II. The unfacts of ‘request’ caesarean section

Authors


Mother: Maternal Mortality

The previous Confidential Enquiry for the UK did not attempt a case fatality rate for caesarean section1,2, temporarily restricting the safety debate. When trying to isolate mortality due to elective caesarean section in previously healthy women, Lilford et al.3 found a 3.8 relative risk (elective vs vaginal delivery 23:6 deaths /100,000). This would equate to a maternal death rate of 1 in 4262 for mothers having elective caesarean section which seems high, particularly compared with unproven claims of avoidable intrapartum fetal death. A more recent study from The Netherlands4 found a similar relative risk of 3.25 (and a death rate of 13/100,000). Extrapolated estimates (Table 1) show that elective caesarean section death rates can be as much as two to eight times higher5, and in the 1994–1996 Confidential Enquiry period would be at least three times higher6, than vaginal delivery. Numerous studies have recorded the higher risk of caesarean section delivery, not all of which can be accounted for by the complications which necessitated the operation7–10. Proponents of caesarean section point out that the relative risk of elective caesarean section might fall further with increasing safety procedures (e.g. use of epidural anaesthesia, antibiotics, and thromboprophylaxis). However, we must use what figures we have, as vaginal delivery also becomes less mortal with the growing use of the same procedures in an increasingly healthy population. There are rare risks in multiple medical interventions, such as allergic anaphylaxis, which may become more significant in the future, and maternal mortality must be watched particularly in countries with the highest and rising caesarean section rates.

Table 1.  Case fatality rate by mode of delivery 1988–1990 and 1994–1996. 1997–1999 Confidential Report1 used a different classification of caesarean section2.
Mode of delivery1988–19901994–19961997–1999
n (000s)Direct deathsRate/million casesn (000s)Direct deathsRate/million casesn (000s)Direct deathsRate/million casesRelative risk (95% CI)
  1. 1National Institute of Clinical Excellence, Scottish Executive Health Department, Department of Health, Social Service and Public Safety:Northern Ireland. Why Mothers Die 1997–1999.

  2. 2Lucas DN, Yentis SM, Kinsella SM, Holdcroff A, May AE, Wee M, Robinson PN. Urgency of caesarean: a new classification. J Roy Soc Med 2000;93:346–350.

Vaginal20823717.818463820.617102916.91.0
Elective CS12819148.3154958.5130538.52.3 (0.88–5.86)
Scheduled CS      78112.80.8 (0.10–5.55)
Urgent CS      13714102.26.0 (3.18–11.40)
Emergency CS15038252.719836182.06914202.912.0 (6.32–22.65)

Even if vaginal delivery is safer than elective caesarean section, the real risk–benefit calculation is between labour (which might end in an emergency caesarean section) and elective caesarean section, and thus becomes critically dependent on the emergency caesarean section rate. The only hard data on this comes from a meta-analysis of all the randomised breech trials which found an increased risk of maternal death or severe early morbidity (RR 1.29, 95% CI 1.03–1.61), and this despite a high emergency caesarean section rate of 45% in the labour arm11. This increase of nearly 30% is likely to be an underestimate for cephalic presentation with its higher vaginal delivery rate. A laissez-faire attitude to elective caesarean section sends a mistaken signal to the public and professionals alike that all caesarean sections are safe and the request debate can be misinterpreted as such. A reduction in the threshold for emergency caesarean section increases the dangers for labouring women12 in this and future pregnancies as 67% of women will have further caesarean sections13. A vicious cycle can result whereby high emergency caesarean section rates fuel further loss of confidence and raise caesarean section rates, making elective caesarean section relatively more attractive. Working towards optimum safety in childbirth must mean increasing the rate of uncomplicated vaginal delivery. As emergency caesarean section in labour is the worst mode of delivery in terms of maternal morbidity, concentrating on the quality of intrapartum care should be our priority. Surely the real obstetric debate today, with the highest intervention rates on record, is how to achieve this.

