BJOG Editor's Choice


Caesarean section and subsequent abnormal placentation

In the acute clinical setting clinicians often make ‘best interest’ decisions about the information they give to patients when obtaining consent. They balance the details of the information they give to patients with their own evaluation of the need and degree of urgency for the procedure, which is reflected in time used to obtain consent. Anyone who has worked on a busy labour ward for any length of time will appreciate that, in the clinical setting of an emergency caesarean section for fetal distress, obtaining consent is often a rushed and cursory affair, chiefly because of time constraints. Immediate common complications of caesarean section such as wound and urinary tract infection, haemorrhage and damage to internal organs are often discussed, but rare, or uncommon, late complications are seldom mentioned. In this setting clinicians feel, quite rightly, that to frighten an already worried woman by mentioning rare and complex problems that may only be relevant years in the future is inappropriate when, in the obstetrician's judgment, the need for immediate caesarean section is great. However, the obligation to discuss such problems changes in situations where there is more time and the justification for caesarean section is less clear. This is especially so when there is no medical reason for caesarean section, and a woman has chosen to have the operation.

Caesarean section has been estimated to be the most common surgical procedure in women in the developed world and, although the large majority of caesarean sections are performed for clear clinical indications, an increasing proportion of women request elective caesarean sections without such reasons. For example, in the USA the figure more than doubled over an 8 year period, rising from 3% in 1996 to 7% in 2004 (NIH State-of-the-Science Conference Statement on Cesarean Delivery on Maternal Request.378 NIH Consens Sci Statements 2006;23(1):1–29). For women who request a caesarean section without a clear clinical indication, the discussion of even small risks of severe complications assume higher importance. Two such risks are placenta praevia and abnormal placentation (placenta accreta, increta and percreta) in subsequent deliveries, all of which are increased by previous caesarean section. In this issue Kamara et al. on page 879 demonstrate in a retrospective matched case–control study from Australia (65 cases of abnormal placentation and 102 controls from a total of 82 667 births during a study period between 1993 and 2008) that women who undergo a caesarean section without labour are three times more likely to develop an abnormally implanted placenta in a subsequent pregnancy complicated by placenta praevia compared with women whose primary caesarean section occurred after the onset of labour. Furthermore, this increased risk appeared to be independent of gravidity, parity or vaginal delivery before the first caesarean section. The risk remained unchanged regardless of a history of uterine surgery before the primary caesarean section and regardless of subsequent successful labour after the first caesarean section. Nearly half of the abnormal placentations were increta or percreta (Figure 1).

Figure 1.

Distribution of abnormal placentation among cases.

