological changes associated
(MBRRACE-UK) report published in November 2018, placed with pregnancy.
obstetric haemorrhage as the second most common direct cause
of death and seventh as the overall leading cause of maternal
Postpartum haemorrhage
deaths for the period 2014e2016.1 The mortality rate was 0.78
per 100,000 maternities, which was not significantly different to The aetiology of PPH can be classified as per the ‘Four Ts’:
the previous triennial period (2013e2015).2 The lack of signifi-  tone e abnormalities of uterine contraction
cant decrease in mortality therefore stresses the need to improve  thrombin e abnormalities of coagulation
the care of women with haemorrhage.  trauma e genital tract injury
Paucity of observations following delivery and lack of  tissue e retained products of conception, abnormally
recognition of haemorrhage were implicated in many of the invasive placenta.
cases, more evidently when women were being cared for from Risk factors are summarized in Table 1.4 The occurrence of
outside the theatre, such as recovery or postnatal ward. Human secondary postpartum haemorrhage is associated with retained
factor was identified as a persistent problem, with the lack of products of conception or puerperal sepsis.
leadership and failure to escalate appropriately featuring in
many deaths.1
Management (Figure 1)
Obstetric haemorrhage causes significant morbidity and is the
most common cause of obstetric-related intensive care Team preparation
admissions.3 Early involvement of appropriate senior staff (including the
anaesthetic team and relevant clinicians) is fundamental to the
management of PPH, especially those who are at risk of
abnormal placentation.4 There is some evidence that intensive
Rosanne Ching MBChB is a Trainee in Anaesthesia in the North West educational programmes, comprehensive obstetric haemorrhage
Deanery, UK. Conflicts of interest: none declared.
treatment protocols and post-event team review may individually
Thomas Mount MBChB BSc (Hons) FRCA is a Senior Anaesthetic Trainee or in combination lead to reductions in the incidence of severe
in the North West Deanery, UK. Conflicts of interest: none declared. PPH and thus morbidity.5 The enquiry into maternal morbidity
Kirsty MacLennan MB ChB MRCP FRCA is a Consultant Obstetric from haemorrhage identified improvements in care for almost
Anaesthetist at St Mary’s Hospital, Manchester, UK. Conflicts of 90% of women; in 74% this would have made a difference to
interest: none declared. their outcome.1

ANAESTHESIA AND INTENSIVE CARE MEDICINE 20:9 484 Ó 2019 Published by Elsevier Ltd.
 OBSTETRIC ANAESTHESIA

temperature, change in pallor, pathological cardiotocography

Risk factors and the associated levels of risk for PPH due to placental hypoperfusion to be monitored and detected,
Risk factor The four Ts OR (95% CI) along with the basic observations of tachycardia and
hypotension.
Multiple pregnancy Tone 3.30 (1.00e10.60) Pathways such as massive haemorrhage protocols allow the
4.70 (2.40e9.10) obstetric MDT to obtain blood products quickly to enable rapid
Previous PPH Tone 3.60 (1.20e10.20) empirical treatment of major blood loss when necessary.
Pre-eclampsia Thrombin 5.00 (3.00e8.50) Resuscitation should be informed by laboratory results where
2.20 (1.30e3.70) possible, and there is an increased awareness of the importance
Fetal macrosomia Tone 2.11 (1.62e2.76) of correcting hypofibrinogenaemia in obstetric haemorrhage.6
2.40 (1.90e2.90) More appropriate administration of red cells, blood compo-
Failure to progress in second stage Tone 3.40 (2.40e4.70) nents and coagulation factors may be guided by point-of-care
1.90 (1.20e2.90) testing of haemoglobin values, e.g. Haemocue, and thromboe-
Prolonged third stage of labour Tone 7.60 (4.20e13.50) lastography (TEG) or thromboelastometry (ROTEM). Evidence
2.61 (1.83e3.72) has stated that the use of such tests have an increased sensitivity
Retained placenta Tissue 7.83 (3.78e16.22) in identifying deficits in the coagulation cascade when compared
3.50 (2.10e5.80) to laboratory based tests and an increase in accurate assessment
6.00 (3.50e10.40) of the deficit phase of coagulation cascade.7
Placenta accreta Tissue 3.30 (1.70e6.40) Maintenance of normothermia is also important as both sepsis
Episiotomy Trauma 4.70 (2.60e8.40) and hypothermia increase the oxygen demand of the women and
2.18 (1.68e2.76) may lead to disseminated intravascular coagulation. Calcium
1.70 (1.20e2.50) must also be replaced to support the coagulation system.
Perineal laceration Trauma 1.40 (1.04e1.87)
2.40 (2.00e2.80) Pharmacological management
1.70 (1.10e2.50) Alongside the initial resuscitation of the patient, the first-line
General anaesthesia Tone 2.90 (1.90e4.50) treatment for obstetric haemorrhage is pharmacological inter-
vention to encourage uterine contraction (Table 2).
Adapted from Green-top Guideline No. 52 published by the Royal College of
Obstetricians and Gynaecologists, December 2016.4
Haemostatic agents
Additional agents can be used to promote haemostasis when
Table 1
treating life-threatening haemorrhage.

