Nutrition

Increasing evidence is emerging on the importance of nutrition to health. With specific reference to haemorrhagic deaths, folate supplementation before pregnancy reduces the risk of placental abruption. In low-resource settings, iron deficiency is the most common cause of anaemia. Nutritional advice should include advice to aim for a diet with adequate folate and iron. In addition, in most low-income settings, iron and folate supplements are recommended for all pregnant women. A particularly common dietary idiosyncrasy in some cultural settings is pica (sand-eating). Sand is thought to chelate iron in the diet inhibiting absorption and aggravating iron-deficiency anaemia.

A key general strategy to reduce deaths due to haemorrhage is to avoid unnecessary caesarean section, both because caesarean section increases the risk of haemorrhagic death in the current pregnancy and because it increases risk in future pregnancies related to uterine rupture, morbidly adherent placenta and the need for repeat caesarean section.

Placenta praevia

Placenta praevia occurs in about five per 1000 pregnancies. It is characterized by bright red, painless bleeding, which may be profuse, and a healthy baby with high presenting part or abnormal lie. Ultrasound is useful for confirming the diagnosis. APH at term due to placenta praevia is managed by delivering the baby. Vaginal birth is possible if the degree of placenta praevia is minor (type 1 posterior – into the lower uterine segment but not reaching the internal cervical os or type 2 anterior – reaching the internal cervical os but not covering the cervical os). Major placenta praevia requires caesarean section. If the diagnosis or degree of placenta praevia is uncertain, cautious vaginal examination is performed in the operating theatre with everything in place for rapid caesarean section if haemorrhage is provoked.

Placental abruption

Bleeding from separation of a normally situated placenta typically presents with persistent abdominal pain with or without dark vaginal bleeding. The uterus may be tense to palpation and in severe cases the baby may die before arrival at the health facility. Tocography may show a characteristic undulating uterine contraction pattern with 9–10 contractions per 10 min and no rest period between contractions. If the fetal heart is still beating, there may be reduced variability and recurrent late decelerations producing an undulating mirror image of the tocograph pattern.

Major placental abruption is associated with massive blood loss (often partially concealed) and coagulopathy. Avoiding maternal mortality requires active management to deliver the baby, replacing blood products, controlling haemorrhage and managing concomitant organ failure (e.g., renal). The vaginal route of delivery is usually safest, though caesarean may be undertaken at times for fetal distress.

Uterine rupture may present as either antepartum or PPH (see below).

Avoiding unnecessary caesarean section and labour induction and augmentation

Caesarean section rates are increasing worldwide. Apart from increased risk of maternal mortality overall, caesarean section specifically increases the risk of PPH. This risk may be further increased by the use of general rather than neuraxial anaesthesia . In the 2011–2013 report of the Committee for Confidential Enquiries into Maternal Deaths in South Africa, caesarean section was highlighted as a contributing factor to deaths due to haemorrhage. Caesarean section also increases the risk of death due to haemorrhage in future pregnancies due to placenta accreta, ruptured uterus or the need for repeat caesarean section.

An important driver for increasing caesarean section rates is overdiagnosis of fetal distress, a problem inherent in cardiotocographic assessment of fetal well-being. In settings where ‘non-reassuring’ fetal heart rate patterns are accepted by courts as evidence of fetal distress, health workers may feel compelled to perform caesarean section in large numbers of cases with benign fetal heart rate decelerations, just in case there should subsequently be a poor outcome. Overdiagnosis of fetal distress may be reduced by using fetal scalp blood sampling in women known not to be infected with HIV.

Similarly, reducing unnecessary labour induction and augmentation may reduce the risk of PPH.

Primary prevention

There is evidence from systematic review that routine ‘active management of the third stage of labour’ (AMTSL) reduces severe PPH (blood loss >1000 ml) by about two-thirds. Adverse effects include hypertension (due to the use of ergometrine), abdominal pain and reduced placental–fetal transfusion. There is less certainty about the relative contribution of the individual elements of AMTSL: routine administration of uterotonics, early cord clamping and active delivery of the placenta with controlled cord traction (CCT).

