Maternal and neonatal individual risks and benefits associated with caesarean delivery: multicentre prospective study
BMJ 2007; 335 doi: https://doi.org/10.1136/bmj.39363.706956.55 (Published 15 November 2007) Cite this as: BMJ 2007;335:1025
All rapid responses
Further to Consultant Obstetrician Mike Wyldes' comments on his
investigation into the CEMACH dataset ['Elective Caesarean section safest
form of childbirth', above]:-
In April 2008, the UK's Birth Trauma Association also analyzed data
taken from the latest CEMACH report, and found that planned cesarean
delivery had the lowest maternal mortality rate compared with all other
births: "of the 2,113,831 women who delivered a baby after 24 weeks
gestation between 2003 and 2005, one in 10 had a caesarean before labour
had begun. Seven women died, giving a mortality rate of 0.31 per 10,000.
This compared to 74 deaths amongst the remaining women who had a natural
birth or an emergency caesarean section, giving a mortality rate of 0.39
per 10,000."(1)
Certainly, planned cesarean delivery results in greater abdominal
morbidity than PVD, but on the other hand, PVD results in greater pelvic
floor trauma than planned cesarean delivery (see this week's news report
on the U.S. 'Fourth International Consultation On Incontinence (ICI)' for
example(2) and my comment on the cesarean benefit of protection against
fecal incontinence(3)).
Women should be honestly informed of each set of birth risks and
benefits, supported during their decision-making process and have their
final choice respected - whether it's vaginal or cesarean delivery.
Finally on this issue, a 2003 HealthGrades Quality Study(4) in the
U.S. "identified an association with higher vaginal complication rates in
those hospitals that did fewer than expected preplanned cesarean sections"
and likewise, lower vaginal complication rates in hospitals with more
preplanned cesareans than expected. The report said that this finding was
"suggestive of, but not definitive of, inappropriate under-utilization of
preplanned first time C-sections in those hospitals", and that further
studies are needed.
Frustratingly, five years after this report, the debate over planned
cesarean delivery and PVD comparisons remains hampered by a lack of
appropriate and relevant clinical studies.
References:
(1)http://www.telegraph.co.uk/news/uknews/1584671/Women-choosing-
caesarean-have-low-death-rate.html
(2)http://www.medicalnewstoday.com/articles/116063.php
(3)http://www.medicalnewstoday.com/youropinions.php?associatednewsid=116063
(4)http://www.healthgrades.com/media/english/pdf/Patient_Choice_Csection_St...
Competing interests:
I am editor of the website www.electivecesarean.com
Competing interests: No competing interests
The effects of anesthetic risk on maternal morbidity and mortality
were not taken into account.
It is accepted that after maternal hemorrhage, anesthesia related problems
are the leading cause of perinatal maternal mortality.
Competing interests:
None declared
Competing interests: No competing interests
Sir,
Caesarean section is the last resort as Mother Nature takes care of the
labour. The doctor's lack of insight and the people's demand to decide the
date time for delivery takes the logic out of decision-making.
Man's defiance to suit his convenience has left him more mechanical than
biological. It is taught in medical schools to wait for as long as
possible unless there are strong indictions to resort to Caesarean
section.In developing countries what was exception earlier has now become
a rule.Studies such as these need to be conducted in such a setting before
jumping to conclusions soon lest the glamour this procedure spoil Nature's
Beauty just for a few pounds more!
Prof Mudassir Azeez Khan
Head,Department of Community Medicine,
Mysore Medical College,
Mysore
Competing interests:
None declared
Competing interests: No competing interests
Sir,
I have a total of 23 years Obstetric experience in the UK, 13 years as a
Consultant in a large maternity unit and with 2 years full-time research
in perinatal mortality statistics during training. Villar and Colleagues
(BMJ 2007;335:1025 (17 November), oi:10.1136/bmj.39363.706956.55
(published 30 October 2007) present data from South America showing that
elective Caearean delivery (El CD) is associated with a 1.66 fold increase
in risk of neonatal death in cephalic presentation (table 3). This is a
worrying statistic for any women deciding the best mode of birth for her
baby, and would suggest that a vaginal birth is safer for the baby than El
CD.
