Antepartum haemorrhage (APH) refers to bleeding from the genital tract after 20 weeks of gestation, which is 4−6 weeks below the lower limit of fetal viability. Establishing the cause of APH is important to distinguish scenarios at risk of substantial haemorrhage, such as major placenta praevia or abruption, from a range of possibilities that pose much lower risks ( Table 19-1 ). Cases with serious underlying causes of vaginal bleeding, such as caesarean section scar pregnancy, may present with vaginal bleeding before 20 weeks. Furthermore, the identification of a benign lower genital tract source of minor vaginal bleeding does not preclude an additional more serious uterine cause of APH. The two most serious causes are placental abruption and placenta praevia; the latter is becoming more common, due in part to a greater prevalence of: previous uterine surgery, assisted reproductive technologies, multifetal pregnancy and advanced maternal age. Major abruption is much less common due to general advances in maternal health, including a large reduction in smoking, and improvements in antenatal care. Historically, the first description of placenta praevia in 1885 was by the Parisian physician, Paul Portal (1630–1703), who was the first to describe the attachment of the placenta to the lower uterine segment.
|Placenta praevia increta|
|Antepartum fetal death|
|Cancer of the cervix|
|Lower genital tract||Vulvo-vaginal varices|
‘I put my fingers into the orifice and felt the after birth which covered the orifice of the matrix from all sides and adhered in all its parts with the exception of the middle’.
La Pratique des Accouchements Soutenue d’un Grand Nombre d’Observations. Paris: G. Martin, 1685 .
The elegant, yet poignant, drawings of this disease from partially dissected dead women by the Scot, William Hunter (1718−1783), living in London, are a vivid reminder of the danger of placenta praevia, which still exists in many countries today. The terms ‘unavoidable ’ referring to placenta praevia and ‘accidental ’ referring to placental abruption are attributed from 1775 to Edward Rigby (1747–1821) of Norwich, England. The historical background of placenta praevia has been documented.
Placenta praevia occurs when the entire placenta, or in part, implants in the lower uterine segment after 20 weeks’ gestation. The incidence varies by population but significant disease occurs in about 1/200 deliveries. Perinatal mortality is increased almost twofold compared to non-praevia pregnancies adjusted for smoking, maternal age, parity and in vitro fertilization. The risk factors for placenta praevia are summarized in Table 19-2 . Some of these, in particular multiple previous caesarean deliveries and previous placenta praevia, confer recurrence risks of up to 5%. Finally, the increasing use of 18−20-week ultrasound examinations to assess fetal anatomy has increased the rate of diagnosis of asymptomatic minor degrees of placenta praevia. An abnormally large placenta surface area predisposes to placenta praevia, the most common cause being multifetal pregnancy; amongst the rare causes, careful consideration of a succenturiate lobe in the lower segment is important, since it is associated with vasa praevia, which, when undiagnosed, may result in fetal mortality. Twin pregnancies have a 50% greater risk of placenta praevia.
|Advanced maternal age|
|Obstetric history||High parity, twins|
|Assisted reproductive technologies|
|Previous placenta praevia|
|Chronic fetal anaemia|
|Abnormal placental development|
|Succenturiate lobe (vasa praevia)|
The classical method describes four types, or degrees, of placenta praevia as illustrated in Figure 19-1 . With additional descriptive terminology, these are as follows:
Type 1 (low-lying): the lower edge of the placenta is inside the lower uterine segment but does not reach the internal cervical os.
Type 2 (marginal): the lower edge of the placenta extends to but not across the internal os.
Types 1 and 2 are commonly observed in asymptomatic women at the 18−20-week transabdominal fetal anatomy ultrasound examination. The distinction between the two, by transvaginal ultrasound, is unimportant at this stage, since in both instances, the likelihood of clinically significant placenta praevia in the third trimester is very small.
Type 3 (partial): the lower edge of the placenta extends asymmetrically across the internal os; however, since the portion of the placenta covering the internal os is thin, it may pull away with minimal vaginal bleeding during cervical effacement and dilation to permit safe vaginal delivery.