Temporal trends in postpartum haemorrhage

Medical students on labour wards are often amused, sometimes astonished, when they learn that the majority of obstetricians and midwives routinely use euphemisms for blood loss when they are recording estimated blood loss (EBL) after delivery. An experienced midwife will routinely write ‘EBL 300 ml’ when the blood loss has been what they consider average for a delivery, and an obstetrician will record ‘EBL 450 ml’ for a caesarean section when he or she considers the blood lost has been average. It is well established that when blood loss is accurately measured by techniques such as swab weighing or dilutional techniques that the true average blood loss is around 400–500 ml for a normal delivery and around 600–1000 ml for a caesarean section. Midwives and obstetricians are understandably reluctant to use these correct figures for ‘average blood loss’ because they will be seen to be outliers in subsequent audit, and their surgical or delivery skill may be called into question. Perhaps a third of the midwives’ deliveries would then be registered as a postpartum haemorrhage (PPH) and more than half of the caesarean sections for the obstetrician. When blood loss is clearly greater than average, more attempts are made to accurately measure this, although even in these circumstances clinicians’ estimates are usually underestimates of the true loss. The practice of using euphemisms and systematically recording underestimates for EBL when the blood loss is considered to be low by the clinician, may be changing. On page 853 Mehrabadi et al. from Canada report the results of a population-based retrospective cohort study of factors associated with atonic and severe atonic PPH using data from 371 193 women obtained from the British Columbia Perinatal Data Registry for each year during the period 2001–09. PPH was defined as an EBL of ≥ 500 ml after vaginal delivery or > 1000 ml after caesarean delivery. Severe atonic PPH was defined as atonic PPH in conjunction with blood transfusion or as atonic PPH plus any of the following: transfusion of three more units of blood or packed red cells, emergency hysterectomy, uterine (and vaginal) packing, ligation of pelvic vessels, embolisation of pelvic vessels. The second definition was included to account for potential changes in obstetric practice that may have reduced the threshold for blood transfusion. The factors they examined that may influence PPH included maternal characteristics (e.g. age, parity and body mass index) and obstetric interventions (e.g. labour induction, augmentation and caesarean section). The atonic PPH increased from 4.8% in 2001 to 6.3% in 2009; atonic PPH with blood transfusion increased from 16.6 in 2001 to 25.5 per 10 000 deliveries in 2009; and atonic PPH with blood transfusion ≥3 units or procedures to control bleeding increased from 11.9 to 17.6 per 10 000 deliveries. Changes in rates of atonic PPH and severe atonic PPH were not accounted for by obvious changes in obstetric practice or maternal characteristics. It seems quite possible that the change in atonic PPH may be the result of subtle changes over time in the way that EBL was being reported, but the increase in major PPH is more likely to be genuine. Whether the increase in major PPH was the result of changes in oxytocin used for labour induction or augmentation, magnesium sulphate use, management of the third stage of labour or even the use of selective serotonin reuptake inhibitors (which can impair platelet aggregation), could not be deduced from this study as these factors were not studied, but clearly these are areas that need examination.

Factors predicting obstetric sphincter injuries

Obstetric anal sphincter injuries (OASIS) are an important cause of gynaecological complications such as perineal pain, dyspareunia, and anal and urinary incontinence. Their incidence appears to be rising. The introduction of classification systems and training in their use may partially explain this rise but some of the important risk factors for OASIS, such as high birthweight and high maternal age in primiparous women are increasing within the general obstetric population (Figure 2).

Figure 2.

Familial risk of obstetric anal sphincter injuries.

On page 831 Baghestan et al. report the results of a population-based cohort study from Norway examining familial factors in OASIS. The study was very large, studying nearly 400 000 mother–daughter pairs, 300 000 mother–son pairs, around 130 000 each of mothers whose sisters later became mothers, fathers whose brothers later became fathers and mothers whose brothers later became fathers, in addition to around 90 000 fathers whose sisters later became mothers. The results were intriguing. Not only did mothers or sisters who had OASIS increase the risk in the index pregnancy two-fold, but the risk appeared to be increased by the influence of the brother. If OASIS occurred in one brother's partner at delivery, the risk of OASIS in another brother's partner was also (modestly) increased. If OASIS occurred in one sister at delivery, the risk of OASIS in her brother's partner was also a little increased. It therefore appears that there is a familial component to OASIS that is strongest in the maternal line but which is still weakly present in the paternal line. Some of the factors that predispose to OASIS such as maternal smoking, delivery position, body mass index and duration of second stage may be influenced by genetic or behavioural factors within families. For example smoking habits appears to be highly hereditable (Vink et al. Behav Genet 2005;35:397–406). Data concerning these factors were not examined in this study. Adjustment for other familial influences such as birthweight and head circumference did not have any effect on the findings. There are a now known to be more than 200 inherited connective tissue disorders, ranging from Ehlers–Danlos syndrome to the relatively benign and more subtle forms of joint hypermobility syndrome. The current study raises the intriguing possibility that there may be inherited variations in quality of collagen and connective tissue. However, the findings in this study must be interpreted cautiously as bias due to unmeasured confounding factors may have affected the findings.