To estimate this risk for a UK Population I have analysed the
comprehensive dataset collected by the Confidential Enquiry for Maternal
and Child Health (CMACH), with very different results. During the period
1995 to 2006 there were 780,370 total births in the West Midlands region
of England. During that time there were a total of 4054 neonatal deaths
(up to 28 days), of which 176 were following El CD. Of these 111 were
caused by lethal congenital malformations and an additional 50 were due to
prematurity. Amongst this massive cohort there were only 15 deaths
following El CD after 37 weeks, with a neonatal mortality rate for El CD
at term in normally formed infants of 0.19 per thousand (one in 5,263),
compared to the headline figure of 7.7 per thousand (one in 130) presented
in your paper. A difference of more than 40 fold.
In my opinion the paper, and associated editorial gives the false
impression that El CD is a cause of neonatal death. The fact is that
elective caesarean section at term is a very safe method of birth for the
baby, and is safer (for the baby) than planning vaginal birth. The
presentation of data in this way gives me no confidence that the morbidity
data, which is much more difficult to quantify, is accurate and unbiased.
I suspect that the vast majority of these cases are in fact deliveries
associated with prematurity and other obstetric problems. In my opinion
the WHO and BMJ have joined together to present a completely false
statistic for the risk associated with El CD, and this has not been picked
up or corrected by either the peer review process or the associated
editorial.
It is regretable that the false impression given has been widely reported
and disseminated, but the true mortality rates - which are well collected
and reported for all UK deaths through CMACH are not given the same
prominence.
Yours sincerely
Mike Wyldes MA FRCOG
Consultant Obstetrician
Clinical Lead for Labour Wards at Heart of England Foundation Trust
Data from West Midlands Perinatal Institute, CMACH Regional data
collector.
Competing interests:
Obstetrician in NHS and Private Practice in UK
Competing interests: No competing interests
This study needs to be interpreted with extreme caution when
counselling women who have to decide whether to undertake planned
caesarean section (CS) or not. The data can be represented to show that
among women delivered in Latin America, including those delivered
prematurely and at term, the observed outcomes for delivery by planned CS
versus the alternative (vaginal birth or intrapartum CS) were: 1:2640 v
1:4503 risk of maternal death; 1:37 v 1:134 risk of admission to ICU;
1:102 v 1:199 risk of blood transfusion; 1:285 v 1: 928 risk of
hysterectomy; 1: 39 v 1:85 risk of hospital stay >7 days; 1:570 v 1:172
risk of 3rd or 4th degree perineal laceration or post-partum fistula; 1:
323 v 1:249 risk of fetal death; 1:230 v 1:341 risk of neonatal death.
There are a number of problems with applying these findings to
prospective decision-making in circumstances where planned Caesarean
Section is not considered mandatory e.g. for ‘low-risk’ women and fetuses
at term. Firstly, the exclusion from the study of women who underwent what
the authors term emergency Caesarean Delivery e.g. for fetal distress,
eclampsia, or severe haemorrhage negates any usefulness of the study in
the above-mentioned respect since a woman choosing to deliver by CS at 38
weeks gestation, for example, avoids by that choice the development of
such a complication that may lead to perinatal or maternal death at 39
weeks thereafter unless she was instead to choose induction of labour,
which may or may not be successful and would have its own attendant risks
and drawbacks. Secondly, the reported maternal mortality in a group from
whom most high-risk women had already been excluded is completely out of
line with UK maternal mortality figures of approximately 1: 59,000 of
vaginal births and 1: 12,000 caesarean births (1). This alone demonstrates
either a totally different form of care or demography between the study
group and the UK population. Third, in the above-mentioned decision-making
scenario CS is almost always performed at term. In the study group a
perinatal death rate of 1:323 and neonatal death rate of 1:230 was
reported in a way that suggests that these may have been attributable to
delivery by CS. The authors do not report on the proportion of women in
the study with secure pregnancy dating based upon reliable early
ultrasound scans, but if such scans are not universal in Latin America
inaccurate estimation of gestation could potentially be an important cause
of morbidity among women delivered by planned CS. These deaths would be
expected to be almost totally avoidable among low-risk babies with secure
pregnancy dating and delivered by CS at term in the UK, thus tipping the
balance strongly in favour of planned CS where fetal outcome is of
especially high significance e.g. older women with a history of
subfertility. Fourthly, the report of a 1:172 risk of 3rd or 4th degree
tears or vaginal fistula among the vaginal delivery group is 7 times lower
than other reported estimates for the clinical diagnosis of such injuries
(2), and even further out of line with data based on objective
investigations (3).
