Placental development begins upon implantation of the blastocyst into the maternal endometrium during the initial stages of human embryogenesis. Implantation involves localisation to the most optimal position; most commonly the mid-to upper-anterior or posterior uterine wall. This physiological process can be defective and in approximately 6.3 per 1000 pregnancies (Health and Social Care Information Centre, 2007) the pathophysiological condition known as placenta praevia occurs.
Placenta praevia is characterised by either whole or partial implantation of the placenta in the lower uterine segment (Figure 1). Pregnancy complicated by placenta praevia or placenta accreta presents numerous and varied challenges for both the woman and her maternity care team due to the associated levels of maternal and fetal morbidity and significant demand on health resources (Royal College of Obstetricians and Gynaecologists (RCOG), 2011). When insertion of the placenta pervades the deciduas basilis and through the myometrium, a morbidly adherent placenta (placenta accreta) is indicated (RCOG, 2011). A continuous increase in the incidence of placenta praevia, including placenta accreta, is anticipated due to rising caesarean rates. It is suggested that the damage caused to the myometrium and endometrium due to surgical disruption of the uterine cavity during caesarean section is linked with an increased risk of placenta praevia in a subsequent pregnancy (Faiz and Ananth, 2003). More recent studies regarding risk factors for placenta praevia support this and additionally link placenta praevia to increased maternal age, multiple gestation, high parity and smoking (Gurol-Urganci et al, 2011).
Types of placenta praevia
Classification of placenta praevia pertains to the degree to which the leading edge of the placenta encroaches on the cervix and is categorised as either major placenta praevia or minor/partial placenta praevia. Placenta praevia is classified as major when the placenta lies over the internal cervical os and minor or partial when it is in the lower uterine segment (Figure 1) (RCOG, 2011).
Diagnosis
Historically, a diagnosis of placenta praevia was based on the clinical suspicion raised when a woman presented with vaginal bleeding post 20 weeks gestation (particularly unprovoked or painless), and a high presenting part or an abnormal fetal lie. While clinical assessment is essential and suspicion of placenta praevia must always be raised if clinical signs and symptoms are present (regardless of previous imaging results); the definitive diagnosis of placenta praevia is most accurately achieved via ultrasound scan. Current UK practice includes placental localisation at the routine 20-week anomaly scan. Although this practice is not specifically recommended by the UK National Screening Committee (UKNSC, 2013) due to lack of high-level evidence, it is supported by UKNSC, the National Institute for Health and Care Excellence (NICE, 2014) and RCOG (2011).
Diagnosis of placenta praevia or low lying placenta at or prior to 20-weeks gestation needs accurate confirmation later in pregnancy as it is estimated that 90% of placenta praevia diagnosed at this gestational age migrate (RCOG, 2011).
The advent of transvaginal ultrasonography has significantly influenced more exact diagnosis of placenta praevia with increased precision in the measurement of the relationship between the placental edge and the internal cervical os and has therefore been recognised as a more favourable diagnostic tool than transabdominal ultrasonography. The high level of accuracy with transvaginal ultrasonography is demonstrable (sensitivity 87.5%, specificity 98.8%, positive predictive value 98.8% and negative predictive value 97.6%) (Leerentveld et al, 1990); whereas, transabdominal ultrasonography incurs up to a 25% false positive rate (McClure and Dorman, 1990). This rate of false positivity is associated with the limitations of transabdominal imaging in obese women (Trimor-Tritsch and Rottem, 1987) posterior placental position (Edlestone, 1977), presenting part interfering with lower segment visualisation (King, 1973) and underfilling or overfilling of the maternal bladder (Townsend et al, 1986; Lauria et al, 1996). Up to 60% of placenta praevia diagnosed via transabdominal ultrasound will be reclassified by transvaginal imaging in the second trimester (Farine et al, 1988; Smith et al, 1997). Third trimester transvaginal ultrasonography has also demonstrated disparate placenta praevia diagnoses (Oyelese et al, 2004). Notably, transperineal or translabial ultrasound (using a transabdominal probe) is associated with higher levels of accuracy than transabdomnial scanning (Dawson et al, 1996). Although it is not widely advocated and transvaginal ultrasonography is considered the gold standard approach, it can be beneficial in the absence of transvaginal ultrasound resources. The safety of transvaginal ultrasonography in the presence of placenta praevia (even in cases of current vaginal bleeding) has been confirmed by prospective observational study and one small randomised controlled trial (Leerentveld et al, 1990; Timor-Tritsch and Yunis, 1993).
