SciELO - Scientific Electronic Library Online

vol.101 issue5Abortion practices undermining reformist laws: expertsCommentary on the need to obtain consent for vaginal delivery author indexsubject indexarticles search
Home Pagealphabetic serial listing  

Services on Demand



Related links

  • On index processCited by Google
  • On index processSimilars in Google


SAMJ: South African Medical Journal

On-line version ISSN 2078-5135
Print version ISSN 0256-9574

SAMJ, S. Afr. med. j. vol.101 n.5 Pretoria May. 2011




Haemorrhage associated with caesarean section in South Africa - be aware



S Fawcus; J Moodley; for the National Committee on Confidential Enquiries into Maternal Deaths (NCCEMD)



The National Committee for Confidential Enquiries into Maternal Deaths (NCCEMD), which since 1998 has produced triennial reports of maternal deaths in South Africa, has identified the increasing rate of haemorrhage during and after caesarean section (CS) as a problem. Obstetric haemorrhage is the third most common cause of maternal death, accounting for 491 of the total 3 959 deaths in 2005 -2007, but constitutes one of the most avoidable causes of maternal death, over 80% of cases being thought to be 'clearly avoidable'.

Deaths from obstetric haemorrhage are further categorised as follows: abruptio placentae, placenta praevia, uterine atony, retained placenta, uterine rupture, uterine inversion and 'other uterine trauma'. This latter group predominantly comprises severe bleeding during and/or after CS, and is the largest group, accounting for 141 deaths (28.7% of all haemorrhage-related maternal mortality) in 2005 -2007.1 Of concern is that this has increased compared with 2002 -2004, when it accounted for 78 deaths (17.6% of haemorrhage mortality). Also, some deaths classified under other causal groups such as uterine atony, abruptio placentae and placenta praevia were due to severe bleeding at CS. This gives a total of 201 maternal deaths during 2005 -2007 with severe bleeding at CS (41% of the total deaths from obstetric haemorrhage). During 2005 -2007 a total of 477 210 CSs were performed in public sector facilities; 35% at district hospitals, 40% at regional hospitals and 25% at tertiary hospitals, giving a national public sector CS rate of 18.4% (District Health Information Systems data).

Therefore, of a total of 477 210 women who underwent CS during the years 2005 -2007, 201 died from haemorrhage at CS, giving a CS haemorrhage fatality rate of 0.042% (4.2 deaths from haemorrhage at CS for every 10 000 CSs performed). Further analysis of 103 of these deaths revealed the following:

• Of the CSs 92.2% were emergencies, compared with 7.8% elective operations.

• Of the CSs 37.9% were performed at district, 41.7% at regional and 20.4% at tertiary hospitals.

• The commonest indications for emergency CS were obstructed labour, abruptio placentae, patient with a previous CS presenting in labour, placenta praevia and fetal distress.

• In many cases, postoperative bleeding followed intra-operative problems with haemostasis, frequently due to an atonic uterus or uterine tears from a distended lower segment after difficult delivery of an impacted head. In a few cases the bleeding originated from the placental bed, such as in placenta praevia or previous CS, and in some surgery was difficult due to multiple adhesions.

• Twenty-seven per cent had a hysterectomy, either at initial CS or at re-laparotomy; 5% had a re-laparotomy but no hysterectomy. Of great concern is that 68% of the total group had no additional conservative surgical measures, no re-look laparotomy and no hysterectomy.

• Blood was transfused in 70% of cases, no blood was given in 28%, and 2% declined blood products.

• Deaths were assessed as clearly avoidable in 85% of cases. Lack of and insufficient blood was a problem in 28%, and staff shortages, particularly for postoperative monitoring, were a factor in 15.5%. Non-functioning theatres at level 1 were a problem for some, and delayed access to theatre due to other emergencies occurred in 6% of cases at level 2 and 3 hospitals. Deficiencies in clinical care, including poor monitoring in the postoperative period (35%), were identified in most cases and led to serious delays in detecting haemorrhagic shock, poor resuscitation (17.5%) and incorrect management (7.8%). There was grossly substandard care in 60% of cases, including failure to secure haemostasis at the initial CS, non-use of potentially effective treatment such as parenteral ergometrine, uterine compression sutures and balloon tamponade, and delays in definitive management such as hysterectomy.

Over the past few months several deaths associated with bleeding during and immediately after CS have been reported, despite dissemination of information on the prevention of deaths from obstetric haemorrhage.


Towards solving the problem

Knowledge and clinical skills in resuscitation, clinical decision making, clinical interpretation and surgical techniques must be improved. Teamwork and administrative support are also essential to prevent and manage severe bleeding associated with CS.

