DOI: 10.1111/j.1471-0528.2009.02113.

x
www.blackwellpublishing.com/bjog
Review article

Balloon tamponade in the management of

postpartum haemorrhage: a review
C Georgioua,b
a Graduate School of Medicine, University of Wollongong, Wollongong, New South Wales, Australia b Wollongong Hospital, Department of

Obstetrics and Gynaecology, Illawarra, New South Wales, Australia

Correspondence: Dr C Georgiou, The Wollongong Hospital Academic Suite, Wollongong Hospital, Block C, Level 8, Crown Street, Wollongong,
NSW 2500, Australia. Email georgiou@uow.edu.au

Accepted 31 December 2008.

Obstetric haemorrhage is a significant contributor to worldwide available including the Bakri, Foley, Sengstaken–Blakemore, Rusch
maternal morbidity and mortality. Guidelines for the management and condom catheter. This paper reviews these uterine tamponade
of postpartum haemorrhage (PPH) involve a stepwise escalation of technologies in the management of PPH.
pharmacological and eventual surgical approaches. The method of
Keywords Balloon tamponade, intrauterine, management, post-
uterine tamponade using balloons has recently been added to the
partum haemorrhage, review.
armamentarium for managing PPH. There are various balloons

Please cite this paper as: Georgiou C. Balloon tamponade in the management of postpartum haemorrhage: a review. BJOG 2009;116:748–757.

Background Uterine tamponade

Obstetric haemorrhage is a significant contributor to world- One of the earliest methods of achieving a tamponade effect
wide maternal morbidity and mortality.1,2 In Australia and to control PPH was by uterine packing.15
the UK, haemorrhage features within the top four causes of The possibilities of trauma, infection and ineffective pack-
direct maternal death as reported in the latest triennial ing resulting in concealed bleeding together with the
reports.3,4 increasing effective pharmacological agents to treat uterine
Guidelines for the management of postpartum haemor- atony such as ergometrine and syntocinon resulted in a grad-
rhage (PPH) involve a stepwise approach including the exclu- ual reluctance in use.16 Despite the declining popularity, in
sion of retained products and genital tract trauma. Uterine units where uterine packing was commonly employed, data
atony, which is the most common cause,5 is dealt with uterine suggested that it was effective.17,18 In one series of 163 cases,
rubbing and various uterotonic agents such as oxytocin, ergo- 158 (97%) of these resulted in ‘immediate control of
metrine, misoprostol and prostaglandin F2a (PGF2a).3,6–8 bleeding’.19
If these attempts prove to be unsuccessful and the woman is Sterile gauze was invariably used for uterine packing, but
not already having a caesarean section, a laparotomy is con- more recently, balloon technology has been used to tampo-
sidered. During this time, various surgical interventions may nade the postpartum uterus to control haemorrhage. This
be used. These include internal iliac artery ligation, uterine involves inserting a rubber or silicone balloon into the uterine
compression sutures and peripartum hysterectomy to control cavity and inflating the balloon with normal saline. The
the life-threatening haemorrhage.9–11 balloons in descending order of relative cost include the
Recently, uterine balloon tamponade has been added to Sengstaken–Blakemore tube, the Bakri balloon, the Rusch
this armamentarium in the management of PPH.7,12–14 The balloon, Foley catheters and the condom catheter balloon
purpose of this paper was to review the various uterine tam- (Tables 1 and 2, Figures 1 and 2).
ponade technologies currently available for the management Currently, the intrauterine balloon is believed to act by
of PPH. exerting in inward-to-outward pressure ‘that is greater than

748 ª 2009 The Authors Journal compilation ª RCOG 2009 BJOG An International Journal of Obstetrics and Gynaecology
 Balloon tamponade in the management of postpartum haemorrhage