Morbidity

New, unexpected long term risks of caesarean section continue to be reported such as ectopic pregnancy14, haemorrhage and hysterectomy following uterine evacuation15, latex allergy16, cutaneous endometriosis17, adenomyosis18, increased hospital readmission and even an increase in gallbladder disease and appendicitis19. Straightforward tables of complications could be formulated with inclusions depending on the severity and degree of risk. However these do not rank risks (e.g. the risk of urinary tract infection versus potentially life threatening haemorrhage). Nor are there tools to help us agree how much low risk should be accepted in order to avoid a rarer but potentially serious or lethal complication. While some work has begun to look at decision analysis20 we do not have an ethical framework. Without this preparatory analysis, all counselling is personal and biased, affected by clinical experience. Using a more research-based approach to reviewing complications reveals a voluminous amount of information on the complications of different modes of delivery, but it is variable in quality and coverage and can be biased by enthusiasm or omission of negative trials. While the medical profession debates the risks and benefits for different modes of management, the press and the public hear that the debate is about rights and wrongs and so popular beliefs, myths and dogma are generated.

Morbidity of Mythological Proportions

Problems arising after birth are often attributed to labour and delivery. This scares and undermines women's ability to undergo successfully a normal process. Often this blaming occurs without good supporting evidence or, worse still, when opposing evidence suggests it is the pregnancy or life event changes that are contributory. Doctors must be sympathetic to women injured during vaginal delivery, but should distinguish between morbidity occurring after birth and that caused by it. A good example is the belief that childbirth inevitably damages the pelvic floor, often makes women incontinent of urine, and that caesarean section is protective. Evidence suggests that much pelvic floor weakening is due to pregnancy, most incontinence settles or is treatable and caesarean section is not completely protective. Nevertheless the dogma sticks.

The two common reasons for urinary incontinence in women are genuine stress incontinence and detrusor instability. While pregnancy is likely to be a factor for genuine stress incontinence there is no known cause for detrusor instability in healthy women. Researchers have found that women in whom stress incontinence develops late in life are probably destined to do so from an early age. Heredity and pregnancy, rather than parturition, reveal the defect21,22. The damage that is inflicted appears to be in those predisposed23 (by development of antenatal stress incontinence24,25 or collagen disorders26). Sixty-two percent of primigravidae with incontinence say it started in pregnancy27–29 and other workers find incontinence rarely, if ever, starts after childbirth30,31. Where women are continent, caesarean section is only 20% protective and only for the first caesarean section. Where repeat caesarean sections are performed the protection is lost. By the third there is no benefit with 35% of women suffering the symptom27. The mechanism is thought to be due to vesical denervation32 which also increases the risk of detrusor instability, which is more important when looking at incontinence in old age. In older women detrusor instability (probably due to age changes in both the central nervous system and in the lower urinary tract) is the main cause for incontinence, made worse by immobility and suboptimal nursing. Ten percent of old people have genuine stress incontinence and detrusor instability33.

So, while there might be an argument about diminution of incontinence (still as yet unproven) it would only apply to a minority of women. Future work should address these women and information should be given about possible treatments whichever route delivery takes. Antenatal pelvic floor exercises are helpful in prevention but not widely promulgated27. Other risk factors for incontinence are having more than four children27 and obesity consistent with nonpregnant weights of >120% average weight27,34. If caesarean section were to be proven protective, evidence-based incontinence prevention might be to offer it to the more obese planning large families while obstetricians could reassure slim women planning one child they are low risk.

Another recent concern is that women will become incontinent of faeces. Sultan et al.35 work showed a 0.6% frequency of documented anal sphincter tears and demonstrated a disappointing inadequacy of primary repair. However, there is confusion about symptoms and ultrasound findings. An ultrasound finding in normal controls of 33% anal sphincter disruption36 does not mean a third of women will have faecal incontinence, just as 100% rectus sheath disruption at caesarean section does not mean all suffer abdominal wall morbidity. It has been suggested that a moratorium on episiotomies would have greater impact on women's faecal continence and quality of life than further increases in caesarean section rates37. While recent researchers blame anal sphincter tears, others blame long second stages with consequent nerve damage38 although the role of pudendal nerve damage and its recovery is now debated39–41. The role of constipation has currently been left out. It worsens with increased parity, and straining may have an effect on the pudendal nerve42. The protective effect of caesarean section even after sphincter repair is debated by researchers in this field43,44. Long term, the incidence of faecal incontinence in older men is as high as in women, suggesting other mechanisms apart from childbirth are important45–47. Awareness of risks and targeting high risk women for anoendoscopy and neurophysiology should help ensure women do not live with chronic incapacitating symptoms, with most women benefiting from normal delivery48. Undoubtedly, nowadays we are more aware of pelvic floor problems but are not yet doing enough to prevent and obviate them. Despite randomised trial evidence of antenatal perineal massage preventing trauma, few units practise this routinely49. If concerns about childbirth damage are overplayed, anxiety in both obstetricians and women will be increased.