A combination of prediction and prevention, early recognition Tranexamic acid is a synthetic derivative of the amino acid
and rapid coordinated action in treating PPH will aid in pre- lysine. It is an anti-fibrinolytic agent that competitively inhibits
empting worst-case scenarios. Excellent communication the activation of plasminogen to plasmin, which causes fibrin
amongst the multidisciplinary obstetric team is paramount. degradation. Tranexamic acid reduced death due to bleeding in
women according to the WOMAN Trial Collaborators.8 The risk
Recognizing the haemorrhage of serious adverse events is minimal. The use of tranexamic acid
As placental blood flow at term can exceed 750 ml/min, obstetric should be considered in addition to oxytocin, at caesarean sec-
haemorrhage can be rapid and catastrophic. Failure to recognize tion to reduce blood loss in women who are at increased risk of
the severity of the situation was identified in 26% of deaths in the PPH.4
MBRRACE-UK report.1 This was more evident when the patient
was being cared for out of theatre. Fresh frozen plasma is a blood product that contains all coag-
The initial recognition of obstetric haemorrhage is often ulation factors. If there is ongoing bleeding without imminent
challenging. The haemorrhage may be concealed or in the case of haemostatic results, FFP should be infused at a dose of 12e15
an overt haemorrhage, estimation of total loss is notoriously ml/kg after 4 units of red blood cells transfusion. If prothrombin
inaccurate. Amniotic fluid, loss of blood into sheets and pads, time/activated partial thromboplastin time is more than 1.5 times
blood loss occurring out of hospital and operator underestima- normal and haemorrhage is continuing, volumes of FFP in excess
tion of blood loss are all causes of inaccuracies. of 15 ml/kg are likely to be needed to correct coagulopathy.
The ability of the pregnant patient to compensate for hae-
morrhage with an increase in heart rate, stroke volume and Fibrinogen is a glycoprotein that is converted to fibrin and
vascular tone mean that haemorrhage may go unrecognized until subsequently a blood clot. During ongoing PPH, a plasma
physiological extremes have been reached. In some cases, more fibrinogen level of greater than 2 g/l should be maintained with
than 40% total blood loss can occur before the physiological the use of cryoprecipitate which contains fibrinogen, von Wille-
signs of haemorrhage become recognizable. brand factor, factor VIII, factor XIII and fibronectin.
The use of standardized observation systems such as modified The routine use of recombinant factor VIIa (rFVIIa) is not
early obstetric warning scores (MEOWS) can be vital in recog- currently licensed for use in obstetrics. Although there is some
nizing evolving physiological trends over time. These observa- positive results in case studies, the use of rFVIIa is limited to
tion systems allow signs such as tachypnoea, poor urine output, patients with refractory haemorrhage; where embolization is

