Memorandum by Alison Macfarlane and Rona
1.1 This memorandum focuses on the evidence
available to support choices about where and with whom to give
birth. Decisions about this have a major influence on the pattern
of care provided, especially for women with uncomplicated pregnancies.
There are, of course, many other factors which can also play a
part, but they are outside the scope of the evidence we have reviewed
over the years.
1.2 At the time the House of Commons Health
Committee set up its enquiry into Maternity services in 1991,
the government's view had long been that every woman should give
birth in a district general hospital. This was based on the assumption
that the safest place to give birth was a consultant obstetric
unit and that the reduction in perinatal mortality in preceding
decades was a direct consequence of the increase in the proportion
of births taking place in large hospitals with consultant obstetric
1.3 We submitted to the Committee a memorandum1
based on our review of the available evidence published in the
British Journal of Obstetrics and Gynaecology2 a fuller review
Where to be born?3 published in 1987 and research published subsequently.
Among our conclusions were two which contradicted the basis for
the government's policies. These were that " ... any statistical
relation between the increase in the proportion of hospital deliveries
and the fall in the crude perinatal mortality rate seems unlikely
to be explained by a cause and effect relation." and that
"... there is no evidence to support the claim that the safest
policy is for all women to give birth in hospital, or for the
policy of closing small obstetric units on the grounds of safety".1
1.4 In this memorandum, we first outline
subsequent policy changes, then discuss the extent to which they
have actually been implemented before closing by examining our
conclusions in the light of research since 1992, including the
second edition of our review4 and subsequent research.
2. POLICY CHANGES
2.1 In its report on Maternity services,
published in 1992,5 the Committee concluded that "On the
basis of what we have heard, this Committee must draw the conclusion
that the policy of encouraging all women to give birth in hospital
cannot be justified on the grounds of safety."5 It went on
to comment that "Given the absence of conclusive evidence,
it is no longer acceptable that the pattern of maternity care
provision should be driven by presumptions about the applicability
of a medical model of care based on unproven assertions."5
2.2 When the government replied to the Committee
in July 1992, its position showed little sign of change, but it
did recognise a need to provide a choice. "The safety of
mother and child must be the prime consideration. It is impossible
to predict all the problems which may arise in labour. Women have
therefore been encouraged to give birth in hospital where back-up
facilities are available in an emergency but the Government recognises
that there are women for whom home birth may be an option they
would prefer. Health authorities are obliged to recognise a woman's
right to choose, and see that a midwifery service is available
for a woman to give birth at home, if that is her choice."6
2.3 The government then appointed the Expert
Maternity Group, chaired by Baroness Cumberlege. Its report, Changing
childbirth,7 published in August 1993, signalled a considerable
change in the direction of policy for maternity care and stressed
that the service must be "woman centred". To achieve
this, it recommended that midwives should have a more prominent
role, with every pregnant woman having a named midwife even if
her care was led by a GP or obstetrician. Increasing choice was
a key theme of the report, with recommendations that each woman
should be provided with accurate and unbiased information about
the options available so that she could make informed decisions,
for instance about where she received care.
2.4 In discussing place of birth, the report
commented that: "Whether a mother with an uncomplicated pregnancy
is putting herself and her child at any greater risk by choosing
to have her baby away from a general hospital is a topic that
has been argued with vehemence and emotion for decades. The inability
to reach agreement after this length of time suggests there is
no clear answer. The job of midwives and doctors, therefore, must
be to provide the woman with as accurate and objective information
as possible, whilst avoiding personal bias or preference."7
2.5 Changing childbirth contained ten indicators
of success to be achieved within five years. It proposed that
every woman should have and know a "lead professional"
and a named midwife. At least 30% should have a midwife as the
"lead professional". Midwives should have direct access
to some beds in all maternity units. At least 30% of women should
be admitted under the management of a midwife and at least 75%
should know the person who cares for them during delivery. It
called for the numbers of antenatal visits to be reviewed and
for all ambulances to have a paramedic able to support midwives
transferring women from home to hospital during delivery. Finally
it suggested that all women should be entitled to carry their
own case notes and have access to information about local services.7
The 10 indicators of success were also listed in an "Executive
letter", entitled "Woman-centred maternity services".8
and a Patient's Charter leaflet.9
2.6 It was widely believed that there was
a legal obligation to provide a home birth service, until the
United Kingdom Central Council for Nursing, Midwifery and Health
Visiting (UKCC) took legal advice and published a position statement
in 2001.10 While fully supporting "the expectation that choice
will be offered to women, and that the option of home birth will
be available," this pointed out that "There is no statutory
duty to provide any specific aspect of the maternity service."10
2.7 This came at a time when the NHS Plan
and related government policy documents were emphasising the need
to reduce geographical variations in the availability of services.11
In parallel with this, the Maternity Care Working Party, set up
by professional and voluntary organisations, published a toolkit
for primary care trusts in 2001. In this, it recommended that
"All women should be given an even-handed explanation of
the options for location, ie home, hospital or community unit.