the systemic arterial pressure’ to prevent continual bleeding.35 Foley catheters, filled to 30–50, 80 and 110 ml, respectively,
More recently, an alternative mechanism of action has been were used in postpartum haemorrhage following unsuccessful
proposed, which involves the hydrostatic pressure effect of the use of pharmacological agents and despite uterine curette.34
balloon on the uterine arteries.28
Many of these balloons have previously been used to con- Sengstaken–Blakemore tube
trol haemorrhage at other anatomical sites, including the
bladder36 and oesophagus,37 as well as to control PPH from The volume of a postpartum uterus was considered too large
vaginal lacerations.38 for an effective tamponade to be achieved by using a 30-ml
Furthermore, these same technologies have been used in Foley catheter balloon as used in gynaecological procedures.45
gynaecological conditions in which bleeding is problematic, for Therefore, the Sengstaken–Blakemore two-balloon tube, orig-
example following first- and second-trimester termination of inally designed for the management of bleeding oesophageal
pregnancy,20,39 cervical pregnancy,40–44 knife cone biopsy,33 laser varices, was used.37 The distal, gastric balloon was filled with
ablation of the endometrium, dysfunctional uterine bleeding,45 300 ml of normal saline to control uterine atony following
multiple vaginal lacerations46 and bleeding from a cervical stump vaginal delivery and manual removal of the placenta23
following a post-caesarean section subtotal hysterectomy.47 (Table 1, Figures 1 and 2).
Subsequently, the proximal oesophageal balloon of the Seng-
staken–Blakemore tube was used24 (Table 1, Figures 1 and 2).
Bakri balloon
Bakri first published the concept of intrauterine balloon tech- Rusch balloon
nology in the management of haemorrhage secondary to pla- The greater cost of the Sengstaken–Blakemore tube in com-
centa praevia–accreta during caesarean section with or parison to the Bakri balloon and the premise that the uterine
without bilateral hypogastric arterial ligation.48 Multiple uri- cavity requires a balloon capable of being insufflated to a large
nary Foley catheters were inserted together with a ‘haemostatic volume resulted in the use of the urological Rusch balloon29
substance’ applied to the oozing inner surface of the lower (Figures 1 and 2). This balloon is reported having an insuf-
uterine segment to function as a ‘haemostatic cushion’. The flation capacity of 1500 ml.30
uterine incision site was then closed, and each of the balloons The Rusch balloon was first used for continual uterine
was inflated with 35–75 cm3 of saline or water. Gentle traction bleeding after removal of a morbidly adherent placenta and
was then applied to obtain a continuous tamponade effect, an adherent succenturiate lobe. In the former case, balloon
and the vagina was packed. application followed a failed Sengstaken–Blakemore tube
The catheters were then tied together, and an examination application and unilateral uterine artery embolisation. The
glove or plastic bag was used for the collection and measure- other uterine artery being inaccessible for embolisation. In
ment of blood loss. This was suggested to help prevent blood both cases, the Rusch urological balloon was inflated with
collection inside the uterine cavity and provide an accurate 400–500 ml of warm saline and removed after 24 hours fol-
estimation of bleeding. lowing deflation at a rate of 20 ml/hour29 (Table 1).
Later, a ‘balloon device for controlling capillary/venous bleed-
ing and surface oozing’ in cases of ‘placenta praevia with variable
Condom catheters
degrees of accretism’ was described49 (Figures 1 and 2). The now
termed Bakri ‘SOS’ (Surgical Obstetric Silicone) balloon was The principle of a fluid-filled structure exerting a tamponade
described with a capacity of up to 500 ml of saline achieving a effect to stop bleeding36,37 has also been exemplified by the use
‘pressure and tamponade effect to control the bleeding state’. of condom catheters in the management of PPH. This ‘Sayeba’s
A subsequent article described the successful use of the method of PPH control’ was used in a prospective study of 23
Bakri (SOS) balloon in four women with PPH resulting from cases31,32,50 (Figures 1 and 2). A latex condom was inserted into
a low-lying placenta/placenta praevia20 (Table 1). the uterus by means of a size 16 rubber catheter and inflated
with 250–300 ml of isotonic saline until the bleeding was con-
trolled. The condom catheter was then removed after 24–48
Foley catheters
hours (Table 1).
Both single and multiple Foley catheters have been used in the Two further cases using this condom catheter were
management of PPH48 (Table 1, Figures 1 and 2). In one case, described in the management of PPH in women with
despite uterine curette and failed uterine packing using dry impaired coagulation32 (Table 1). In the first case, a condom
sterile gauze, five Foley catheters were inserted into the uter- tied with silk to the tip 3–4 cm of the Foley catheter was
ine cavity.33 They were inflated with 80 ml normal saline to placed in uterus. The condom was inflated (250 ml saline)
achieve haemostasis and subsequently removed after 36 hours until bleeding was reduced. The proximal end of catheter was
without further bleeding (Table 1). Three other cases of single ligated to prevent backflow, and the vagina was packed with

ª 2009 The Authors Journal compilation ª RCOG 2009 BJOG An International Journal of Obstetrics and Gynaecology 749
 Georgiou

Table 1. Balloon tamponade devices

Balloon Number Timing of use Cause of PPH Use of Failed surgical Additional surgical Route of
device of cases* (number of cases) (number of cases) uterotonic measures measures balloon
agents (number of cases) placement

Bakrid, t, p 4 Caesarean section (2); Placental site (2) Not mentioned Hypogastric At caesarean
postdelivery (2) a. ligation (2) section from
above and
transvaginal
Bakri 5 Caesarean section (5) Uterine atony (5) Yes Uterine artery ligation B-Lynch suture Transvaginal
(presumed)
Sengstaken– 22 Caesarean section (9); Uterine atony (11); Yes Embolisation (7)
Blakemorea, e, t vaginal delivery (10) retained placenta (5)
or Bakri
Sengstaken– 1 Vaginal delivery Uterine atony/DIC Yes Transvaginal
Blakemorea, g
Sengstaken– 1 Vaginal delivery Uterine atony (coagulopathy) Yes Uterine curette Transvaginal
Blakemorea, e
Sengstaken– 16 Caesarean section (6); Uterine atony (10); Yes Transvaginal
Blakemorea, e vaginal delivery (10) retained placenta (4);
cervical laceration (1);
haematological
I condition (1)
Sengstaken– 17 Caesarean section (9); Uterine atony (10); Yes Ligation of uterine, Embolisation because Transvaginal/
Blakemoreb, c, e vaginal delivery (8) placenta accreta (7); round and utero- bleeding resumed through
genital tract trauma (2) ovarian pedicles when balloon was hysterotomy
withdrawn 2, 4 and
7 hours (3)
Sengstaken– 1 Caesarean section Placenta accreta Yes Oversewn placental bed Transvaginal
Blakemoreb, g
Sengstaken– 1 Vaginal delivery Uterine atony Yes Transvaginal
Blakemoreb, p
Rusch 2 Vaginal delivery Morbidly adherent placenta (1); Yes Left uterine artery Transvaginal
retained cotyledons embolisation, right side
of adherent succenturiate not accessible Failure of SBT
lobe (1)
Ruschp 8 Caesarean section (4); Uterine atony (4); Yes After caesarean
instrumental delivery (1); adherent placenta (3) section (2);
vaginal delivery (3) transvaginal (6)
Condom 23 Caesarean section (6); Uterine atony (20); placenta Yes
catheterp instrumental delivery (3); praevia/morbid adhesion (3)
vaginal delivery (14)
Condom 2 Vaginal delivery Uterine atony (1) Yes Transvaginal
catheterp
Foley 1 Vaginal delivery Uterine atony (recurrent) Yes Uterine packing with
sterile gauze, Uterine curette
Foleyd 2 Vaginal delivery Placental remnants Yes Uterine curette Transvaginal