A related fear about damage to the perineum is that of dyspareunia and of wrecking couples' sex life and their relationship to one another. The myth has grown that caesarean section keeps the vagina honeymoon fresh and presumably therefore relationships intact. Apart from the contradiction of honeymoon fresh in a pregnant woman, what does this belief reveal? How much of a woman's identity and self-esteem reside in her vagina? What factors in men's and women's sexuality make them adapt or fail to adapt to the change childbirth brings? Whatever the culturally based attitude, what approach should doctors take? Recent work on postnatal symptomatology shows that dyspareunia and failed resumption of sexual intercourse at six months post childbirth are the same whether delivery is spontaneous vaginal, instrumental or caesarean section50. Something more complex is going on. If a doctor performs a caesarean section purportedly to keep the vagina the same, not only may it fail to preserve a fragile relationship (threatened by the new family member, sleeplessness or prior commitment), it may inadvertently reinforce problems of adaptation after the birth. Offering surgery as the primary treatment of normal anxiety around childbirth does not acknowledge the genuine concerns around changing relationships and the loss of control of daily life that children bring. If concerns are not addressed, men are left out, couples are not aided through this life transition, and parents may be less well prepared for child-rearing.

At its extreme, for example after sexual abuse, a bad experience of health care or previous labour, a woman may be petrified of labour (tokophobia), or have posttraumatic stress disorder. In these situations, the primary appropriate management would be psychological support and only then consideration of a caesarean section for preventative mental health indication. It has been documented that in some instances refusal of caesarean section results in severe psychiatric problems51. Referral to an interested psychologist or psychiatrist should be performed in the same way as referral to medical colleagues for medical problems affecting pregnancy. If women with extreme fear of delivery are referred to a psychosomatic clinic for individualised counselling, as many as 56% opt for trial of vaginal delivery and 44% for elective caesarean section52. Within the group of women requesting elective caesarean section are some with bizarre attitudes who may bring forward the date of delivery by request, emotional pressure or deceit53. Obstetricians have to take requests seriously, but not necessarily accept them at presented face value. The danger in not recognising fears and phobias but renaming them requests or even rights is that they remain unaddressed and reinforced. If obstetricians consider labour destructive and dangerous, they will not be able to pass on any confidence in women to labour.

The evidence to date for normal delivery with minimal complications is that women do better with continuity of care and supportive companions54. Good obstetricians who recognise our respective contributions should be working with and fighting for midwives, doulas and other support staff. There is a national shortage of midwives, compounded in London by inexperience and high turnover. An insufficient number of midwives is a dismal indication for caesarean section but may be part of the explanation as to why this debate has come from the capital.

Baby: Mortality

There has been a consistent reduction in the perinatal mortality rate over the last 100 years, although this is as likely to be due to improved social conditions, healthier mothers and neonatal intensive care as obstetric practice55. The relationship between rising caesarean section rates and the perinatal mortality rate is not consistent, questioning whether caesarean section benefits the newborn baby56. In the United States where caesarean section rates have increased year on year since 1965, there is no evidence that maternal and child health has improved as a result57. The World Health Organisation pointed out in 1985 that the countries with some of the lowest perinatal mortality rates in the world had caesarean section rates under 10% and that thus there was no justification in any specific geographical region to have more than 10%–15% caesarean section births58.

The current UK perinatal mortality rate is 7.9 per 1000 births59, largely still related to prematurity and congenital abnormality and with a propensity to occur in lower social classes. It is odd that the current vogue for elective caesarean section has come from the more affluent south east of England and social classes that are not traditionally associated with high perinatal mortality rate.

The antenatal stillbirth rate has remained static compared with the overall reduction in perinatal mortality rate. Proponents of elective caesarean section have speculated that since there is one intrauterine death between 38 weeks and delivery per 600 pregnancies60 (1 in 730 stillbirths)61, elective caesarean section at 38 weeks could be offered to prevent this. It is even postulated that one death per 1500 births of babies of >1.5kg in labour62, one term intrapartum stillbirth per 5000 births, one case of hypoxic–ischaemic encephalopathy per 1750 births and 10% of cases of cerebral palsy would be avoided by a policy of elective caesarean section63. As seductive as this sounds, the logic is that considerably more stillbirths and late pregnancy complications would be avoided by offering caesarean section at 32 weeks. This is not suggested because it is more obvious that iatrogenic damage has been left out of the calculations. The assumption that caesarean section is protective for babies is neither robust nor proven. Perinatal deaths occur even in normal babies after elective caesarean section at term (as high as 1.6% in the term breech trial64 and 0.5% in an observational study of repeat caesarean section)65. The analytical difficulty is that elective caesarean section numbers are smaller than emergency, statistics are not available, and CESDI does not analyse by mode of delivery. With less than 10% of deliveries occurring by elective caesarean section, and even less in low risk pregnancies, rare adverse outcomes and death would not be immediately apparent.