A 24 hour midwifery service should be available for birth at home."12
2.8 These policies have been reiterated
in the Emerging findings of the Children's National Service Framework
which emphasise the need to promote normality and offer greater
3. TO WHAT
3.1 No formal system was set up to monitor
progress towards the targets in Changing childbirth. A survey
undertaken in 1997 of NHS trusts in England with maternity units
found that between a third and two fifths reported that they collected
data about easily quantifiable Changing childbirth indicators
of success, with only 18 collecting all of them.14
3.2 Data from routine systems give some
insight about the extent of births in different settings, although
there are problems in interpreting them. No data are collected
at a national level about lead professionals but the Hospital
Episode Statistics record the person conducting the delivery and
the type of ward in which the woman gives the birth. The address
at which birth takes place is recorded at birth registration.
A code is allocated to denote the institution in which the birth
took place. Births outside hospitals or other institutions are
subdivided into those at the woman's usual place of residence
and those elsewhere.
Place of birth
3.3 Most women give birth in large maternity
units. According to data from birth registration, nearly a quarter
or 22.8% of women who gave birth England and Wales in 2002 did
so in institutions with 4,000 or more registrable maternities
and a further 22.8% did so in institutions with 3,000-3999 maternities.15
Only 12.8% of maternities took place in institutions with 1,000-1999
maternities and 3.0% on sites with under 1,000 maternities.
3.4 The impact of the European Working Time
Directive and the implementation of neonatal networks is likely
to lead to further closures and mergers, increasing the proportions
of women who deliver in large units and the distances they will
need to travel to do so, even for uncomplicated deliveries. The
alternative to this is to keep maternity units open as midwife-led
units when consultant obstetrics services are withdrawn and to
open new midwife-led units in areas when none currently exist.
This would be in line with the government's wider policy of increasing
the extent to which care is given outside secondary and tertiary
3.5 The percentage of births taking place
at home remains low. At the time of the Committee's previous enquiry
into maternity services, the percentage of maternities at home
in England and Wales had risen slowly but steadily from an all
time low of 0.89% in 1987 to 1.13% in 1991.16 This rise continued
to 2.26% in 1997, after which it levelled off around 2.1%, as
Figure 1 shows.16,17
3.6 There are considerable variations throughout
the country, as Figure 2 shows . This was based on pre-April 2002
health authorities,18 but variations at a more local level are
even wider.18 The highest figure was 6.1% of all maternities in
the former South and West Devon Health Authority area, but within
that, 8.3% of maternities to residents within the Torbay Unitary
Authority occurred at home.19 Within the area served by the West
Herts Hospitals Trust, the proportion of maternities at home in
2001 ranged from 0.7% for residents of the Watford local authority
district and 1.2% in Three Rivers to 3.5% in Dacorum and 3.7%
in St Albans. A survey undertaken by the National Childbirth Trust
which asked for data by trust giving care rather than area of
residence found similarly wide variations.20
Lead professional and type of care
3.5 In England as a whole, the percentage
of hospital deliveries conducted by midwives declined from 75.6
in the financial year 1989-90 to 66.2% in 2001-02, while the percentage
conducted by hospital doctors rose from 23.7 to 33.8%.21 This
mirrors almost exactly the rise in caesarean sections and the
decline in spontaneous delivery. Regional differences in 2001-02
showed the same association. The marked differences between units
within the same region are likely to be a sign of corresponding
differences in the proportion of women delivering with a midwife.
3.6 Relatively few women appear to have
midwife-led care. In 1996-97, when the category was first used,
1% of women in England gave birth in a midwife or other ward.