Type 4 (complete or central): the placenta is almost centrally placed within the lower uterine segment.
With few exceptions the distinction between type 3 and 4 placenta praevia is not important in high-resource countries, because the risk/benefit ratio of planned caesarean delivery outweighs that of attempting vaginal delivery. Exceptions would include: previous vaginal deliveries in a highly motivated and well-informed individual; and anticipation of a difficult caesarean, for example, due to morbid obesity. In other healthcare settings, especially those in which accurate transvaginal ultrasound imaging is not readily available to distinguish these types, the use of the term ‘type 3’ to describe transabdominal ultrasound findings implies that a proportion of women in labour with a small amount of placenta across the os can deliver safely by the vaginal route because during cervical effacement and dilation the small area of disrupted placenta may not bleed significantly.
The wide application of high resolution transabdominal and transvaginal ultrasound has largely obviated the need to describe four categories of placenta praevia, such that the disease is now commonly described as minor praevia (types 1 and 2) and major praevia (types 3 and 4) that respectively do not or do require elective caesarean delivery.
Physiology of the Lower Uterine Segment
In the non-pregnant state the uterus is comprised of just two parts, a corpus and cervix, the boundary between which is a fibro-muscular junction described originally by Danforth. The lower uterine segment begins to form in the second trimester, once the gestation sac has fully occupied the uterine cavity; thereafter maternal tissue below the apex of the fetal membranes is considered the cervix and the junction is the internal os ( Fig 19-2a ). The lower uterine segment gradually forms, via myometrial growth and thinning, from the tissue above and below the internal os. As such, the cervix gradually shortens as pregnancy advances while failure to do so increases the risk of caesarean section for dystocia. Formation of the lower uterine segment provides one-third of uterine volume for fetal growth and is normally occupied by the fetal head from 34 weeks’ gestation. Sonographically, the upper margin of the lower uterine segment is the reflection of the utero-vesical peritoneum at the upper edge of the semi-filled bladder ( Fig 19-2a ). In labour, the upper active uterine segment (the fundus) provides the driving force for labour, placental detachment and subsequent mechanical haemostasis. By contrast, the lower uterine segment is a passive structure in normal labour; at caesarean section for obstructed labour it may balloon out significantly and be a source of primary postpartum haemorrhage. In the labour and delivery setting, the lower uterine segment is defined pragmatically as that 6−8 cm portion of the uterine cavity palpable digitally in women with either regional or general anaesthesia following delivery of the placenta.
Gradual formation of the lower uterine segment is sometimes described as ‘placental migration’. The forces underlying this phenomenon may cause the lower placental edge to bleed, even in women with type 1 or 2 praevia. Nevertheless, this migration means that at least 90% of low-lying placentas will resolve, leaving about 1 in 200 women with clinically significant placenta praevia after 34 weeks. The almost universal use of ultrasound at 19−20 weeks has the capacity to over-diagnose minor degrees of placenta praevia in asymptomatic women ( Fig 19-2a ). Advice on safe mode of delivery in this context can all too easily err on the side of caution and caesarean delivery; yet many options, described in Chapter 20 , exist to manage postpartum haemorrhage effectively. The previous generally accepted standard was to recommend caesarean where the lower placental edge is < 2 cm from the internal os in the third trimester. However, the lower uterine segment continues to form, especially with cervical shortening and effacement from Braxton Hicks contractions. More recent evidence suggests that the majority of women with no other risk factors can achieve safe vaginal delivery if the lower placental edge is > 1 cm from the internal os. A flexible policy of serial transvaginal ultrasound in borderline situations can save some women from unnecessary caesarean delivery ( Fig 19-2b,c ).