In addition, there might be an excess infant morbidity or mortality in the first year and beyond, which we will presently miss. Imagine two babies conceived on the same day, with the same due date. If the first is delivered electively by caesarean section at 38 weeks and the second by spontaneous labour at 40 weeks, the first has to live to a year and two weeks to reach the same time point as the second for a true comparison of stillbirth and infant mortality risks. Assessing the first at one year after birth artificially decreases the time period of risk by 4%. Infant mortality statistics cannot pick up the effect of iatrogenic early delivery if the reduction in born life is not accounted for. Conventional measures of infant mortality or cot death (known to be associated with prematurity)66 will underestimate any adverse effects as the extra born life after prelabour delivery is not being taken into account in current audit or research calculations.

Morbidity

Whether caesarean section prevents death, fails to prevent death, or even causes death, this is separate from immediate and long term effects of premature abdominal delivery on children's health. Cerebral palsy has not been shown to fall as a result of an increasing caesarean section rate67. Regression analysis of 17,000 vertex births stratified by 500g birthweight found no impact of caesarean section on outcome68. Brain damage occurs without difficult labour or perinatal hypoxia and caesarean section is no guarantee against it. We have been warned before that “it is irrational to ascribe a child's so called brain damage to labour or delivery without considering other factors. There is an interaction of numerous factors, prenatal, perinatal and postnatal and it is simplistic to ascribe ‘brain damage’ to single factors without considering the antecedent causes of these factors”69,70. The corollary is also true. Birth depression and severe hypoxia or acidosis at birth are not largely followed by evidence of ‘brain damage’71,72.

Performing an elective caesarean section before the onset of labour inevitably leads to the baby being born earlier and of lower weight. Does this lead to harm? There is a doubling of respiratory morbidity with each week earlier that elective caesarean section is performed between 37 and 40 weeks73, more transient tachypnoea and respiratory distress74. Although a doubling of morbidity with each week earlier is highly significant as elective caesarean section rates rise, reports also exist of fourfold increases in intermediate or intensive nursery care, mechanical ventilation and oxygen therapy74 and up to 120-fold increase in mechanical ventilation after caesarean section at 37 to 38 weeks75. As many as 1 in 18 (5.5%) babies born by elective caesarean section as opposed to 1 in 63 (1.6%) after vaginal delivery cannot support themselves in room air74, a fact that might worry mothers-to-be, were they informed. Gestational age calculation is not accurate, and up to 9% of babies born by elective repeat caesarean section are thought to be <38 weeks76. As there is no possibility of iatrogenic prematurity and the undisputed immediate respiratory sequelae are less if caesarean section is performed in labour75 those who want to protect babies should consider the rational policy of planned caesarean section at labour onset. Convenience comes into conflict with optimising safety for the babies.

Along with respiratory compromise and admission to the special care baby unit, elective caesarean section brings associated risks such as maternal separation and anxiety, poor feeding, jaundice, cannulation and cross-infection, sometimes with multiresistant hospital acquired organisms. Scalpel lacerations have been reported as 1% with cephalic presentation, up to 8% with abnormal lie77. We have no idea of the longer term consequences. Increasing work on programming in early life brings understanding of how small changes can cascade or be amplified with time. Seemingly small differences in respiratory health, birthweight or breastfeeding might well make significant differences to long term health or development. For example, long term follow up of term infants treated for severe persistent pulmonary hypertension (a condition associated with elective caesarean section78) shows ongoing problems with neurodevelopmental disability, reactive airways disease and slow growth79. Other work shows caesarean section exacerbates maternal–neonatal discordance in gut flora80, which may be related to rising incidence of atopic disease, potentially relievable by perinatal probiotics81.