Oxytocin and ergometrine

Oxytocin intravenously or intramuscularly reduces the risk of PPH (blood loss >500 ml) by about half and is associated with fewer side effects than ergometrine . Addition of oxytocin to ergometrine does not improve effectiveness. The use of oxytocin 10 IU is associated with less need for additional oxytocics than 5 IU, and is therefore the generally recommended dose. From a global health perspective, the greatest limitation of oxytocin is that it requires cold storage and parenteral administration, which limits its availability in rural settings and births not attended by a health professional able to administer injections.

Misoprostol

In the past 20 years, considerable interest has been seen in the use of misoprostol to prevent and treat PPH. The potential game-changing properties of misoprostol are that it is a heat stable and orally active uterotonic, creating the opportunity for self-administration by women giving birth without access to oxytocin. Enthusiasm from early observational studies showing spectacular effectiveness was tempered by less favourable results from two large randomized trials conducted by WHO. First, oral misoprostol 600 μg was found to be much less effective than oxytocin for prevention of PPH (blood loss >1000 ml) . Second, oral misoprostol 800 μg was found to have no benefit over placebo for treatment of PPH in women who had received conventional uterotonics . Misoprostol has been found to reduce blood loss in settings without access to injectable uterotonics, and on this basis it is recommended for use in these settings by WHO. Despite lower effectiveness and higher side effects than oxytocin, such strategies have been considered cost-effective. Distribution of misoprostol to pregnant women with instructions to be taken after birth is an attractive option, which is being rolled out in many countries, though robust evidence of the effectiveness of such strategies is lacking . There remains uncertainty whether any effect on reducing blood loss outweighs adverse effects in terms of overall morbidity and mortality. A systematic review of 78 studies (59,216 women) found no statistically significant difference in maternal mortality for misoprostol compared with control groups overall (11/19,715 vs. 4/20,076 deaths); or for the trials of misoprostol versus placebo (6/4626 vs. 1/4707 deaths) or for misoprostol versus other uterotonics (5/15,089 vs. 3/15 deaths). All 11 deaths in the misoprostol arms occurred in studies of misoprostol ≥600 μg. There was no statistically significant difference in the composite outcome ‘maternal death or severe morbidity’, excluding hyperpyrexia. Pyrexia >38 °C was increased with misoprostol compared with controls and the effect was greater for trials using misoprostol 600 μg or more. In view of the possibility that adverse effects of misoprostol on organ systems such as thermoregulation, cardiovascular and coagulation may counteract the beneficial effects on uterine contraction, we recommend further research to determine whether lower doses may reduce the risk/benefit ratio.

Controlled cord traction

As originally described, CCT involves immediate clamping of the umbilical cord, palpation for contraction of the uterus and then cord traction with transabdominal counter pressure on the uterus to prevent uterine inversion. Following evidence of the benefits of delayed cord clamping to both preterm and term infants, CCT has been modified to allow delayed cord clamping for 1–3 min. Systematic review of trials of CCT found evidence of reduced blood loss >500 but not >1000 ml and reduced manual removal of the placenta, which may have been related to the use of ergometrine in one study and a policy of manual removal after 30 min of third stage of labour in another . In view of limited benefits, and the investment needed in training health providers, the option of omitting routine CCT from the package of AMTSL may be considered.

Uterine massage

Uterine massage is widely practised as a routine method of maintaining uterine contractions and reducing haemorrhage, yet there is surprisingly little research conducted on its effectiveness, as the procedure is uncomfortable for women and costly in terms of health worker time. Systematic review of two studies found inconsistent results and overall limited evidence of effectiveness . The potential to show an effect may have been limited by the fact that all women received routine uterotonics. Further research is needed, particularly in settings in which uterotonics are not available.