rolled gauze to prevent the condom catheter from slipping They differ with respect to balloon shape, volume and drain-
out. The bladder was continuously drained, and oxytocin was age of the uterine cavity (Figure 1, Table 2).
administered for 12 hours. The condom catheter was
removed after 32 hours. Balloon shape
The second case involved a woman at 31 weeks of pregnant The shape of the balloons not only differs with respect to each
who presented with placental abruption. Following induction other (Figure 1) but also as they are filled with fluid (Table 2).
of labour and stillborn delivery, a PPH ensued. A condom Furthermore, for those balloons with a drainage channel, the
catheter was inserted and then removed after 24 hours. Both degree to which the distal surface of the balloon contacts the
cases were successfully treated (Table 1). uterine fundus will depend on the length of the drainage tip
(Table 2, Figure 2). In the case of the Sengstaken–Blakemore
balloon, the tip is usually cut to allow a better fit between the
Balloon design
balloon and the uterine fundus. In other studies, the distal
Although the various balloons attempt to achieve a tampo- gastric balloon is folded back when the oesophageal balloon is
nade effect on the uterus, they are not identical in design. insufflated26 (Table 1).

750 ª 2009 The Authors Journal compilation ª RCOG 2009 BJOG An International Journal of Obstetrics and Gynaecology
 Balloon tamponade in the management of postpartum haemorrhage

Balloon Volume in Oxytocin Antibiotic usage Specific Duration of Failure of Failure of Successful References
device balloon infusion (duration) pain relief balloon placement balloon balloon management
(average) after (average) placement tamponade of PPH
placement

Bakrid, t, p 500 ml 20–24 hours 4 Bakri et al.20

Bakri 60–250 ml Prophylactic 10–24 hours (11 hours) 5 Nelson and
(100 ml) (while balloon O’Brien21
in place)
Sengstaken– Bakri: 120–750 ml Epidural 2–59 hours (18 hours) 3 1 18 Dabelea et al.22
Blakemorea, e, t (282 ml); anaesthesia
or Bakri Sengstaken– or intravenous
Blakemore: sedation
(28 6 ml)
Sengstaken– 300 ml Yes Yes (48 hours) 48 hours, deflated 1 Katesmark et al.23
Blakemorea, g at 20 ml/hour
Sengstaken– 50 ml Deflated at 10 hours, 1 Chan et al.24
Blakemorea, e removed at 30 hours
Sengstaken– 70–300 ml 8 hours, 40 units Yes (24 hours) Minimal 8 hours 55 minutes – 2s 14 Condous et al.25
Blakemorea, e (167 ml) in 500 ml analgesia or 43 hour 40 minutes
regional and (26 hours 14 minutes)
general
anaesthetic
Sengstaken– 120–370 ml 2–82 hours Broad spectrum Regional or 3.5–82 hours, deflated 2s 15 Seror et al.26
Blakemoreb, c, e (256 ml) general over 30 hours – mean
anaesthesia
(for
manual
exploration)
Sengstaken– 180 ml 8 hours Cefuroxime/ Spinal 8 hours 1 Frenzel et al.27
Blakemoreb, g metronidazole anaesthesia
Sengstaken– 320 ml Yes Anaesthetic 1 Cho et al.28
Blakemoreb, p
Rusch 400 and 500 ml 24 hours 2 Johanson et al.29
Ruschp 240–1000 ml Over 24 hours, Yes (24 hours) Regional and 6–24 hours, deflated in 1 7 Keriakos and
40 units in 1 l general stages 100–200 ml Mukhopadhyay30
normal saline anaesthesia
Condom 200–500 ml ‘At least 6 hours’ Prophylactic 24–48 hours, deflated 23 Akhter et al.31
catheterp (336.4 ml) antibiotics over 10–15 minutes
amoxycillin,
metronidazole,
gentamicin
(7 days)
Condom 250 ml 12 hours Broad-spectrum Anaesthesia 24 and 32 hours, 2 Bagga et al.32
catheterp response deflated and removed
to ‘fever’
Foley 400 ml–5380 ml ‘Continuous Prophylactic 36 hours 1 De Loor and
infusion’ antibiotics van Dam33
Foleyd 80 and 110 ml Intravenous 1 g cefazolin 7 hours and 24 hours 2 Marcovici
oxytocin before procedure and Scoccia34

a, tip cut; b, tip modification not mentioned; c, more than one cause identified; d, drainage volume 100–460 ml; e, oesophageal balloon used;
g, gastric balloon used; p, vaginal pack used; s, includes genital tract trauma; t, traction used.
DIC; disseminated intravascular coagulopathy SBT, Sengstaken–Blakemore Tube.
*Greater than 20 weeks of gestation.