This rose to reach 3% in 2001-02.21 This coding covers both freestanding
midwife-led units and birth centres as well as midwife led units
in hospitals with consultant units. Deliveries coded in this way
varied markedly by region. They accounted for 15% of women delivering
in the South West Region, 11% of those in the Yorkshire and Humber
Region, 4% of those in the London Region and well under 3% of
those in the other regions.21
3.7 Deliveries coded as occurring in GP
wards declined from 5% in 1989-90 to 2% 1996-97 and 1% in 2001-02.21
This is not surprising as "isolated GP units" had been
closing for many years. Most of the care in the units was given
by midwives and during the 1990s many of the remaining GP units
changed their designation to midwife-led units.
3.8 In 2001-02, 40% of deliveries were coded
as occurring in combined "consultant/GP wards", even
though only an estimated 200 women were delivered by GPs.20 This
suggests that either the designation of wards or the way in which
they are coded had not been changed to reflect the declining involvement
4. ARE EARLIER
4.1 In this section we summarise the conclusions
of research and audit undertaken since the Committee's last report.
In doing so, we draw on the second edition of our review Where
to be born?, published in 1994.4 Between us, we have summarised
more recent research in successive editions of the MIDIRS peer
reviewed informed choice leaflet on the subject, the most recent
of which was published in 2003.22 As in earlier research, it is
impossible to make direct comparisons between the safety of birth
in different settings. This is because there are very few deaths
among babies born to women who book to give birth in midwife led
units or at home.
4.2 Three randomised trials have compared
midwife-led units with consultant obstetric care in the same hospital.23-25
All found a higher rate of electronic fetal monitoring and episiotomy
among women receiving consultant care and in two trials more fetal
distress was detected. No significant differences were found in
mortality or life-threatening morbidity, but much larger trials
would be needed to detect any such differences reliably. Two trials
found that the women allocated to midwifery units were more satisfied
with their care.24,25 One of these found that women allocated
to midwifery care reported receiving more consistent advice, saw
fewer care providers and experienced greater continuity of intrapartum
care.26 The third reported no difference in overall satisfaction.27
Midwife-led units in hospitals without consultant
4.3 Women who booked to deliver at the midwife
led unit at the Royal Bournemouth Hospital and at the Edgware
Birth Centre were compared with women who satisfied the criteria
for delivering at the units but booked to deliver in consultant
units elsewhere.28,29 Booking for the midwifery units was associated
with lower rates of induction of labour and episiotomy and similar
rates of perineal tears. One of the studies found higher rates
of augmentation of labour and use of anaesthesia among women booked
for hospital28 and the other found higher rates of ventouse delivery
and elective caesarean sections.29
4.4 A survey of planned and unplanned home
births from 1981 to 1994 in the former Northern Region estimated
a perinatal mortality rate for women booked for home delivery
and delivering there to be 1.7 per 1,000. When women who transferred
to hospital were included, the estimated perinatal mortality rate
was 4.8 per 1,000.30 This is less than half the rate for the region
as a whole, the women booked for home birth did not include those
classed as "high risk".
4.5 A survey by the National Birthday trust
fund of 6,044 planned home births in 1994 and 4,724 births in
hospital to broadly similar women found no maternal deaths and
low perinatal mortality rates were among the women who booked
for home birth.31 An economic analysis of the data suggested that
the better outcome alongside the lower expected costs per case
lead us to conclude that . . . a real option of a home birth for
all women who want it would also be a cost effective option.'31
4.6 Overall perinatal mortality rates for
England and Wales continued to fall when percentages of home birth
first rose from the late 1980s and then levelled off in the mid
to late 1990s. If Figures 1 and 3 are compared, it can be seen
that there is no longer any statistical association between them.
4.7 National statistics do not distinguish
between planned and unplanned home births, but the perinatal mortality
rate for all births at home in England and Wales fell in the mid
1990s. Since 1994 it has been lower than that for all births in
hospital, as Figure 3 shows. This is because the numbers of probably
unintentional births of low birthweight babies at home did not
increase as overall numbers of home births rose through the 1990s.16,32
This contrasts with the position in 1979, when a national survey
showed that a third of births at home were unplanned and they
contributed substantially to the high perinatal mortality rates
seen in Figure 3.33 It was the high overall rates for 1975-77
which led the Committee's predecessor, the House of Commons Social
Services Committee to draw incorrect conclusions about the safety
of planned home births.