All women with placenta praevia identified at the 20-week anomaly scan or who present with clinical symptoms indicative of placenta praevia require secondary, and if necessary, subsequent imaging. Physiological development of the uterus, essentially the differential growth of the lower segment during the second and third trimesters of pregnancy leads to seemingly ‘apparent’ migration of the placenta (Becker et al, 2001; Dashe et al, 2002; Mustafa et al, 2002). The variable degrees of migration are considered in the literature to be correlated with the relationship between the leading placental edge and the internal cervical os and the stage of pregnancy gestation that diagnosis of placenta praevia occurs. Evidence suggests that in placenta praevia reaching or overlapping the internal os at 18–23 weeks on transvaginal ultrasound, only 9% will continue to remain placenta praevia at birth (Taipale et al, 1997). Early diagnosis of placenta praevia at 11–14 weeks gestation is associated with a lower incidence of 6.2% placenta praevia at birth (Hill et al, 1995). Two studies have shown that overlap of 10 mm or less at 9–16 weeks is extremely unlikely to be associated with placenta praevia at term and the rate of placenta praevia at term is essentially zero if the placenta reaches the os but does not overlap it at early or routine anomaly ultrasound scan (Taipale et al, 1997; Rosati and Guariglia, 2000).
Despite placental migration occurring most commonly in the second trimester, and the relationship between migration and degree of placental overlap at this stage of pregnancy being integral to predicted incidence of placenta praevia at birth, there continues to be value placed on third trimester ultrasonography. One study identified only an 11.5% incidence of placenta praevia at delivery when the placenta lay between 20 mm from the os and 20 mm overlap at 29 weeks gestation (Oppenheimer et al, 2001). While an overlap of more than 20 mm after 26 weeks gestation is highly predictive of the need for caesarean section (Oppenheimer et al 2001), the role of serial scanning can be considered as a tool for determining average migration rates. Migration rates greater than 1 mm per week are predictive of normal outcome (Oppenheimer, 2007). It is recognised that migration is still possible if the placenta overlaps the internal os by 23 mm at 11–14 weeks gestation (Mustafa et al, 2002), by more than 25 mm at 20–23 weeks gestation (Becker et al, 2001) and greater than 20 mm at 26 weeks gestation (Oppenheimer et al, 2001). Placental migration is increasingly demonstrated in cases of anterior placenta praevia and in the absence of a previous caesarean section (Ghourab and Al-Jabari, 2000; Oyelese et al, 2004).
The recommended timing of subsequent diagnostic ultrasonography for placenta praevia is influenced by the degree of placenta praevia primarily diagnosed, obstetric history of the woman and presence or absence of clinical symptoms. It is nationally recommended that women who remain asymptomatic with suspected minor placenta praevia are followed up at 36 weeks gestation. Whereas women with suspected major placenta praevia or questionable placenta accreta are reviewed at 32 weeks gestation in order to plan appropriate third trimester management. Symptomatic women should experience further imaging in accordance with their individual needs (RCOG, 2011).
It is recognised that women with a history of caesarean section are at decreased likelihood of placental migration. In a retrospective study of 714 women with placenta praevia at 20–23 weeks gestation and a history of previous caesarean section, 50% of women required caesarean section due to persistent placenta praevia. This finding was inclusive of women with only partial placenta praevia at the time of initial diagnosis (Dashe et al, 2002). The incidence of placenta praevia is also reported to escalate with the number of previous caesarean sections (Gurol-Urganci et al, 2011). The causal relationship between previous caesarean section and placenta praevia is not fully understood but is considered to be secondary to the reduced differential growth of the lower segment in a scarred uterus and consequential lack of ascension of the placental position during pregnancy (Laughon et al, 2005) (Table 1).