The study is interesting for highlighting some differences between
obstetric practices and outcomes between Latin America and the UK but of
little relevance to UK practice and the ongoing debate regarding the risks
and benefits of maternal choice and planned CS. It is to be hoped that its
findings will not be taken out of context.
1. Confidential Enquiries Into Maternal Deaths. Why mothers die 1997-
1999: The
fifth report of the Confidential Enquiries into Maternal Deaths in the
United
Kingdom. No. 5. Regent's Park, London: RCOG Press; 2001.
2. Stamp G, Kruzins G, Crowther C. Perineal massage and prevention of
perineal trauma: randomised controlled trial. BMJ 2001; 322: 1277-1280.
(26 May.)
3. Fines M, Donnelly V, Behan M, O'Connell PR, O'Herlihy C. Effect of
second vaginal delivery on anorectal physiology and faecal continence: a
prospective study. Lancet 1999; 354: 983-986
Competing interests:
None declared
Competing interests: No competing interests
If we want to know the effects of C-section, we must have two groups
of patients with similar obstetric risk factors. The C-section group in
the presented study is destined to have worse maternal and neonatal
outcomes, no matter what the mode of delivery, because they are all sick
mothers or high-risk pregnancies.
On the other hand, it is clear the C-section saves more fetuses in
breech presentation - which is not a surprise. I would expect C-section is
beneficial in most normal presentation cases in the current study too -
these fetuses were likely distressed intrauterine. Even the short-term
benefit is less clear, the reduced exposure to hypoxia intrauterine may
result in better long-term neuro-developmental health.
Competing interests:
None
Competing interests: No competing interests
This study by Villar et al in this weeks BMJ concerning caesarean
childbirth tells us nothing a first year registrar could not have told us.
Women undergoing caesarean section - both emergency and elective - are
more likely to suffer morbiditiy and mortality than those that give birth
vaginally at least in part because they are more likely to have suffered
placenta praevia, placental abruption, diabetes, major medical problems,
preeclampsia and prolonged or obstructed labour - conditions not
adequately controlled for in Villar's study. Babies born by caesarean are
more likely to die or suffer harm because they will be either smaller or
larger than those delivered vaginally. They will have suffered greater
degreees of acute or chronic intrauterine compromise than those delivered
vaginally and of course emergency (either in or before labour) caesarean
is a likely - and indeed appropriate - response to evidence of fetal
compromise. Comparing women who suffer medical complications and obstetric
morbidity with those who don't does not help us.
Equipoise exists. The time has come for a large prospective long term
observational study comparing outcomes for women without absolute
indications for caesarean who plan to give birth by caesarean versus those
planning to give birth vaginally.
Competing interests:
None declared
Competing interests: No competing interests
The study of Villar et al is clearly an important contribution
towards understanding the associations between caesarean section and
maternal and infant outcomes. One of their key findings should, however,
be interpreted with some caution. They found an absolute risk of neonatal
mortality associated with elective caesarean section of 77 per 10,000.
This was significantly higher than vaginal birth. However, this absolute
risk is much higher than large scale studies of planned repeat caesarean
section at term. A report of 9,104 procedures in Scotland between 1992 to
1997 found a risk of 1.1 per 10,000 among women with a cephalic
presentation at term (1). That study excluded deaths attributed to
congenital anomaly. A report of 15,014 procedures at term in large centres
in the USA found a risk of 5 per 10,000, which included all causes of
death (2). The most likely explanation for the greater absolute risks of
Villar et al is that they included both preterm and term procedures. Given
the disproportionate contribution of preterm births to the risk of
neonatal mortality, a relative risk for neonatal death associated with
caesarean section from a population including both preterm and term births
may not be applicable for the large majority of women considering elective
caesarean delivery. These authors should present absolute and relative
risks of neonatal death for planned caesarean section having stratified
their study group into preterm and term deliveries.
Yours sincerely,
Gordon C S Smith, MD PhD.
References
1. Smith GC, Pell JP, Cameron AD, Dobbie R. Risk of perinatal death
associated with labor after previous cesarean delivery in uncomplicated
term pregnancies. JAMA 2002;287:2684-90.
2. Landon MB, Hauth JC, Leveno KJ, Spong CY, Leindecker S, Varner MW et
al. Maternal and perinatal outcomes associated with a trial of labor after
prior cesarean delivery. N Engl J Med 2004;351:2581-9.