Assessment of Vaginal Bleeding from Placenta Praevia
Around 80% of all women with major placenta praevia will have one or more bleeds before delivery. The first is a warning, or ‘sentinel’ bleed. Any subsequent bleeds are likely to be heavier. In general, major degrees of placenta praevia bleed earlier, more frequently and more heavily than do minor degrees. Nevertheless, even a complete placenta praevia may not bleed until either the onset of labour (if undiagnosed) or simply presentation with an oblique or transverse lie in the clinic after 34 weeks. Bleeding from placenta praevia is most commonly caused by disruption of small uteroplacental veins as the anchored placenta is gradually stretched. This maternal source of blood may escape through the decidua and enter the myometrium. Thrombin, the local product of pathological haemorrhage, is a powerful myometrial irritant and may explain why uterine contractions can accompany bleeding from placenta praevia. Maternal tocolysis may arrest ongoing APH and extend the duration of pregnancy.
Transvaginal ultrasound may be useful to predict preterm birth in placenta praevia; at 32−33 weeks, a cervical length < 30 mm conferred a threefold risk of subsequent delivery for haemorrhage and preterm birth. A prominent marginal sinus at the lower edge also predicts the need for caesarean delivery due to vaginal bleeding. Transvaginal ultrasound therefore provides useful predictors of recurrent bleeding and thus the need to remain in hospital if undelivered following the sentinel bleed. Unexpectedly severe vaginal bleeding from a known minor placenta praevia may be due to a prominent marginal placental sinus where uterine venous blood drains out of the lower edge of the placenta. Transvaginal ultrasound is a valuable assessment tool in stable women following a sentinel bleed. Where vaginal bleeding with placenta praevia is seen, the rare accompaniment of gross haematuria should immediately raise the suspicion of associated invasive placentation (see below and Figures 19-2d and 19-9 ).
The importance of careful evaluation of women with a ‘warning bleed’ from placenta praevia was dramatically illustrated by Munro Kerr in an earlier edition:
‘I arrived one morning on my wards to learn that a patient with placenta praevia had died. She had had one or two slight bleeds to which the family physician had not attached much importance; then a severe one occurred, and he sent her into hospital. On her admission the house surgeon examined her vaginally; a most profuse bleeding occurred which neither he nor the more senior resident could control. Before a senior member of the staff arrived by taxi-cab the patient was moribund and could not be rescued. Here the family doctor was to blame for not sending the patient in after the first, but still more to blame was the house surgeon for having examined the patient (vaginally); a senior member of staff should have been summoned immediately’.
Today, complicated obstetrics is more often characterized by multidisciplinary care in larger centres where senior staff direct the care of complex patients in a context of structured post-graduate medical education. I think Munro Kerr would approve of obstetric safety teaching sessions and be a strong exponent of unit-specific massive haemorrhage guidelines, instrument organization and participation in the creation of national guidelines. Paradoxically, the risk of massive APH from a previously undiagnosed complete placenta praevia is rare in settings with wide availability of obstetrical ultrasound. By contrast, a more common cause now of unanticipated major APH is placental abruption (see below). The main clinical features differentiating placenta praevia from abruption are summarized in Table 19-3 . Note that cases of mild abruption may present in a similar fashion to placenta praevia, while in some cases of placenta praevia extravasation of blood into the myometrium may cause uterine irritability.
|Placenta Praevia||Abruptio Placentae|
|History of ‘warning’ bleed(s)||Less likely to be preceded by ‘warning bleed’|
|Apparently causeless||May be associated with hypertensive disorders, trauma, etc|
|Shock and anaemia correspond to apparent blood loss||Shock and anaemia may be out of proportion to apparent blood loss|
|Uterus has normal tone and is not tender||Increased uterine tone and tenderness|
|Malpresentation and/or high presenting part||Normal presenting part|
|Normal fetal heart rate and fetal assessment||More likely absent or abnormal fetal heart rate and fetal growth restriction|
Asymptomatic placenta praevia : Despite the risk of APH, the most common outcome for placenta praevia is to remain asymptomatic until planned admission for caesarean section. As antenatal care evolves, the following clinical factors are important:
Ensure maintenance of haemoglobin > 100g/L, treat iron deficiency − if necessary with IV iron.
Consent for blood transfusion, and specialized (see below) management of the small subset of women who are Jehovah’s Witnesses.
Careful assessment for invasive placentation (see below) in women with major placenta praevia and previous uterine surgery.