ANAESTHESIA AND INTENSIVE CARE MEDICINE 20:9 485 Ó 2019 Published by Elsevier Ltd.
 OBSTETRIC ANAESTHESIA

Initial management and resuscitation

Monitoring and
Communication Resuscitation Arrest the bleeding
investigations

Airway – assess and Continuous monitoring of Consider four Ts and

Alert midwife in charge
administer 15 L/min oxygen pulse, BP, ECG, saturations treat accordingly
via non-rebreathing bag and respiratory rate
Tilt patient if APH to relieve
aorto-caval compression Pharmacological and
Alert obstetrician
Rapid assessment of Hb mechanical measures if
by point-of-care testing, uterine atony present
Breathing – assist if required e.g. blood gas analyser or
securing the airway a Haemocue device
Alert anaesthetist
Remove retained products
Circulation – establish Bloods to the lab including
2 x 14G intravenous arterial blood gas
Alert neonatologist if APH cannula. 20 ml blood sample
sent for full blood count, Repair trauma
coagulation including Consider arterial line
monitoring
Alert haematologist electrolytes, liver function
tests and four units cross Replace clotting factors
match as urgent. Foley catheter to monitor
hourly urine output
Alert blood bank and blood products

Temperature measurement
Transfusion aims:
every 15 minutes
Call porters Haemoglobin (Hb) >8 g/dL
Platelet count >75 x 10/L
Prothrombin <1.5 x mean
control Thromboelastograph to
Assign one member of Activated prothrombin times guide clotting product
team to record events, <1.5 x mean control administration
Fibrinogen >1.0 g/L

Disability – monitor
conscious level, blood
sugar, administer analgesia

via rapid infusor,

minimize exposure

Figure 1

unavailable; or hysterectomy is the only alternative. The initial Retained products of conception should be removed and any
dose administered is 90 mg/kg, and it is specified that prior to genital tract trauma should be identified and repaired.
administration the patient must have a fibrinogen level of greater If uterine atony is the cause of the haemorrhage, uterine
1 g/dl, adequate platelets, normal calcium levels, correction of compression should be initiated. Emptying the bladder is also
acidosis and be normothermic. The main side effects are the advocated.4
occurrence of a thromboembolic event or absence of response. External uterine tamponade by means of bi-manual uterine
compression should be initiated. If this manoeuvre is successful,
Surgical intervention a B-Lynch suture can be sited.
Internal uterine tamponade can be achieved using a Bakri or
Surgical manoeuvres and pharmacological agents are used side
Rusch Balloon. The balloon is placed inside the uterus and is
by side in the treatment of persistent haemorrhage.

ANAESTHESIA AND INTENSIVE CARE MEDICINE 20:9 486 Ó 2019 Published by Elsevier Ltd.
 OBSTETRIC ANAESTHESIA

Pharmacological management
Drug name Action Dose and frequency Side effects Contraindications

Oxytocin A naturally occurring Bolus dose 5 IU slow IV Reflex tachycardia Caution in hypovolaemic
polypeptide (bolus can be repeated if Hypotension and cardiac patients due to
Acts on oxytocin-specific needed) Short acting so vasodilatation and
receptors in uterine often administered as 10 IU/ hypotension
myometrium promoting hr infusion over 4 h
uterine smooth muscle
contraction
Ergometrine Acts on a-adrenergic, 250e500 mg IV or IM Tachycardia Avoid in patients with pre-
dopaminergic and serotonin Hypertension eclampsia and
5-HT2 receptors Vomiting cardiovascular disease
Uterine smooth muscle Nausea
contraction not clearly
associated with a specific
receptor
Carboprost Prostaglandin F2a analogue 250 mg IM. Dose can be Bronchospasm Avoid in asthmatics and pre-
Profound smooth muscle repeated every 15 min Intrapulmonary shunting existing respiratory and
contractor (maximum of eight doses) Hypoxia cardiac disease (not for IV
Vomiting administration)
Diarrhoea
Shivering
Pyrexia
Misoprostol Prostaglandin E1 analogue 800e1000 mg rectally/ Shivering Allergy to misoprostol or
vaginally/sublingually Pyrexia other prostaglandins
Diarrhoea