Users of the Sengstaken–Blakemore balloon suggest that ‘the Balloon volume

tubular oesophageal balloon of the tube would conform more From the product literature, there are different recommended
to the shape of the uterine cavity to achieve a haemostatic effect capacities for each of the balloons or balloon components
compared to the stomach balloon or a Foley catheter’.25 Others (Table 2). This, however, has not prevented them from being
describe the Rusch balloon and the condom catheter as ‘con- used beyond these recommended capacities.22 It is unclear
forming naturally to the contour of the uterus’.29,31 whether these recommended volumes are specific to the

ª 2009 The Authors Journal compilation ª RCOG 2009 BJOG An International Journal of Obstetrics and Gynaecology 751
 Georgiou

balloon material (e.g. silicone or rubber) or due to the

intended site of placement (e.g. diameter of oesophagus). If
Condom catheter the recommended volumes are due to an inherent maximum

a, no drainage of uterine cavity if tip folded; b, distance from balloon base (* in Figure 1) to shaft base (* in Figure 2); c, number of balloons each with insufflation and drainage channels;
www.tyco.com
Tyco (Catheter)

16F (5.3 mm)

of the particular balloon, there is a theoretical potential of

Rubber/latex
—
balloon rupture (see ‘Failures and complications’).

Variable
The product literature on the Bakri balloon (Cook Medical,

N/A
No

No
Bloomington, IN, USA) suggests that a ‘predetermined’ vol-

2
2
ume should be used. This volume would be difficult if not
impossible to calculate as the uterine cavity is likely to distend
www.telemedical.com

as the balloon is insufflated to achieve haemostasis. The

advantage of the ‘tamponade test’ is that it is volume inde-
Rusch

16F (5.3 mm)

500–1500 ml
Rubber/latex

pendent and reaches a clinical end-point of no further bleed-

Telemedical

ing. (See ‘Clinical effectiveness: the tamponade test’.)

28.5
N/A
No

No
2
2

Uterine cavity drainage

Some of the balloons, such as the Rusch balloon and the
Gastric 250 ml, oesophagus 150 ml

condom catheter, do not allow drainage of the uterine cavity.

Sengstaken–Blakemore

Despite being dual channel devices, there is no continuity of

the inner channel with the uterine cavity (Figure 1).
By contrast, the remaining balloon devices do allow drainage
Balloon

of the uterine cavity (Table 2). The Bakri balloon has a relatively
www.tyco.com

large bore drainage channel, whereas the other relatively nar-

16F (5.3 mm)

row bore devices may block due to fibrin formation as drainage

Yes/Noa
Silicone

for these balloons is predominantly by gravity.

Tyco

9.2b

In the case of the Sengstaken–Blakemore tube when the

Yes
49

4c
2

distal tip is folded, the previously available drainage channel

is potentially eliminated26 (Figure 2, Table 2), whereas cut-
www.cookmedical.com

ting the distal tip creates a single wide bore channel for
drainage.23
Bakri

24F (8 mm)

Indications, contraindications and

Silicone
500 ml
Cook

43.5

timing of use
Yes

3.6

No
1
1

At present, the Bakri balloon is the only balloon product that

is specifically designed for ‘the control of postpartum uterine
10–24F (3.3–8 mm)

bleeding’ (Cook Medical; enclosed instruction leaflet

www.tyco.com

J-SOS1106). However, in settings where it is unavailable, or

Foley

considered expensive, other balloons have been used to

Silicone

achieve a similar effect.

30 ml

30.5
Tyco
Table 2. Specifications of tamponade balloons

Yes

Yes
2.5

1
1

Indications for use

The various balloon devices have been used alone or in com-
Manufacturer recommended volume

bination with other surgical interventions, such as internal

iliac artery ligation and the B-Lynch suture.21,51 There is no
Length of catheter shaftb (cm)

Balloon number(s) in device

Length of drainage tip (cm)

specific hierarchy for the sequence of surgical interventions.

Drainage of uterine cavity
Manufacturer/distributor

Their indication for use is usually after pharmacological

Diameter of catheter

N/A, nonapplicable.
Channel number(s)

methods such as oxytocin, ergometrine and misoprostol have

Insufflation valves
Balloon material

proven to be ineffective for uterine atony (Table 1). By com-

Contact details

parison, during certain gynaecological procedures in which

heavy bleeding is anticipated, such as the removal of a cervical
ectopic, the Foley catheter balloon has been inserted as a pro-
phylactic measure.42

752 ª 2009 The Authors Journal compilation ª RCOG 2009 BJOG An International Journal of Obstetrics and Gynaecology
 Balloon tamponade in the management of postpartum haemorrhage

Figure 1. Distal component of tamponade balloons. (Asterisk represents position of suture to attach condom to Foley catheter. Also represents distal point
of measurement of balloon shaft length – Table 2 and Figure 2.)

The Sheffield guidelines suggest the use of the Rusch bal- Obviously, the use of rubber/latex products, such as the
loon ‘as a prophylactic method in cases of women who are at Rusch balloon and the condom catheter, is contraindicated
increased risk of PPH and when PPH would jeopardise the in those with such an allergy.
pre-existing maternal condition’.30
Timing of use
Contraindications Some reports describe the use of the various balloons at lap-
Few contraindications have been highlighted in the use of the arotomy or at caesarean section (Table 1). If a balloon device
balloons. Uterine infection has been mentioned in one report is used prior to laparotomy following a vaginal delivery, it
requiring readmission for endometritis, despite receiving may negate the need for a laparotomy.22,24 If unsuccessful, it
antibiotics for 24 hours. The infection was not solely attrib- will not result in significant delay as insertion is easily
uted to the Rusch balloon as the woman had a prolonged achieved. Furthermore, it may also reduce continuing bleed-
second stage and instrumental delivery.30 Another report ing prior to transfer to the operating theatre or while pre-
commented on a ‘fever’ that responded to antibiotics.32 paring for a laparotomy.25 The early use will allow time for

Figure 2. Proximal component of tamponade balloons. (Asterisk represents proximal point of measurement of balloon shaft; b, d, e, f, i, j, k and m indicate
insufflation portion of balloon; a, c, g and h indicate drainage portion of uterine cavity. Note that l and n do not contribute to balloon tamponade or
drainage. They inflate a balloon within the actual tamponade balloon. There is no drainage of the uterine cavity when using the Rusch and condom catheter
– Table 2 and Figure 1.)