Refine gestational age for caesarean delivery based on several factors, including abnormal lie, recurrent APH, co-morbidities (e.g. hypertension), fetal wellbeing, a prior history of preterm delivery and cervical length.
Consent for surgery in advance, focusing on the wisdom of a midline skin incision and classical caesarean section, especially with persistent transverse lie, large fibroids, previous laparotomies and morbid obesity.
Ensure that women are counselled about the option of tubal ligation.
Admission with major APH : Should this occur, women should be advised to call 911 (or equivalent) and rely on medical services, rather than private transport, in order that they are stabilized (if necessary) and transported to the nearest general hospital with obstetrical services. Ideally, they will arrive with intravenous access. The initial triage assessment should involve ABC (airway/breathing/circulation), including vital signs, oxygen saturation on air, abdominal examination and a non-stress test. Admission as an emergency > 26 weeks with active APH is an indication to perform this initial assessment in the operating room, with co-assessment by the anaesthesia team. No pelvic examination should be undertaken; rather, the patient’s medical records should be reviewed and an emergency transabdominal ultrasound should be performed with portable equipment. Ultrasound will quickly determine fetal viability (fetal death or bradycardia is rare unless the diagnosis is abruption), fetal lie and presentation, in order to verify the appropriate skin and likely uterine incision, should the extent of APH dictate immediate caesarean section under general anaesthesia. In parallel with this obstetric assessment, anaesthesia, supported by nursing and anaesthesia assistants, can establish two large-bore IV access lines to obtain blood samples, start resuscitation measures including infusion of crystalloids +/− colloids and prepare the woman for general anaesthesia (airway assessment, pre-oxygenation) as needed. Initial blood tests are a complete blood count, coagulation screen, baseline electrolytes, blood type, antibody screen and cross-match at least two units of packed red cells. Where APH is substantial and immediate general anaesthesia is agreed on, the local policy to alert all systems for massive blood transfusion (often termed ‘code Omega’) should be activated. Surgical and post-operative considerations for women delivered by caesarean section for placenta praevia and major APH are described below. Despite the acute setting, it is important that rhesus-D negative women receive Anti-D.
Admission with minor APH : Typically in this situation the vaginal bleeding settles during triage assessment of a haemodynamically stable woman. In the absence of any uterine contractility or concerns with the non-stress test, such women can be admitted to the antenatal unit. Where no formal diagnosis of placenta praevia has been made, elective high-quality transabdominal and transvaginal ultrasound should be arranged to establish or refute this diagnosis. Women transferred to the labour and delivery area with intermediate-level APH and/or uterine contractions < 32 weeks are at greater risk of delivery in the subsequent 48 hours, and therefore co-care with anaesthesia is important. Some women merit blood transfusion, to stay ahead of blood loss. Tocolysis may be useful in this setting. All women admitted in the window 24−32 weeks with APH should be given an intramuscular course of antenatal steroids to promote fetal lung maturation. For viable deliveries < 32 weeks women should also be started on a 12-hour IV regimen of magnesium sulphate for fetal neuro-protection. These two evidence-based interventions make sense in high-resource settings but may be understandably omitted where resources must be focused on survival of the mother and term new-borns.
Reassessment of care : Depending on organizational structure, women under the care of a midwife or family physician should have their care transferred to an obstetrician. Where a diagnosis of major placenta praevia is made following admission to a small birth unit lacking 24/7 in-house support and/or major support services, consideration should be given to transport to a regional centre when stable.