Table 2

inflated with 500e1000 ml of sterile water. The balloon should transfusion-related immunosuppression, microchimerism,
be used alongside an oxytocin infusion to further promote uter- alloimmunization and the potential of an increase in blood-borne
ine contraction. cancers in recipients.10
Transfusion-related acute lung injury (TRALI) is another real
Uterine artery ligation risk and the most common cause of death following transfusion,
Ligation of the bilateral uterine or internal iliac arteries may accounting for 51% compared with 27% caused by non-ABO and
help to reduce haemorrhage. Ligation of the arteries can be ABO-related haemolytic transfusion reactions.9
technically difficult, and the presence of collateral vessels Risks aside, the availability of blood is also reducing as more
may reduce effectiveness. Early senior involvement is donor restrictions are enforced.
essential. Interventional radiology can be utilized to facilitate For all of these reasons blood conservation techniques
cannulation of the internal iliac arteries. This allows cathe- should be embraced commencing with the recognition of
terization of the uterine arteries, which enables uterine artery antepartum anaemia and treatment aimed at replenishing iron
balloon occlusion or selective uterine artery embolization. stores.
This specialist interventional radiology procedure can be
used in either elective cases (placenta praevia or accreta) or Cell salvage
in the emergency setting. Due to its specialist nature this Intraoperative cell salvage (IOCS) have been endorsed by several
service may not be available in all hospitals, particularly out authoritative bodies including CEMACE, Obstetric Anaesthetists’
of hours. Association and NICE guidelines. It is considered as a clinically
effective and cost-effective means of providing autologous red
Hysterectomy
cells.4 The NICE guideline11 addressing intraoperative blood cell
In drastic cases where uterine bleeding cannot be controlled by
salvage in obstetrics also acknowledges that it may also be useful
any of the above measures, emergency hysterectomy may be
when there are difficulties with cross matching. However, there
needed.
was no statistically significance in the overall reduction observed
Blood conservation techniques in donor blood transfusion associated with the routine use of cell
Early replacement of blood products should be considered. salvage during cesarean section in the SALVO study, a random-
However, transfusion of blood and blood components are not ized controlled trial of IOCS during caesarean section in 3028
without complications.9 There are the associated risks of blood women.12

ANAESTHESIA AND INTENSIVE CARE MEDICINE 20:9 487 Ó 2019 Published by Elsevier Ltd.
 OBSTETRIC ANAESTHESIA