ª 2009 The Authors Journal compilation ª RCOG 2009 BJOG An International Journal of Obstetrics and Gynaecology 753
 Georgiou

resuscitation of the women, obtaining cross-matched blood the balloon is insufflated, it will expand to fit the least resis-
and arrival of senior help (see ‘Clinical effectiveness: the tam- tant space. This may be the vagina in the case of a dilated
ponade test’). cervix unless the balloon is somehow maintained within the
The United Kingdom Obstetric Surveillance System data uterine cavity.
on peripartum hysterectomy have demonstrated that early A device such as a Rampley forcep or the operators fingers
timing of hysterectomy and by inference less blood loss results may be used to gently maintain the distal portion of the
in less maternal morbidity.11 Therefore, the early intervention balloon at the uterine fundus as the balloon is being insuf-
of a balloon device may result in less maternal morbidity flated. However, subsequent traction of the balloon, as recom-
secondary to reduced blood loss. mended in the cases of placenta praevia,20 may result in the
balloon being displaced into the vagina if it is not insufflated
sufficiently through a dilated cervix.
Practical considerations The option of ‘over-inflating’ the balloon in the uterus to
In using the various balloons for the management of PPH, prevent migration may cause other problems.30 The first is
there are a few practical considerations that arise. These that distension of the uterus causes significant pain (see ‘Pain
include insertion of the balloon device, use of a vaginal pack, relief ’) and therefore one should aim for the minimal amount
continuing oxytocin infusion, antibiotic usage, pain relief, of uterine distension to accomplish haemostasis. The second
rate of balloon deflation, timing of removal and clinical effec- problem is a theoretical concern of uterine rupture (see ‘Fail-
tiveness. These issues are discussed below. ure and complications’).
If a vaginal pack is to be used, then a positive tamponade
Insertion of balloon device test needs to be demonstrated prior to placement of the vag-
The various uterine balloons are described as being inal pack. Otherwise, there is a danger that the pack will
‘inserted’,23–26,29 ‘placed’32 or ‘introduced’,33,34 although there obscure any continuing bleeding leading to a delayed diagno-
are few specific details as to exactly how this is accomplished. sis of ineffective tamponade.
The Rusch and the Bakri balloon have been described as being
inserted transvaginally using ring forceps to hold the cervix Oxytocin infusion
and inserting the Rusch balloon with a sponge holder forceps. Despite the use of the various balloons for different causes of
Alternatively, the balloon is ‘inserted digitally in the same PPH, there is no evidence that an oxytocin infusion is oblig-
manner as an intrauterine pressure catheter’.22 The Bakri atory for all causes of PPH (Table 1). The majority of pub-
(Cook Medical) product information leaflet suggests ‘using lications describe the use of continual oxytocin infusion
ultrasound guidance’. Ultrasound scan can also used to con- following balloon placement (Table 1). However, little spe-
firm correct placement.28,32 cific information is available with respect to the various con-
At laparotomy following a caesarean section, some reports centrations, rates and duration of use.
describe the balloon being placed abdominally and then If the syntocinon is continued for the duration of balloon
insufflated after the uterine incision is closed. This may placement, this can range from 2 to 82 hours (Table 1). In
potentially result in balloon failure secondary to damaging such cases of prolonged syntocinon use, there is no mention
the balloon by the suturing needle (see ‘Failure and of monitoring the plasma sodium ion concentration. The
complications’). possibility of hyponatraemia secondary to the cross-reactivity
An alternative approach is to close the uterus first and of the oxytocin with antidiuretic hormone receptors and
then insert the balloon from the vagina, applying the tam- resultant need to fluid restrict is overlooked. This may be
ponade test before closing the laparotomy site. This has the further exacerbated as these women are usually loaded with
advantage of allowing visualisation of the uterus following fluid (blood/blood products/saline) in an attempt to resusci-
insufflation. tate them.
Carbetocin, a synthetic analogue of oxytocin, with a half-
Use of a vaginal pack life of 4–10 times that of oxytocin is available.52 There were no
The early publications involving the Bakri balloon suggested significant changes in sodium, potassium or chloride values
the use of a vaginal pack to maintain the balloon in the vagina from predrug levels after a single dose of carbetocin when
(Table 1). The use of a vaginal pack in the form of ribbon measured at 6, 24 and 72 hours after intravenous injection
gauze is recommended in the Sheffield guidelines for the use in nonpregnant women.53 Therefore, this may be a preferred
of the Sengstaken–Blakemore tube.30 This was also used in the drug in the presence of a uterine balloon for prolonged uter-
studies involving condom catheters, or alternatively, a second ine contraction.
inflated condom was used in the vagina31,32 (Table 1). Although not specifically mentioned, another means of
However, the vaginal pack may only be necessary in cases of increasing uterine tone is to encourage breastfeeding. How-
PPH involving a dilated cervix. The reason for this is that as ever, this may be impractical or declined by the mother.