Tranexamic acid

Given the effectiveness of tranexamic acid for reducing mortality from traumatic blood loss, until recently there has been surprisingly little research on the use of tranexamic acid to prevent or treat PPH. A potential advantage of tranexamic acid is that its effect is on fibrinolysis rather than uterine contraction; hence, it would be expected to be useful to reduce bleeding both from uterine hypotonicity and from genital tract trauma.

Moderate quality evidence from systematic review of randomized trials of tranexamic acid used prophylactically in low-risk women in addition to uterotonic agents found a decrease in PPH and blood transfusions both after vaginal birth and caesarean section . Minor side effects were increased with the use of tranexamic acid. Because of potential side effects, tranexamic acid is unlikely to be used prophylactically for normal birth, but there may be a place for prophylactic use in women at high risk of haemorrhage and for caesarean section.

Correction of coagulopathies

Observational studies have shown an increased risk of PPH in women with Factor XI deficiency . The risk appears to be reduced with prophylaxis, but no evidence from interventional studies is available.

Secondary prevention: Monitoring blood loss after birth

Recording blood loss after birth is a cornerstone of post-partum care. However, the volume estimated by clinicians is notoriously inaccurate, mainly due to underestimation. A recent review found that additional training in blood loss estimation and multifaceted interventions had limited impact, and that use of blood volume measuring devices did not improve outcomes . The authors suggested that the rate and nature of blood loss may be more important than volume in the early diagnosis of PPH.

Place of birth

The WHO has the vision that ‘every pregnant women and newborn receives quality care throughout pregnancy, childbirth and the postnatal period’ . Although factors that increase risk of PPH may be identified or excluded, PPH may occur unpredictably in apparently low-risk women. Although the risk of PPH may be reduced and certain treatment strategies may be initiated at a primary-level care and even in the home, women with PPH unresponsive to primary-level care need higher-level interventions such as blood transfusion and surgical interventions. Death due to PPH may be rapid, limiting the time available for transfer to a higher level of care. For this reason, in addition to attempts to ensure the presence of a skilled birth attendant at every birth, attention should be given to public health strategies to facilitate access for women in labour to rapidly available higher-level care.

For low-risk women, midwife-led care during childbirth has clear benefits over physician-led care . The principle that all low-risk obstetric care should take place in primary care facilities away from the secondary- and tertiary-level hospitals, is entrenched in many public health services. Although appropriate for antenatal and postnatal care, this paradigm needs to be further investigated with respect to care during labour. Approximately one-third of women with apparently low-risk pregnancies develop complications during or after labour requiring emergency referral to a higher level .

Women in early labour arriving at secondary-level facilities are often turned away to go to a primary-level facility at some distance from the hospital. Women develop strategies to bypass the system by delaying presentation at the hospital until birth is imminent in the hope that it will be too late for them to be redirected. This is counterproductive as it increases the rate of unattended labours and births before arrival at a facility. Within a metropolitan setting, we found that women preferred to give birth on hospital premises where they felt safer because of the availability of higher-level facilities, and most women were able to get to a hospital as easily as to a primary birth facility.

For the above reasons, we have implemented a model of ‘On-site, primary care, midwife-led birth units’ at two hospitals in the Eastern Cape, South Africa . The units are managed and staffed by the primary care services and provide midwife-led, family-centred childbirth care similar to that which would be provided in a clinic at a distance from the hospital. The difference is that the facility is located on the hospital premises. This facilitates transfer to the appropriate level of care. Women arriving at the hospital in labour can be assessed for risk and triaged to the primary- or secondary-level facility as appropriate. When complications arise in the primary care unit, women can be transferred to the hospital labour ward within minutes, with no need for an ambulance transfer. The result has been an increase in births on the hospital premises, but fewer in the secondary-level service, with overall improvement in cost-effectiveness and outcomes. Similar models have been found to be cost-effective in a well-resourced setting, to be safe in a Japanese setting, and in China to be associated with improved satisfaction and labour outcomes . Several on-site birth centres have been reported on from Australia .