Basic Medical Concepts

Good medicine starts with an understanding of physiology. Normal pregnancy and childbirth are physiological processes, although disorders exist. Much, although not all, death and damage due to disease can be avoided or treated. Doctors, with their powerful diagnostic tools and treatments for pathology, can also cause harm. Risk–benefit changes with the prevalence of disease, and interventions become relatively more harmful in the well rather than the ill. In general, pregnant women are strong enough to bear children and their babies are capable of surviving. It is true that the outcome of labour is unknown and that nature is not perfect, but it is almost as if some think female physiology flawed and childbirth always pathological. There is plenty of justifiable work so is it well intentioned, misguided or arrogant to think obstetricians can improve on healthy physiology? Respected gynaecological thinkers have written about the evolutionary mistakes of womankind, for example with labour82 or the menopause83,84. Gynaecologists who wished to cure women of being female would find much resonance in the ‘caesarean section for prophylaxis’ argument. By seeing the normal as pathological, the need for everyone to have a doctor is created. Not only does this disregard the evidence about the better outcomes of midwifery-led care85 and diminish our sister profession, it undermines those obstetricians trying to develop evidence-based practice in complicated pregnancies. Focusing on elective caesarean section for choice distracts from the highest risk women on whom the burden of complications of high caesarean section rates and repeated caesarean sections will fall (e.g. older, high parity, high body mass index, socially disadvantaged, ethnic minority or medically complicated women).

Professional Vested Interests

Many reports have identified an individual obstetrician effect on caesarean section rates86–88. The trend of request intervention is more established in the private sector. It has been shown repeatedly that caesarean section rates are higher in the private sector89,90 with lesser strength of indication91. One explanation is that physicians in private practice use different criteria for decision making which would support the changing rates being ‘doctor-led’. There is a financial disincentive for doctors to challenge their clients' requests, and incentives both to make the well worried and perform procedures during office hours. Third-party payers (state or insurance companies) also have an interest in the threshold for interventions, and, as in the USA, it is possible they may wish to step into standard setting to control medical behaviour as has happened in gynaecology.

It is interesting that trends that come from within, and suit, the profession can be projected as coming from outside (the phenomenon of blaming the patient or the lawyer), whereas detailed qualitative work actually demonstrates mechanisms whereby doctors induce the so-called demand for caesarean section by choice92,93.

From the outside, and to historians of medicine, the elective caesarean section debate might appear to be a simple professional boundary dispute about who should control normal pregnancy and childbirth. Turf wars in Britain and North America have carried on through the centuries94,95 with little evidence to support obstetric claims of safety96. The subject has changed, from the registration of midwives, introduction of forceps, or home versus hospital delivery debate to caesarean section. After the limited midwifery success of Changing Childbirth97 it is fascinating to see how the powerful rhetoric of choice and control has moved from the naturalist to the interventionist camps.

The evidence base supports midwifery-led care for low risk women which obstetricians should accept. There is a professional contradiction if an individual obstetrician supports elective caesarean section on request and yet is not active on the delivery suite, helping women who choose labour to get a good outcome from it. To their credit, some of the proponents of elective caesarean section are also active practising obstetricians, but some are not.

Women's Changing Status

Women have made tremendous gains towards personal and financial independence in Western societies in the last century. However, this new self-esteem, which has been hard to win and maintain in the male-orientated culture of work is fragile and brings new fears of loss of self-control. There is a paradox of the strong, controlled career woman who has succeeded by her brain and wits and yet who has lost confidence in her female body. If a generation of women are losing confidence in their innate abilities, this is a tragedy. If deep and genuine fears are driving the requests for caesarean section, then doctors must address them sensitively, but it is difficult to envisage mass surgery as an appropriate solution.

Competing notions of good and bad motherhood are also at stake. It is intriguing that concepts of ‘suffering on behalf of the child’ and ‘putting her own needs first’ resonate on both sides of the debate. Women who want home births without interventions might be considered to suffer without drugs or pain relief for their babies or to satisfy personal needs. Elective caesarean section with postoperative pain and restrictions might be perceived to be avoiding fetal risk in the last weeks of pregnancy and labour but is also requested to suit parental convenience and putatively protect the perineum for sex. As health professionals we should be wary about fashions that ignore physiology (e.g. remembering the harmful, and even fatal, historical fashions for wet-nursing, and formula feeding). There are powerful psychological pressures on women and it is hardly surprising if they continue to feel anxious, guilty or dependent on one or other health professional with their individual ideology.

How should we Respond to Requests?