| Number of previous caesarean sections | Number of women | Number of women with placenta accreta | Chance of placenta accreta if placenta praevia | Number of hysterectomies |
|---|---|---|---|---|
| 0 | 6201 | 15 (0.24%) | 3% | 40 (0.65%) |
| 1 | 15808 | 49 (0.31%) | 11% | 67 (0.42%) |
| 2 | 6324 | 36 (0.57%) | 40% | 57 (0.9%) |
| 3 | 1452 | 31 (2.13%) | 61% | 35 (2.4%) |
| 4 | 258 | 6 (2.33%) | 67% | 9 (3.49%) |
| 5 | 89 | 6 (6.74%) | 67% | 8 (8.99%) |
In addition to lower rates of placental migration, there must also be a heightened awareness of the additional risks for women with a history of previous caesarean section and presence of an anterior placenta or placenta praevia as they require exclusion or diagnosis of placenta accreta. Overall incidence of placenta accreta is 1: 2500 births (Miller et al, 1997) with a relative risk of 1: 2605 in the presence of placenta praevia. However, it is more commonly associated with one or more previous caesarean sections. The risk of placenta accreta in the presence of placenta praevia and one previous caesarean section is 25% and in the presence of two previous caesarean sections the incidence is further increased to 40% (Silver et al, 2006).
Differing imaging techniques can be valuable, including the use of greyscale/colour and/or 3D power Doppler sonography and magnetic resonance imaging (MRI) (Chou et al 2001; Wong et al, 2008). While the use of MRI continues to be debated in the empirical literature, the benefits of greyscale/colour 3D power Doppler ultrasound modalities in achieving the most useful results is evident (Shih et al, 2009). Fundamentally, the maternal morbidity associated with placenta praevia and previous caesarean section is high and the incidence of caesarean hysterectomy significantly increases with the number of previous caesarean sections.
Antenatal management
The most appropriate place for care in the third trimester of a pregnancy complicated by placenta praevia continues to be in the hospital environment. There is a lack of high-level evidence to guide clinicians regarding the benefits and safety of outpatient vs inpatient management. The most up-to-date Cochrane systematic review of evidence was in 2003 and there have been none subsequently (Neilson, 2003). The available evidence solely pertains to one randomised trial involving 53 women (Wing et al, 1996). There were no significant differences in clinical outcome, despite a 62% recurrence rate of vaginal bleeding. The reduction in bed occupancy was evident and would have significant benefits in terms of length-of-stay targets and effect on patient experience. However, the limitations of the sample size and the resultant lack of statistical significance limits the widespread extrapolation of the findings as it is not predictive of the overall effect on maternal and neonatal safety. A pragmatic approach to hospital vs home antenatal management may be possible in view of the retrospective examinations of Wing et al's (1996) research, which support outpatient management in stable patients and one subsequent retrospective observational study of 161 women with placenta praevia in the third trimester (Love et al, 2004). Currently, national and international guidance predominantly recommends that women with placenta praevia should remain in hospital during the third trimester of pregnancy (Royal Australian and New Zealand College of Obstetricians, 2003; RCOG, 2011).
The role of cervical cerclage and cervical length assessments have been reviewed independently of each other as potential clinical interventions in the antenatal management of women with placenta praevia. The Cochrane systematic review (Neilson, 2003) details two studies examining cervical cerclage and its impact on reducing antepartum bleeding and length of pregnancy. The finding that there was a small reduction in births prior to 34 weeks gestation and low birthweight babies (under 2000g) was diminished by the randomisation element of the research design and thereby the practice of cervical cerclage in placenta praevia is not recommended. The singular study examining the predictive value of cervical length and risk of preterm birth was affected by small sample size and conclusions cannot be effectively evaluated (Arias, 1988).
Labour and birth
The recommended mode of birth is determined by both the distance from the placental edge to the internal cervical os and presence of associated clinical signs and symptoms of placenta praevia in pregnancy. Based on ultrasound findings at 35–36 weeks and the presence of vaginal bleeding and/or unstable fetal lie, the opportunity for vaginal birth or the need for caesarean section will usually be planned. Findings from five studies reviewing likelihood of caesarean section in placenta praevia confirmed that vaginal birth is most commonly associated with a distance of greater than 20 mm between the leading placental edge and the cervical internal os. Vaginal birth rates varied between 63–100% in these situations, whereas caesarean section rates vary from 40–90% when the placenta lies between 0mm and 20 mm from the internal os (Oppenheimer, 2007). Knowledge of the exact distance is fundamental to clinical decision making around mode of birth. It has been suggested that a placenta >20 mm from the os should be classified as a low lying placenta rather than placenta praevia thereby reducing medical bias towards surgical birth. In cases of placenta praevia (20 mm–0 mm from the os) the value of precise measurements in asymptomatic woman can be beneficial in determining suitability for trial of labour in some women. It can be appropriate in this group of women, in the presence of a stable lie and absence of vaginal bleeding (Sallout and Oppenheimer, 2002). However, there is a paucity of evidence regarding this and prospective research is needed. In cases of placenta praevia that overlaps the os, caesarean section is necessary in all cases (Oppenheimer et al, 2001; Bhide et al, 2003). While planning the mode of birth is essentially predisposed by the degree of placenta praevia, the woman's preferences in relation to the clinical picture must be considered and inform the clinical shared decision making process.