The theoretical safety concerns include infusion of fetal cells Postoperative management
(which could potentially cause haemolytic disease in future preg- Patients who suffer an obstetric haemorrhage will require
nancies) and the potential risk of amniotic fluid embolism (AFE).10 ongoing monitoring and close observation. This patient group
The guidance noted that the theoretical safety concerns will need level two or three care postoperatively to continue
include infusion of fetal cells (which could potentially cause resuscitation and administration of supportive medication and
haemolytic disease in future pregnancies) and the potential risk transfusions and to monitor for ongoing bleeding. A
of amniotic fluid embolism (AFE).10
When considering AFE, the risk can be reduced by the use of
REFERENCES
two suctions, with cell salvage commencing after the majority of
1 Knight M, Bunch K, Tuffnell D, et al. eds. Saving lives, improving
the amniotic fluid has been suctioned away to a separate
mothers’ care e lessons learned to inform maternity care from the
container. Leucocyte depletion filters should also be used.
UK and Ireland confidential enquiries into maternal deaths and
The risk of maternal alloimmunization is also considered to be
morbidity 2014-16. National Perinatal Epidemiology Unit, Univer-
small. Fetal red cells may be entrained into the mother’s circu-
sity of Oxford Oxford, 2018.
lation at any time antenatally or during intrapartum. As such,
2 Knight M, Kenyon S, Brocklehurst P, et al. eds. Saving lives,
any incompatibility between maternal and fetal red cell antigens
improving mothers’ care e lessons learned to inform future ma-
risks maternal alloimmunization, with the risk of haemolytic
ternity care from the UK and Ireland confidential enquiries into
disease in future pregnancies. Rhesus factor incompatibility is a
maternal deaths and morbidity 2009-12. National Perinatal
common problem for Rhesus-negative mothers who therefore
Epidemiology Unit, University of Oxford, 2014.
routinely receive anti-D immunoglobulin to counteract this.
3 Zwart JJ, Dupuis JR, Richters A, et al. Obstetric intensive care unit
Kleihauere Betke testing post-natally enables the appropriate
admission: a 2 year nationwide population-based cohort study.
titration of the anti-D dosage.
Intensive Care Med 2010; 36: 256e63.
The use of IOCS in obstetric haemorrhage is recommended
4 Mavrides E, Allard S, Chandraharan E, et al. on. Prevention and
particularly in consenting patients who refuse blood products.
management of postpartum haemorrhage. BJOG 2016; 124:
Ideally obstetric units would have access to a cell salvage service
106e49.
provided by adequately trained personnel available 24 hours a
5 Rizvi F, Mackey R, Barrett T, McKenna P, Geary M. Successful
day. Close attention to coagulation is essential as IOCS only
reduction of massive postpartum haemorrhage by use of guide-
contains suspended red blood cells.
lines and staff education. Br J Obstet Gynaecol 2004; 111: 495e8.
6 Collis RE, Collins PW. Haemostatic management of obstetric
Anaesthetic considerations
haemorrhage. Anaesthesia 2015; 70: 78e86.
Senior anaesthetists must be involved in the early management 7 Solomon C, Collis RE, Collins PW. Haemostatic monitoring during
and resuscitation of patients suffering a major haemorrhage. postpartum haemorrhage and implications for management. Br J
In many cases, the emergent nature of the case predisposes Anaesth 2012; 109: 851e63.
the use of general anaesthesia. The degree of haemodynamic 8 WOMAN Trial Collaborators. Effect of early tranexamic acid
instability should guide the choice and dose of induction agent. administration on mortality, hysterectomy, and other morbidities
Steps should be taken to reduce the risk of aspiration with in women with post-partum haemorrhage (WOMAN): an interna-
antacid prophylaxis (ranitidine, metoclopramide and sodium tional, randomized, double-blind, placebo-controlled trial. Lancet
citrate) and rapid sequence induction. 2017; 389: 2105e16.
Arterial line monitoring provides accurate and rapid blood 9 Banks A, Norris A. Massive haemorrhage in pregnancy. Cont Educ
pressure monitoring, access for frequent blood sampling and Anaesth Crit Care Pain December 2005; 5: 195e8.
serial blood gas analysis. Central venous access may be required 10 Blood transfusions: more is not necessarily better. Int J Obstet
for vasopressor or inotropic support. Anesth October 2009; 18: 299e301.
Due to the associated cardiovascular instability and potential 11 NICE. Intraoperative blood cell salvage in obstetrics. Interv pro-
coagulopathy, neuraxial anaesthesia may be contraindicated in cedures guidance November 2005, https://www.nice.org.uk/
massive haemorrhage. This should be decided on an individual guidance/ipg144 (accessed 30 April 2019).
basis. 12 Khan K, Moore P, Wilson M, Hooper R, et al. On behalf of the
In an elective setting where major blood loss is anticipated, SALVO study group. Cell salvage and donor blood transfusion
the decision between general and regional anaesthesia is decided during cesarean section: a pragmatic, multicenter randomized
by the anaesthetist, the patient and the anticipated degree of controlled trial. PLoS Med December 2017; 14: e1002471.
blood loss.

ANAESTHESIA AND INTENSIVE CARE MEDICINE 20:9 488 Ó 2019 Published by Elsevier Ltd.