The Health Professions Council of South Africa (HPCSA) stipulation of 4 months in obstetrics and gynaecology during the extended 2-year internship, and the logbook requirement for training in CS, are positive developments. Strengthening ongoing training and ensuring that the logbook requirements are fulfilled require more attention.

The Essential Steps in Management of Obstetric Emergencies (ESMOE)2 training programme for interns includes surgical skills with an emphasis on CS. The 'Haemorrhage' module has algorithms for medical treatment of uterine atony and additional surgical measures to arrest haemorrhage at CS.

The NCCEMD Monograph on the Management of Postpartum Haemorrhage is a practical pocket book for health workers at all levels of care.3 It has a section on the prevention and management of severe blood loss at and after CS, with algorithms. There are also posters on the management of postpartum haemorrhage (PPH) for labour wards, and on B lynch compression sutures and uterine artery ligation for theatres.


Algorithms for bleeding associated with caesarean section

Fig. 1 is an algorithm for bleeding detected during CS, and Fig. 2 an algorithm for bleeding diagnosed in the postoperative period.





Abdominal hysterectomy

Proceed straight to hysterectomy in cases of placenta percreta, a ruptured uterus that is irreparable, or when conservative measures are unsuccessful. A subtotal abdominal hysterectomy (STAH) is usually sufficient to control the bleeding unless there are tears down into the cervix or in some cases of lower segment bleeding following a major placenta praevia, when the cervix must also be removed, i.e. a total abdominal hysterectomy (TAH) must be performed.

If haemostasis is not satisfactory after STAH, a suction drain can be left in situ. If coagulopathy is evident after the STAH/TAH, consider abdominal packing to tamponade the abdominal cavity. At least 5 paediatric swabs followed by abdominal swabs can be used. The patient must be kept ventilated and the packs removed after 24 -48 hours.


Level of care

Stepwise medical treatment of uterine atony, uterine compression sutures, balloon tamponade and uterine artery ligations are skills that must be learnt by doctors performing CS and can be done in a level 1 hospital with emergency blood available. STAH skill may not be available at level 1 hospitals, but should be available at all level 2 and 3 hospitals, which are more appropriate to manage patients with massive haemorrhage


What can we do to reduce the frequency of this problem?

Algorithms, posters and booklets distributed to the relevant health workers are useful, but are insufficient on their own. 'Hands-on' surgical training for all doctors performing CS, and an approach to and demonstration of additional surgical skills to arrest excessive bleeding, must be emphasised. Experienced specialist assistance for difficult cases, preferably on site or in the form of easily available telephonic advice for remote rural hospitals, is also important. Clinical outreach can help maintain skills and surgical training.

Clinical managers of maternity services should monitor bleeding associated with CS as an indicator, and ensure that guidelines are in place, that surgical training occurs and that experienced assistance is available.

Prevention is important, e.g. earlier recourse to CS in cases of prolonged labour and the use of oxytocics before removal of the placenta. Early detection of postoperative bleeding enables earlier treatment. It is important to recognise that postoperative CS haemorrhage may be concealed because the bleeding is intra-abdominal.

The NCCEMD sent the following recommendation to all heads of institutions and training schools involved in maternity care:



For a more detailed approach to the management of bleeding associated with CS, see the PPH Monograph (pages 21 - 24), available free of charge from the NDOH or your provincial Maternal Child and Women's Health Cluster. The information can also be obtained from the website


1. National Committee on Confidential Enquiries into Maternal Deaths. Saving Mothers. Fourth Report on Confidential Enquiries into Maternal Deaths in South Africa 2005-2007. Pretoria: Department of Health, 2009.         [ Links ]

2. National Committee on Confidential Enquiries into Maternal Deaths. Saving Mothers. Essential Steps in the Management of Common Conditions Associated with Maternal Mortality. Pretoria: Department of Health, 2007.         [ Links ]

3. National Committee on Confidential Enquiries into Maternal Deaths. Monograph on the Management of Postpartum Haemorrhage. Pretoria: Department of Health, 2010.         [ Links ]



J Moodley is Professor Emeritus in the Department of Obstetrics and Gynaecology, University of KwaZulu-Natal.
Professor Fawcus is Head of the Department of Obstetrics and Gynaecology, Mowbray Hospital, Cape Town, and an Associate Professor in the Department of Obstetrics and Gynaecology, University of Cape Town.
Members of the NCCEMD: D M Chweneyagae, S Fawcus, R Fuentes, N Godi, N Khaole, B Kunene, M Lekhoathi, M Masasa, Z Mbambisa, N Mbombo, R E Mhlanga, N Molefe, R Molupe, J Moodley, N Moran, R C Pattinson, C Rout, R Seabe, M Schoon, J Sithole.



Corresponding author: J Moodley (

Creative Commons License All the contents of this journal, except where otherwise noted, is licensed under a Creative Commons Attribution License