  • 1The first response must be to listen. Why is this woman requesting a caesarean section? What is her prior knowledge of childbirth and caesarean section and what are the sources of that information? It may be that the obstetric history of her mother, sisters or friends is particularly relevant (and the relationships she has with them). What are her beliefs about her body and its capabilities? What happened in a previous labour?
  • 2Only then should unbiased information be given and management options explored.
  • 3Having a negotiated plan (that is open to renegotiation in late pregnancy or labour) helps women to have control, without being left with all the responsibility, and may help her believe that her professionals understand.
  • 4A proper discussion about labour, the realistic available local midwifery services and continuity of care by a known midwife should be undertaken by a senior midwife.
  • 5For women with a past bad labour experience, reassurance and a written negotiated plan to avoid a replay may be all that is required (e.g. to avoid induction, set a maximum length of labour, different pain relief, avoidance of instrumental delivery, have a known midwife, consider a home delivery etc.).
  • 6For women with nonrecurrent stillbirth/neonatal loss, fears may be controlled with monitoring, plans for delivery at term or even admission.
  • 7For women with intractable anxiety or overwhelming fears referral to a psychologist or psychiatrist should be contemplated (whatever the decision regarding mode of delivery) since anxiety and fear can be disabling and continue after the baby is born.
  • 8There is evidence to support caesarean section for psychological or psychiatric indications as a phobia realised can be very damaging. However, this is something that should be assessed with expert advice as there are other potential modalities for dealing with anxiety and phobia. Fear is not a rational emotion and the range of its manifestation means the obstetrician should be able to provide a range of managements rather than merely reach for the scalpel.
  • 9Doctors who in good conscience feel they cannot recommend a caesarean section (or any other procedure) on demand should refer on. Examples of nonindicated elective caesarean sections might be those in normal women, burnt out herpes, or explained nonrecurrent perinatal death. A second opinion after a refusal would be good practice so that women are given kindly care even if not acquiescence. It is important to avoid conflict in the doctor–patient relationship and offering a referral to a colleague can be a way of continuing the dialogue with the patient, taking her seriously and avoiding any sense of bias or bullying98.
  • 10We are not suggesting obstetricians get in cahoots to refuse all request caesarean sections, as we consider a request as merely the start of a continuing dialogue and process. However, a second opinion would appear a wise routine for any doctor considering an operation with no medical indication. A mandatory second opinion would also be a means for doctors or society to demonstrate the seriousness of executing women's choice (generally good and thoughtful and in agreement with the doctor) if that is what is to be done, in the face of a compelling countervailing consideration (e.g. safety, and especially that of her baby). A system with checks and hurdles can be the pragmatic compromise between ‘first of all, do no harm’ and ‘fulfilling women's' choice’. One private hospital currently charges self-payers £600 more for a caesarean section with no medical indication. This is a crude way to say “have second thoughts before treading this path” but may have a similar intention. On the other hand, it may reinforce an idea that it is better and that NHS patients are subject to rationing.

On a national level, it is imperative that we perform more and better research, especially about longer term maternal and child outcomes of labour and delivery. The implication for longer term funding needs to be taken seriously by grant awarding bodies. Obstetricians and midwives might improve their counselling and better inform women better if good referenced information leaflets were available. To deliver good care to mothers and babies, obstetricians have to play a more active part in the labour ward and take responsibility for emergency outcomes. Intervention rates can fall with simple audit and peer pressure methods99–101. The enthusiasm (and support) for this may be driven by the needs of clinical governance102.

Unanswered Research Questions

The radical interventionist view has drawn important research questions to our attention. What outcomes matter to women, and do their values change with age, parity, ethnicity and experience? What are the risks to mother and baby of each successive elective caesarean section and how can both the natural and iatrogenic complications of labour be minimised? How can we predict problems and protect and repair the perineum? Can randomised trials be performed to get the information about the real differences in outcomes? Who requests elective caesarean section and why? What are women's fears and fantasies about labour and how are they altered by unbiased information? How do women perceive risk and assess information? How do professionals provide information properly? What is the effect of the midwife or obstetrician?

Conclusions

Obstetricians who do not want to harm their two patients must be fact, rather than fear-based. There is currently no evidence that elective caesarean section is safer than labour. If there were, all women would, and should, be offered elective caesarean section. The infant risks and benefits have not been calculated and rapidly rising caesarean section rates may harm women and the next generation. There is a paucity of evidence to counteract the known and inevitable increased risk of maternal death and morbidity following caesarean section. The ethic of informed choice for nonindicated surgery is not overwhelming, and we have put forward other explanations for the recent fashion for elective caesarean section without medical indication.

Ancillary