The timing of birth is determined by the degree of placenta praevia (including placenta accreta) and presence of associated symptoms. There is an increased likelihood of preterm birth (40% of births before 38 weeks gestation) and these are unpredictable in the majority of cases (Love and Wallace, 1996; Zlatnik et al, 2007). This is unavoidable as elective birth before 34 weeks gestation is not ordinarily desirable due to the increased neonatal morbidity. However, careful consideration of each case is essential to avoid neonatal mortality associated with prolonging pregnancies complicated by placenta praevia (Gielchinsky et al, 2002). In the most high-risk scenarios such as placenta accreta, planned caesarean section with corticosteroid cover is considered reasonable at 36–37 weeks (RCOG, 2010) and in uncomplicated placenta praevia, birth can usually be planned for at 38–39 completed weeks gestation (Paterson-Brown and Singh, 2010).
Clinical opinion is divided regarding anaesthesia for planned caesarean sections (Schachter et al, 2002). There has been no new research in recent years. Two retrospective trials evidenced positive clinical outcomes for women in terms of haemodynamic state when receiving regional anaesthesia compared to general anaesthesia (Frederiksen et al, 1999; Parekh et al, 2000). However, prolonged surgery, such as that anticipated in cases of placenta accreta benefits from the use of general anaesthesia (Parekh et al, 2000). In view of the lack of consensus in the literature, current national guidance places the decision for method of anaesthesia with the anaesthetist conducting the procedure as part of the multidisciplinary pre-operative planning.
Intrapartum and postnatal complications
Placenta praevia is directly associated with the risk of massive obstetric haemorrhage and the incidence of caesarean hysterectomy is cited as 5.3% (relative risk compared with caesarean sections performed in the absence of placenta praevia is 33) (Oppenheimer, 2007). Planned caesarean hysterectomy may be inevitable in some clinical cases; however, continual risk assessment and planning in relation to antenatal and intraoperative findings is essential as alternative management may be considered.
It is therefore fundamental that preoperative planning involves adequate consent advice for the woman and a multiprofessional team approach due to the high morbidity related to the condition. In cases of previous caesarean section with an anterior placenta, confirmed placenta accreta or placenta praevia with a history of caesarean section; the suspected placenta accreta care bundle should be adopted. This collaborative care model was developed by the National Patient Safety Agency (NPSA), Royal College of Midwives (RCM) and RCOG in response to the confidential enquiry (Confidential Enquiry into Maternal and Child Health, 2007). The six underpinning principles of good care highlighted in the care bundle include consultant obstetric and anaesthetic involvement in planning and delivery, availability of blood products, level 2 critical care bed availability, consideration and consent for potential interventions such as interventional radiology, cell salvage and hysterectomy and multidisciplinary pre-operative planning. Practicality and achievability of the care bundle has been highlighted in pilot studies and its appropriateness to the planning and provision of care in cases of such complexity is theoretically evident.
Conclusion
If a woman is diagnosed with placenta praevia at the routine 20-week anomaly scan, it is important that she is reassured that in the majority of cases placental migration occurs and pregnancy will not be complicated by this condition at the time of delivery. Subsequent ultrasound imaging is essential to either confirming placental migration or diagnosing the degree of placenta praevia. Women with confirmed placenta praevia should be sensitively counselled regarding the increased morbidity associated with the condition, recommended inpatient antenatal care in the third trimester and likelihood of vaginal birth versus caesarean section. Clear communication is integral to positive maternal experience and clinical outcomes.