Subsequent expectant management : If the patient is < 36 weeks’ gestation, has no contractions, the bleeding has settled for 48 hours, the fetus is objectively healthy and there are no maternal co-morbidities that direct the need for delivery (e.g. pre-eclampsia), a period of expectant treatment is reasonable in order to gain time for fetal maturation. Documentation of a thin placental edge > 1 cm away from the internal os is reassuring. In borderline situations with major placenta praevia at 34−35 weeks, particularly with a new or recurrent APH, amniocentesis in the fasted state for fetal lung maturation makes sense, in order that planned caesarean section can take place when a mature lamellar body count is demostrated – rather than exposing women to prolonged hospitalization and the risk of emergency caesarean delivery for a subsequent APH. A positive Betke−Kliehauer (BK) test > 10 ml indicating a significant feto-maternal bleed is rare in placenta praevia and would be an indication for delivery > 32 weeks. Likewise, cervical length < 30 mm on transvaginal ultrasound assessment confers a threefold increased risk of subsequent APH and preterm delivery < 37 weeks, which is presumably due to subclinical labour provoking bleeding. The main elements of expectant management in hospital are as follows:
Admission and bed rest with bathroom privileges
Diagnose and treat anaemia
Continuous cross-match of at least 2 units packed red cells
Identify and arrange a date for elective caesarean and obtain consents – discuss and document type of surgery including desire for tubal ligation
Rhesus-D-negative women with no antibodies should receive Anti-D.
Follow-up of women with minor placenta praevia : Where women remain undelivered for > 2 weeks in the context of minor (type 1 or 2) placenta praevia, repeat consideration should be given to the mode of delivery where vaginal delivery is feasible in the absence of placenta praevia. The modern cut-off for safe vaginal delivery of > 2 cm from the internal os has been challenged by more recent prospective data, showing that two-thirds of women can deliver safely vaginally if the lower edge of the placenta is 1−2 cm from the internal os. Since the cervix gradually shortens by effacement in the late second and third trimesters, a minor praevia in the context of a normal length (> 3 cm) cervix is very likely to pull back behind the fetal head, as a result of cervical effacement ( Fig 19-2b,c ), or during early normal labour. This apparent movement of the lower placental edge takes place even after 36 weeks; therefore even weekly assessments at this stage are worthwhile so as to avoid an unnecessary caesarean section.
Fetal wellbeing assessment and placenta praevia : Despite lower implantation, there is no convincing evidence that this results in any direct association with placental dysfunction. Therefore the initial fetal health assessment in the context of placental praevia and an APH should be as follows: fetal biometry, amniotic fluid, umbilical artery Doppler, biophysical profile and a non-stress test. Additional Doppler studies should be reserved for specific indications: middle cerebral artery Doppler (if the fetus appears growth-restricted), uterine artery Doppler (if the fetus appears growth-restricted or the woman is hypertensive) and anterior lower segment colour Doppler (anterior praevia with previous caesarean section to rule out increta). Since placental function is normal in placenta praevia, fetal growth and tests of wellbeing should follow current advice.
Vaginal Delivery in Placenta Praevia
The widespread availability of good-quality ultrasound means that most women with minor placenta praevia will have had discussions about mode of delivery with their obstetrician in the antenatal clinic setting. As such, the need to clarify the safety of attempting vaginal delivery in an acute setting is rare. Examples today would be either early normal labour with unusually heavy show and fresh bleeding vaginally, or presentation in labour between appointments where no final decision has been made on mode of delivery. In this context use of the ‘double set-up’ examination is valid, since this was the method used to make such decisions in the pre-ultrasound era in haemodynamically stable women. The components of the process are as follows: transfer to the operating room, co-care with anaesthesia, perform complete blood count (CBC) and coagulation screen, cross-match two units of blood, staff present for immediate caesarean section. The major difference today in comparison with former times is that the majority of such procedures are done with a full top-up epidural. Women bleeding heavily would proceed faster through the above steps, with the exception of a general anaesthetic. It is useful to have a portable ultrasound machine at hand, which in acute circumstances may be helpful as follows: (1) the placenta may only be a minor praevia with an engaged fetal head – therefore proceed to vaginal examination as the bleeding may only be due to a rapidly dilating cervix; (2) there is major placenta praevia, but the fetus is a back-up transverse lie, mostly above the umbilicus – therefore use a midline skin entry and be mentally prepared to perform a classical caesarean section (see below).