Treatment of PPH

Because PPH represents an acute, often life-threatening emergency, research to determine the effectiveness of various methods of treatment is difficult to conduct. Much of the currently recommended treatment is based on extrapolation from methods used for prevention, or empirical treatment based on first principles or historical experience.

Obstetric units should have a written policy for management of acute obstetric haemorrhage, quick availability of all the equipments and consumables necessary and hold regular ‘fire drills’ to ensure that staff are able to respond to the emergency in a coordinated and efficient manner.

Treatment decisions are also complicated by the fact that the cause of the PPH may not be apparent, and there may be more than one cause. An important general principle is to progress rapidly from one intervention to the next until the bleeding is controlled.

Non-pneumatic anti-shock garment

The ‘Medical Anti-Shock Trousers’ (MAST) suit was originally developed to maintain upper body circulation in fighter pilots exposed to extreme gravitational forces, by compressing the lower limbs with a suit with inflatable compartments. The non-pneumatic anti-shock garment (NASG) uses elastic neoprene sections with Velcro fasteners to apply compression to the calves, thighs and lower abdomen, including a ball to compress the uterus. It is used as a first-aid device to compress the lower-body vascular system, increasing relative blood volume and circulation to the upper body during resuscitation and transport. A systematic review evaluated the evidence from five observational studies and one cluster randomized trial. The latter found a statistically non-significant reduction in maternal mortality associated with the NASG and suggested that its use to be considered in settings where access to definitive treatment is delayed.

The use of the NASG in combination with uterine balloon tamponade (see below) has been reported in two cases of massive PPH associated with disseminated intravascular coagulopathy (DIC), with successful outcomes .

The FIGO Safe Motherhood and Newborn Health Committee have published recommendations for the use of the NASG in women with PPH and shock in primary- and higher-level facilities .

Uterotonics

The use of uterotonics to treat PPH is mainly empirical, being extrapolated from evidence based on studies on the use for prophylaxis. There is little direct evidence regarding the effectiveness of additional doses of uterotonics, except that the use of misoprostol 800 mg orally for the treatment of PPH in women who have received oxytocin prophylaxis appears to confer no additional benefits . Empirical treatment often includes ergometrine (for women who are not hypertensive) and/or oxytocin by continuous intravenous infusion.

Uterine massage

As mentioned above, evidence for the effectiveness of uterine massage for preventing PPH is conflicting, and limited to studies of women who have received oxytocics. Considering further research, use of uterine massage to treat PPH is probably more likely to do good than harm, particularly if oxytocics are not available.

Bimanual compression of the uterus

Bimanual compression is used as an emergency first-line treatment based on the general principle of direct pressure at the site of bleeding. The body of the uterus is compressed between the attendant’s hand on the abdomen and fist placed in the anterior vaginal fornix.

Intravascular volume replacement

A key aspect of resuscitation of women with PPH is placement of at least one wide-bore intravenous cannula and rapid volume replacement with crystalloid fluids, pending the availability of blood products.

Supportive treatment

Depending on the mother’s general condition, supportive treatment may range from reassurance to full cardiovascular resuscitation. Monitoring of blood pressure, pulse, respiratory rate, oxygen saturation and urinary output are useful for assessing the need for and response to intervention.

Correction of coagulopathies and blood product transfusion

Massive blood loss is often accompanied by a varying degree of coagulopathy. There is very little robust evidence available on the effectiveness of treatments for these coagulopathies, and management is usually empirical with fibrinogen, tranexamic acid and balanced transfusion of blood products. The use of viscoelastic haemostatic assays may be useful as a guide to management .

Empirical, off-label use of recombinant factor VIIA has recently gained favour in settings where such treatment is affordable.