754 ª 2009 The Authors Journal compilation ª RCOG 2009 BJOG An International Journal of Obstetrics and Gynaecology
 Balloon tamponade in the management of postpartum haemorrhage

Antibiotic usage Although reports of ‘Success’ and ‘Failure’ in the use of the
Antibiotic usage is not empirical. The main aim is to reduce balloons for ‘obstetric haemorrhage’ exist14, they do not nec-
the risk of iatrogenic infection caused by contamination of the essarily include specific indications, methods used, balloon
uterine environment by the balloon from the vaginal environ- type or reasons for failure. Detailed analysis of the cases tab-
ment. Antibiotics are generally administered at the time of ulated in Table 1 identifies one true failure of tamponade22
caesarean section or laparotomy. In the studies identified, the not attributed to unidentified genital tract trauma or sponta-
antibiotic used is usually a cephalosporin. The duration may neous expulsion.25,26,30
be prophylactic (single dose), continued for 24–48 hours or Literature with respect to the use of the Sengstaken–
recommended for the duration of balloon usage30 (Table 1). Blakemore tube in the management of oesophageal bleeding
describe a number of potential, but as yet unreported compli-
Pain relief cations. These include ulceration from the pressure effect of the
Analgesia and anaesthesia are not specifically mentioned in balloon in the uterus or vagina especially with prolonged use,54
a number of the studies identified (Table 1). The initial place- unrecognised exsufflation,55,56 uterine rupture from uterine
ment of the balloons following a vaginal delivery may not overdistension57 and uterine perforation during insertion.58
require an anaesthetic, but ‘analgesia (pethidine) may be Other potential complications include inadvertent perfora-
used’.32 It may also be inserted when pain relief has already tion of a previously sited uterine balloon during the admin-
been achieved, for example in the case of a caesarean section istration of intramyometrial PGF2a and air emboli if air is
or laparotomy. used as the distension medium for the balloon.
Similarly, there is no specific mention of pain relief after
insertion. The distended uterus does cause discomfort that
Future pregnancies
can be alleviated by reducing the insufflated balloon slightly.
However, a balance must be achieved with respect to the At present, there is a single pregnancy reported following the
tamponade effect and analgesia requirements. use of the Rusch balloon29 and two pregnancies following the
use of a Bakri balloon in combination with a B-Lynch suture.21
Rate of deflation
Most papers have removed the balloon within 48 hours
(Table 1). Rates of deflation vary from 20 ml/hour to half
Summary
the volume in the balloon at 12 hours.23,30 The timing of Postpartum haemorrhage (PPH) is a potentially life-threat-
removal is also suggested to correlate with the availability of ening event. In the majority of cases, relatively simple meth-
senior staff, in case there is continuing bleeding.30 ods are used to avert a disaster, although these are not always
employed.11
Clinical effectiveness: the tamponade test Uterine tamponade using intrauterine balloons appears to
Various descriptions describe filling the balloons until bleed- be an effective tool in the management of PPH. Overall, from
ing is controlled.22,31,50 This tamponade test25 is considered the case reports, retrospective22,26,30 and prospective stud-
‘positive’ if control is achieved following inflation of the bal- ies,25,31 97/106 (91.5%) cases were successful when the various
loon.7 Although this test was originally coined with reference balloons have been used (Table 1).
to using the Sengstaken–Blakemore tube, it is equally appli- Given that the technology is simple to deploy and with
cable for any of the balloons.25,27 The tamponade test serves to minimal adverse effects, a balloon tamponade method should
formalise the stages of managing the PPH as a ‘negative’ tam- become a familiar component of existing guidelines for the
ponade test (control of PPH not achieved following inflation management of PPH, although not as an isolated form of
of the balloon) suggests that further management, such as therapy.
laparotomy, or an early course to hysterectomy is necessary7 It is hoped that this review paper increases the awareness of
(see ‘Indications, contraindications and timing of use’). the various tamponade balloons and contributes to an evi-
dence-based appraisal of its place in the management of PPH.
Failures and complications
Disclosure of interest
Few studies report difficulties or failures in using the balloons. None.
Some of these ‘failures’ may be interpreted as ‘complications
of placement’. These include obstruction by uterine leiomyo- Contribution to authorship
mata, inadvertent damage to the balloon during preparation CG reviewed the literature and prepared the manuscript
of Sengstaken–Blakemore tube while cutting off the tip,
inability to place the balloon due to the presence of a B-Lynch Details of ethics approval
suture22 and insufficient insufflation requiring two balloons.26 Nonapplicable.

ª 2009 The Authors Journal compilation ª RCOG 2009 BJOG An International Journal of Obstetrics and Gynaecology 755
 Georgiou