Competing interests:
None declared
Competing interests: No competing interests
Sir,
Villar et al (1) have provided further evidence that caesrean delivery is not without risk, and we must carefully balance the risk of interfering with physiology. They confirm the work of others showing the higher risk of respiratory problems for elective caesarean section before the onset of labour. Elective caesarean section interferes with all the physiological processes of labour and delivery and it is the absence of these physiological processes which seem likely to be the underlying cause of the neonatal morbidity. The absence of a surge in corticosteroids, catecholamines and prostaglandins is likely to be a part of it. The lack of compression of the fetal lungs during labour and delivery is another effect of elective caesarean section. The common practice of immediate cord clamping at caesarean delivery is equally non-physiological and has several effects. Firstly it interferes with the normal transition from fetal to adult pattern circulation. The pulmonary circulation is not fully established until the baby has regular breathing. This normally takes about five minutes after birth. Abruptly occluding the placental circulation interferes with this process. Even with a superficial examination of the fetal and adult pattern circulation it can be seen that the serial output of the two sides of the adult pattern heart cannot be equal until the pulmonary circulation is fully established. Obstructing the placental circulation before respiration has started must put a strain on the heart and immediately stops any oxygenated blood returning from the placenta. Secondly the normal closure of the arterial circulation first, then later by the venous circulation, is not followed. The closure of both arteries and vein at the same time prevents the normal "placental transfusion". This leads to hypovolaemia in the newborn. Early clamping interferes with physiology. It is an unnecessary intervention.
It is interesting that in the animal experiments investigating fetal asphyxia and resuscitation, they did not start the resuscitation until after the umbilcal cord had been clamped and cut. (2) The paediatrician is used to being called to attend to a baby with the cord already clamped. The mindset is so established that as soon as the baby is born the placental circulation is considered redundant. In nature and physiology the placenatal circulation closes a few minutes after birth in response to the high oxygen tensions of pulmonary respiration. Even in the sick baby interference with this process does not appear to have obvious dire consequences. The baby is already thought to be struggling to survive. If resuscitation results in an improvement in its condition, the resuscitaion process is given the credit. When resuscitation fails, failure is attributed to the poor condition of the baby at birth. It is never considered that its condition may have been moved from serious to unrecoverable by clamping the cord, by removing its only supply of oxygen and rendering it hypovolaemic. For those who do recover, early cord clamping could contribute to many conditions in later life including the respiratory problems described in this paper. Resuscitation with the placental circulation intact is much closer to physiology than obstructing a functioning placental circulation and then inflating the lungs.
When it is suspected that the fetus is suffering from severe hypoxia in labour it is logical to bring labour to an end by a quick delivery so that the baby can change over to use it own pulmonary respiratory system. The placental respiratory system is failing. Note the word failing not failed. Imagine a battery electric pump which is running low. If the continuous function of the pump was important, we would not remove the old battery then fix up the new. No, we would link up the new battery in parallel before removing the old one. It might be a bit aukward to do it this way but . . .. Why then do we disconnect the failing battery supply of the baby before the new supply is working properly? Is it simply because the power cable (umbilical cord) will not reach as far as the resuscitaire positioned at the wall of the delivery room?
David Hutchon FRCOG
References
1. Villar J et al. Maternal and neonatal individual risks and benefits associated with caesarean delivery: multicentre prospective study. BMJ online 30 Oct 2007
2. Cross KW, Dawes GS, Hyman A, Mott JC. Hyperbaric oxygen and intermittent positive pressure ventilation in resuscitation of asphyxiated newborn rabbits. Lancet 1964 ii 560-2.
Competing interests:
None declared
Competing interests: No competing interests
Re: Maternal and neonatal individual risks and benefits associated with caesarean delivery: multicentre prospective study
After reading your study, it's still unclear why you didn't analyse the isolated results for the outcome of cesarean surgery performed on women WITHOUT any risk factors.
If you compare the total amount of elective cesarean surgery INCLUDING women with risk factors (more than 50%, as reported in your study) with ALL patients in the normal labour group (in which most of them are not in the High risk group), it's clear that the outcome of the elective cesarean group will be worse. After all, you are comparing patients at different risks.
The study should provide more information and ISOLATE the low risk elective cesarean delivery to be able to calculate the impact of the cesarean surgery WITHOUT association with risk factors.
It would be interesting if you could publish this data.
Competing interests: No competing interests