Suturing of lower genital tract bleeding

For PPH, which does not respond promptly to uterotonic treatment, the perineum, vagina and cervix should be inspected, preferably in an operating theatre with good lighting, and any traumatic sources of bleeding controlled with haemostatic sutures.

Uterine artery clamps

A recent innovation is the use of specialized clamps introduced transvaginally bilaterally with one blade inside the cervix and one in the lateral vaginal fornix to occlude the uterine arteries. The clamps may be fitted with Doppler sensors to locate the uterine artery blood flow before clamping.

Retained placenta

Retained placenta is an important cause of PPH. The definitive management is manual removal of the placenta with or without instrumental evacuation of the uterus. In the case of morbidly adherent placenta, surgical intervention including, in some cases, hysterectomy may be needed. Medical management is an attractive option. A systematic review of the use of prostaglandins (prostaglandin E2 and misoprostol) found insufficient evidence to make a recommendation .

Balloon tamponade

In recent years, there has been increasing advocacy for use of an uterine balloon tamponade (UBT) catheter to treat PPH. Various devices have been used, such as the purpose-designed Bakri balloon, the Sengstaken–Blakemore balloon designed for intra-oesophageal use and various handmade devices using a condom or glove tied over the end of a catheter. A systematic review found no randomized trials . Among 13 case reports, series, retrospective and prospective studies, UBT was reported to be successful in 234/241 cases. The authors concluded that UBT is an effective treatment for PPH in resource-poor settings. It is important to keep in mind that uncontrolled studies may greatly overestimate the effectiveness of interventions, and we would regard UBT as an unproven device pending evidence from randomized trials. Use of UBT is somewhat counter-intuitive as it goes against the principle of emptying the uterus to allow maximum contraction of the uterine muscles. Contraction and retraction with occlusion of spiral arterioles is the main physiological mechanism for post-partum haemostasis.

Suction uterine tamponade

The Inpress Device uses essentially the opposite approach to the UBT. Instead of outward pressure, it applies suction to the uterine cavity by means of a racquet-shaped suction probe with a balloon collar to seal the uterine cervix ( http://www.inpresstechnologies.com/ ). The proposed mechanism of action is that suction promotes uterine contraction and facilitates the physiological mechanism of spiral artery occlusion by contracting uterine muscle fibres. This is similar to the principle of vacuum aspiration evacuation of the uterus, which has been shown to reduce blood loss during uterine evacuation compared with non-suction curettage. As for most methods used to treat PPH, the Inpress Device has not been tested in randomized trials.

Arterial embolization

Selective embolization of arteries supplying the uterus has been used in cases of PPH unresponsive to medical management, to avoid major surgical intervention, but it requires specialized equipment and skills, not widely available.

Major surgical interventions

The traditional approach to PPH unresponsive to medical treatment has been immediate hysterectomy. In recent decades, there has been a trend to conservative surgical procedures that are quicker, require less surgical experience and preserve fertility. The potential risk is that if the conservative measures are not effective, the woman’s condition may have deteriorated further by the time hysterectomy is undertaken.

Aortic compression

Digital compression of the abdominal aorta above the sacral promontory may temporarily reduce the rate of bleeding while definitive methods are instituted.

Uterine tourniquet

Another temporizing measure is to place a uterine tourniquet. An elastic catheter such as a Foley catheter is passed around the uterus as low down as possible. The tourniquet may encompass the infundibulo-pelvic ligaments, or may be brought through avascular areas in the broad ligaments, in which case additional clamping of the ovarian vessels may be needed. The tourniquet may be crossed over in the midline anteriorly, pulled tight and clamped at the point of crossing with a small pair of artery forceps. This provides temporary haemostasis while definitive measures are undertaken.

In settings lacking the facilities for definitive treatment, uterine tourniquets have been left in place for some hours during transfer to a higher-level facility.