Funding 20 Bakri YN, Amri A, Jabbar FA. Tamponade-balloon for obstetrical bleed-
None. ing. Int J Gynaecol Obstet 2001;74:139–42.
21 Nelson WL, O’Brien JM. The uterine sandwich for persistent uterine
atony: combining the B-Lynch compression suture and an intrauterine
Acknowledgement Bakri balloon. Am J Obstet Gynecol 2007;e9–10.
The author would like to acknowledge the excellent and efficient 22 Dabelea V, Schultze PM, McDuffie RS. Intrauterine balloon tamponade
library staff at the Wollongong Hospital (Christine Monie, Sharon in the management of postpartum hemorrhage. Am J Perinatol 2007;
Hay and Vivienne Caldwell). j 24:359–64.
23 Katesmark M, Brown R, Raju KS. Successful use of a Sengstaken-Blakemore
tube to control massive postpartum haemorrhage. Br J Obstet Gynaecol
References 1994;101:259–60.
1 Lalonde A, Daviss BA, Acosta A, Herschderfer K. Postpartum haemor- 24 Chan C, Razvi K, Tham KF, Arulkumaran S. The use of a Sengstaken-
rhage today: ICM/FIGO initiative 2004-2006. Int J Gynaecol Obstet Blakemore tube to control post-partum hemorrhage. Int J Gynaecol
2006;94:243–53. Obstet 1997;58:251–2.
2 World Health Organization. Attending to 136 million births, every year: 25 Condous GS, Arulkumarah S, Symonds I, Chapman R, Sinha A, Razvi K.
make every mother and child count: chapter 4: risking death to give The ‘‘tamponade test’’ in the management of massive postpartum
life. The World Health Report 2005.Geneva, Switzerland: WHO; 2005. hemorrhage. Obstet Gynecol 2003;101:767–72.
3 Liston W. Haemorrhage. In: Lewis G, editor. The Confidential Enquiry 26 Seror J, Allouche C, Elhaik S. Use of Sengstaken-Blakemore tube in
into Maternal and Child Health (CEMACH). Saving Mothers’ Lives: massive postpartum hemorrhage: a series of 17 cases. Acta Obstet
Reviewing Maternal Deaths to Make Motherhood Safer—2003-2005. Gynecol Scand 2005;84:660–4.
The Seventh Report on Confidential Enquiries into Maternal Deaths in 27 Frenzel D, Condous GS, Papageorghiou AT, McWhinney NA. The use of
the United Kingdom. London: CEMACH; 2007. pp. 78–85. ‘tamponade test’ to stop massive obstetric haemorrhage in placenta
4 Sullivan EA, Hall B, King JF. Maternal Deaths in Australia 2003-2005. accreta. BJOG 2005;112:676–7.
Maternal Deaths Series No. 3 Cat. no. PER 42. Sydney, Australia: AHW 28 Cho Y, Rizvi C, Uppal T, Condous G. Ultrasonographic visualization of
National Perinatal Statistics Unit, 2007. balloon placement for uterine tamponade in massive primary postpar-
5 Al-Zirqi I, Vangen S, Forsen L, Stray-Pedersen B. Prevalence and risk tum hemorrhage. Ultrasound Obstet Gynecol 2008;32:711–13.
factors of severe obstetric haemorrhage. BJOG 2008;115:1265–72. 29 Johanson R, Kumar M, Obhrai M, Young P. Management of massive
6 Somerset D. The emergency management of catastrophic obstetric postpartum haemorrhage: use of a hydrostatic balloon catheter to
haemorrhage. Obstet Gynaecol 2006;8:18–22. avoid laparotomy. BJOG 2001;108:420–2.
7 Royal College of Obstetricians and Gynaecologists. RCOG Draft 30 Keriakos R, Mukhopadhyay A. The use of the Rusch balloon for man-
Guideline. Prevention and Management of Postpartum Haemor- agement of severe postpartum haemorrhage. J Obstet Gynaecol 2006;
rhage. 2008. 26:335–8.
8 Mousa HA, Walkinshaw S. Major postpartum haemorrhage. Curr Opin 31 Akhter S, Begum MR, Kabir Z, Rashid M, Laila TR, Zabeen F. Use of
Obstet Gynecol 2001;13:595–603. a condom to control massive postpartum hemorrhage. MedGenMed
9 Tamizian O, Arkulkumaran S. The surgical management of postpartum 2003;5(3):1–9. [www.meditorscape.com\viewarticle\459894]. Accessed
haemorrhage. Curr Opin Obstet Gynecol 2001;13:127–31. 29 November 2008.
10 Doumouchtsis SK, Papageorghiou AT, Arulkumaran S. Systematic review 32 Bagga R, Jain V, Sharma S, Suri V. Postpartum hemorrhage in two
of conservative management of postpartum hemorrhage: what to do women with impaired coagulation successfully managed with condom
when medical treatment fails. Obstet Gynecol Surv 2007;62:540–7. catheter tamponade. Indian J Med Sci 2007;61:157–8.
11 Knight M; On behalf of UKOSS. Peripartum hysterectomy in the UK: 33 De Loor JA, van Dam PA. Foley catheters for uncontrollable obstetric or
management and outcomes of the associated haemorrhage. BJOG gynecologic hemorrhage. Obstet Gynecol 1996;88:737–8.
2007;114:1380–7. 34 Marcovici I, Scoccia B. Postpartum hemorrhage and intrauterine bal-
12 Rogers MS, Chang AMZ. Postpartum haemorrhage and other prob- loon tamponade. J Reprod Med 1999;44:122–6.
lems of the third stage. In: James DK, Steer PJ, Weiner CP, Gonik B, 35 Arulkumarah S, Condous G. The ‘‘tamponade test’’ in the management
editors. High Risk Pregnancy: Management Options, 3rd edn. Chapter of massive postpartum hemorrhage. Obstet Gynecol 2003;102:641–2.
77. Philadelphia, PA, USA: Elsevier/Saunders; 2006. pp. 1566–1571. 36 Helmstein K. Treatment of bladder carcinoma by a hydrostatic pressure
13 Baskett TF, Calder AA, Arulkumaran S. In Munro Kerr’s Operative Obstet- technique. Br J Urol 1972;44:434–50.
rics. Centenary 11th edn. Philadelphia, PA, USA: Saunders/Elsevier, 2007. 37 Sengstaken RW, Blakemore AH. Balloon tamponage for the control of
14 Brace V, Kernaghan D, Penney G. Learning from adverse outcomes: hemorrhage from esophageal varices: Sengstaken and Blakemore.
major obstetric haemorrhage in Scotland, 2003-05. BJOG 2007;114: Ann Surg 1950;131:781–9.
1388–96. 38 Tattersall M, Braithwaite W. Balloon tamponade for vaginal lacerations
15 Ramsbotham PH. The Principles and Practice of Obstetrical Medicine causing severe postpartum haemorrhage. BJOG 2007;114:647–8.
and Surgery. Philadelphia, PA, USA: Blanchard and Lea, 1856. p. 371 39 Olamijulo JA, Doufekas K. Intrauterine balloon tamponade for uncon-
p. 415–416. trollable bleeding during first trimester surgical termination of preg-
16 Douglass LH. The passing of the pack. Bull Sch Med Univ Md 1955;40: nancy. J Obstet Gynaecol 2007;27:441–2.
37–9. 40 Thomas RL, Gingold BR, Gallagher M. Cervical pregnancy. J Reprod
17 Drucker M, Wallach RC. Uterine packing: a reappraisal. Mt Sinai J Med Med 1991;36:459–62.
1979;46:191–4. 41 Okeahialam MG, Tuffnell DJ, O’Donovan PO, Sapherson DA. Cervical
18 Maier RC. Control of postpartum haemorrhage with uterine packing. pregnancy managed by suction evacuation and balloon tamponade.
Am J Obstet Gynecol 1993;169:317–23. Eur J Obstet Gynecol Reprod Biol 1998;79:89–90.
19 Lester WM, Bartholomew RA, Colvin ED, Grimes WH, Fish JS, Galloway 42 Fylstra DL, Coffey MD. Treatment of cervical pregnancy with cerclage,
WH. Reconsideration of the uterine pack in postpartum hemorrhage. curettage and balloon tamponade: a report of three cases. J Reprod
Am J Obstet Gynecol 1965;93:321–9. Med 2001;46:71–4.