Double set-up examination : Once the epidural is fully functional, the woman is examined abdominally, to determine lie, presentation and engagement. Any clinical doubts should be resolved using portable ultrasound prior to gowning, sterile preparation and adopting the lithotomy position. The procedure is abandoned in favour of caesarean in women with either transverse lie or (most) situations with breech presentation. Next, the bladder is catheterized with a Foley catheter and bag. A sterile finger is then inserted vaginally to palpate the fornices. This initial step is done to determine if thick placental tissue is present between the lower uterine segment and the fetal head. If the fetal head is easily palpated through a thin lower uterine segment it is then deemed safe to push the examining 1−2 fingers through the cervix, to explore the lower uterine segment for any intervening placental tissue. Blood clot and placental tissue may be difficult to distinguish, though placental tissue is firm and may have a gritty feel. If no placenta is found upon digital exploration of the inside of the lower uterine segment then labour can be safely induced with amniotomy and an oxytocin infusion. If placenta praevia is confirmed, or if there is active bleeding, a caesarean section is performed. If the cervix is long and closed, the examination is inconclusive – which is why portable ultrasound is important, so that the woman can leave the operating room with a clear plan for mode of delivery.
Monitoring in labour : Women attempting vaginal delivery with minor placenta praevia should have one-to-one nursing/midwifery care, in a labour room that is in immediate proximity to the operating room. Written informed consent should be obtained for care, including caesarean if needed as an emergency. Women should have large-bore IV access and two units of blood cross-matched, and be assessed by the anaesthetist and by the most senior on-call obstetrician. Pro-active care, using amniotomy and oxytocin infusion, is preferable since cervical effacement and dilation brings the placenta away from the leading edge of the cervix, while descent of the fetal head may compress the lower placental edge. Active bleeding in early labour is an indication for caesarean section, whereas new bleeding in more advanced labour may be a sign of advanced cervical dilatation and thus the possibility of vaginal delivery.
Postpartum considerations : Following successful vaginal delivery, the woman is at greater risk of primary postpartum haemorrhage because of increased bleeding from the lower uterine segment that is not capable of strong tetanic contractions. Prophylactic measures should be undertaken as described in Chapter 20 .
Vaginal Delivery with Major (Type 3 or 4) Placenta Praevia
In modern obstetric practice, it is occasionally permissible to consider and attempt vaginal delivery when the placenta clearly covers the internal os. The most common situation is with the prenatal diagnosis of a major lethal abnormality (e.g. renal agenesis or skeletal dysplasia) or intrauterine fetal death, typically before 24 weeks of gestation. Assuming there are no other considerations (e.g. previous caesarean), the author’s group have approached such cases using feticide and pre-induction Gelfoam embolization of the anterior divisions of the iliac arteries, followed immediately by a high-dose vaginal misoprostol regimen (600 µg every 4 hours) for induction of labour. Others have approached the challenge in a similar fashion. We would generally not consider this approach after 28 weeks, due to the much greater risk of haemorrhage and the greater likelihood of achieving a lower segment caesarean section.
Abnormal Lie/Malpresentation in Early Labour with Minor Placenta Praevia
Given the potential danger of caesarean section in the pre-blood transfusion/antibiotic era, several techniques were developed to achieve maternal survival via vaginal delivery for the non-vertex fetus. These may remain applicable today when the fetus is dead, pre-viable or has a lethal anomaly and in remote areas with limited or unsafe facilities for caesarean section.
Bipolar podalic version : Rare circumstances exist where fetal manipulation and assisted vaginal delivery may be the safest maternal option in type 1−2 placenta praevia, though at the expense of fetal survival. The Braxton Hicks bipolar podalic version method was developed 150 years ago. The technique demands that the cervix is > 2 cm dilated and the placenta praevia does not cover the internal os – so that the gentle insertion of 1–2 fingers can be used to push up the fetal head between contractions while the external hand manipulates the breech in a downward direction into the pelvis ( Fig 19-3 ). The fingers through the cervix then grasp a foot of the fetus ( Fig 19-4 ) to bring that leg down through the cervix. In this way the breech is used to both dilate the cervix and tamponade the lower placental edge ( Fig 19-5 ). Persistent traction is put on the breech to keep it firmly against the placenta: a bandage can be tied to the fetal ankle and a small weight, for example a bag of saline, is attached to provide sustained traction. For small immature fetuses, sponge forceps can be used to grasp a leg.