Stepwise uterine devascularization

This procedure involves sequential ligation of the uterine arterial supply until haemorrhage ceases . Both uterine arteries; then the descending branches of the uterine arteries, then the ovarian arteries are ligated.

Internal iliac artery ligation

This may be the first procedure, or may follow ineffective stepwise uterine devascularization as described above. The retroperitoneal space is opened, the ureter identified and the internal iliac artery mobilized and double ligated at least 2 cm from its origin.

Uterine compression sutures

Several uterine compression sutures have been described. The B-Lynch suture enters the uterus below and exits above the caesarean section hysterotomy on one side, passes over the fundus, enters the lower uterine segment posteriorly on the same side, exits on the opposite side, passes back over the fundus to enter and exit above and below the hysterotomy. The uterine fundus is manually compressed while the suture is pulled tight and tied.

Hysterectomy

As mentioned above, if bleeding continues despite more conservative options, time should not be lost before proceeding to hysterectomy. For inexperienced surgeons, a subtotal hysterectomy is a quicker and safer option.

Deaths due to PPH in South Africa

The most recent report of the Committee for Confidential Enquiries into Maternal deaths in South Africa covered the triennium 2011–2013. Obstetric haemorrhage was the second most common cause of maternal death, 671 or 15.8% of all deaths, slightly less than 688 deaths between 2008 and 2010. This represents a maternal mortality ratio (MMR) of 23.9 deaths per 100,000 live births. Maternal age >35 years (28.8%) was a risk factor for deaths from obstetric haemorrhage. Common underlying factors were prolonged labour (21.6%) and anaemia (33.7%). The major causes of death due to haemorrhage were bleeding associated with caesarean section (increased to 31.9%), abruptio placentae (16%), uterine rupture (15.3%), uterine atony (7.2%) and retained placenta (6.7%). Uterine rupture remained a concerning statistic. A possible or probable avoidable factor was found to be present in 89.4% of deaths due to haemorrhage, and 61.5% were considered probably avoidable. Substandard care included infrequent monitoring of vital signs after caesarean section, failure to use ergometrine to treat uterine atony in eligible cases and uterine rupture from use of excessive doses of misoprostol for induction of labour, especially in women with intrauterine death.

Analysis of the deaths led to the following key recommendations:

• 1.
  

  Train community health workers to educate pregnant women about obstetric haemorrhage and promote transport plans.

  
  
• 2.
  

  District and facility managers to ensure (a) that district hospitals have 24-h functioning theatres with continuous supply of emergency blood and sufficient staff for the anaesthesia, surgery and recovery areas; and (b) adequately staffed labour wards.

  
  
• 3.
  

  Emergency transport to be onsite for transfers from district hospitals.

  
  
• 4.
  

  Clinical managers to ensure implementation of PPH monograph, essential steps in the management of obstetric emergency skill training, PPH drills and early warning monitoring charts. A particular focus on skills to perform CS safely is required.

  
  
• 5.
  

  Facility auditing of severe PPH and display of trends by graphs/charts in labour wards and theatres.

  
  
• 6.
  

  Use of misoprostol for induction of labour to be closely monitored and all healthcare providers to be trained on the correct use so as to prevent uterine rupture. Foley catheter induction to be promoted as a safer method.

  
  
• 7.
  

  Avoid performing CS for abruptio placenta with an intrauterine death and without appropriate indications.

  
  
• 8.
  

  Syntometrine or ergometrine to be used as second-line treatment for uterine atony in preference to misoprostol, unless contraindicated; and to be considered for prophylaxis at CS.

  
  
• 9.
  

  Monitoring after CS, complicated normal vaginal delivery and manual removal of placenta to be improved in terms of frequency of observations and action on abnormal observations; this is to be facilitated with early warning monitoring charts. At-risk women to be monitored in a special care area.

  
  
• 10.
  

  All women with blood loss in excess of 1000 ml need to be immediately assessed by a doctor.