756 ª 2009 The Authors Journal compilation ª RCOG 2009 BJOG An International Journal of Obstetrics and Gynaecology
 Balloon tamponade in the management of postpartum haemorrhage

43 Bakour SH, Thompson PK, Khan KS. Successful conservative manage- 51 Danso D, Reginald P. Combined B-Lynch suture with intrauterine bal-
ment of cervical ectopic pregnancy with combination of methotrexate, loon catheter triumphs over massive postpartum haemorrhage. BJOG
mifepristone, surgical evacuation and tamponade using a double bal- 2002;109:963.
loon three-way catheter. J Obstet Gynaecol 2005;25:616–18. 52 Dansereau J, Joshi AK, Helewa ME, Doran TA, Lange IR, Luther ER,
44 De La Vega GA, Avery C, Nemiroff R, Marchiano D. Treatment of early et al. Double-blind comparison of carbetocin versus oxytocin in pre-
cervical Pregnancy with cerclage, carboprost, curettage and balloon vention of uterine atony after cesarean section. Am J Obstet Gynecol
tamponade. Obstet Gynecol 2007;109:505–7. 1999;180:670–6.
45 Goldrath MH. Uterine tamponade for the control of acute uterine 53 Sweeney G, Holbrook AM, Levine M, Yip M, Alfredsson K, Capi S, et al.
bleeding. Am J Obstet Gynecol 1983;147:869–72. Pharmacokinetics of carbetocin, a long-acting oxytocin analogue, in
46 Condi RG, Buxton EJ, Payne ES. Successful use of the non-pregnant women. Curr Ther Res 1990;47:528–40.
Sengstaken-Blakemore tube to control massive postpartum haemor- 54 Read AE, Dawson AM, Kerr DNS, Turner MD, Sherlock S. Bleeding
rhage. Br J Obstet Gynaecol 1994;101:1023–4. oesophageal varices treated by oesophageal compression tube. BMJ
47 Fahy U, Sved A, Burke G. Successful balloon tamponade of post cae- 1960;1:227–31.
sarean hysterectomy pelvic bleeding: a case report. Acta Obstet Gyne- 55 Conn HO, Simpson JA. Excessive mortality associated with balloon
col Scand 2003;82:97–8. tamponade of bleeding varices. JAMA 1967;202:135–9.
48 Bakri YN. Uterine tamponade-drain for hemorrhage secondary to pla- 56 Pitcher JL. Safety and effectiveness of the modified Sengstaken-Blakemore
centa previa-accreta. Int J Gynaecol Obstet 1992;37:302–3. tube: a prospective study. Gastroenterology 1971;61:291–8.
49 Bakri YN. Balloon device for control of obstetrical bleeding. Eur J 57 Francis PN, Perkin KW, Pain MCF. Rupture of the oesophagus fol-
Obstet Gynecol Reprod Biol 1999;86:S33–S101; S84. lowing use of the Sengstaken-Blakemore tube. Med J Aust 1963;1:
50 Akhter S, Begum MR, Kabir J. Condom hydrostatic tamponade for 582–4.
massive postpartum hemorrhage. Int J Gynaecol Obstet 2005;90: 58 Conn HO. Hazards attending the use of esophageal tamponade. N
134–5. Engl J Med 1958;259:701–7.

ª 2009 The Authors Journal compilation ª RCOG 2009 BJOG An International Journal of Obstetrics and Gynaecology 757