4.8 Claims have been made in the British
Medical Journal that home birth is no longer safe and that hospital
delivery is three to four times safer than home delivery.34 This
draws on data from a review of by the Confidential Enquiry into
Stillbirths and Deaths in Infancy review of 22 intrapartum deaths
among planned home births.35 Closer examination shows that these
claims are not supported by the available evidence.36-38
4.9 For a variety of different reasons,
women may change their planes during pregnancy or transfer to
a consultant obstetric unit during or immediately after labour,
if problems arise. The Maternity Hospital Episode Statistics tries
to capture data about transfers during labour, but they are incomplete.21
A national survey of maternity units in 2000 found that rates
of transfer from planned home births to hospital ranged from zero
to over 30% at each stage.20
4.10 Many different selection criteria are
in current use and they are based on early epidemiological research
and common practice.39 There is no research evidence to indicate
which would lead to the best outcomes. As `risk scoring' systems
are designed to identify groups of women who are at higher risk
of adverse outcome, their ability to identify individual women
who may experience problems during labour and delivery has not
4.11 Although it is unlikely that many women
perceive that they have a choice about were to give birth, this
does not necessarily imply a lack of interest. A survey commissioned
by the Expert Maternity Group from MORI Health Research asked
1,005 women who had given birth since 1989 whether they would
like to have the option of another type of care and delivery if
they became pregnant again. Seventy two per cent said they would
have liked an option other than delivery in a consultant unit.
Of these, 22% said they would have liked the choice of a home
birth and 44% a "domino" delivery. These accounted for
16% and 32% of all women questioned, respectively.7
The need for better data about the place of birth
5.1 In order to inform choice, better data
are needed about birth in different settings and its outcome,
including transfers in labour and after birth. As originally envisaged,
data about all registerable births, including those at home and
in non-NHS hospitals should be submitted to the Maternity Hospital
Statistics. The feedback of Maternity Hospital Statistics to maternity
units and its wider dissemination should be improved.
5.2 A new classification of maternity units
and definitions of transfer should be devised to reflect the way
care is now given. This should be incorporated into the maternity
care data dictionary, into individual computer systems, the Maternity
Hospital Episode Statistics and other relevant systems. It should
be reviewed and updated on a regular basis.
5.3 The current proliferation of initiatives
relating to the collection of data about maternity care is leading
to duplication of effort and inefficient use of resources which
are needed to improve data collections systems. The initiatives
should be rationalised and the data collection aspects should
be brought within the scope of National Statistics and its Code
Interpreting the available data and evidence
5.4 Despite their deficiencies, the available
data show marked differences in the types of care available in
different parts of the country and that access to midwife led
care, whether in midwife led units and birth centres or at home,
5.5 There is still no evidence that the
safest policy is for all women to give birth in a hospital with
a consultant obstetric unit. There is no longer any statistical
association between perinatal mortality and the percentage of
births at home.
5.6 These conclusions are based on research
undertaken in the United Kingdom. Because of differences in health
care systems, findings from other countries cannot be directly
applied to the situation here.41 It is notable, however, that
research elsewhere has reached broadly similar conclusions.42
5.7 The policy of concentrating services
for women with obstetric complications and sick newborn babies
in secondary and tertiary centres has occurred in response to
a variety of pressures and is in line with government policy in
other areas. In contrast and in addition to the lack of evidence,
it is out of line with the emerging findings of the National Service
Framework and policies in other areas of health care for women
with uncomplicated pregnancies to give birth in such settings,
especially when doing so involves journeys of some distance. When
consultant obstetric services are withdrawn from maternity units,
there are strong arguments for retaining them as midwife led units.
1 Campbell R, Macfarlane AJ. Where to be
born? Written evidence to the House of Commons Health Committee
enquiry into maternity services. In: House of Commons Health Committee.
Maternity Services, Volume III, appendices to the minutes
of evidence. Second report, session 1991-92, HC 29-III. London:
2 Campbell R, Macfarlane AJ. Place of delivery:
a review. British Journal of Obstetrics and Gynaecology 1986;93:675?83.
3 Campbell R, Macfarlane AJ. Where to
be born? The debate and the evidence. Oxford: NPEU 1987.
4 Campbell R, Macfarlane AJ. Where to
be born? The debate and the evidence. Second edition. Oxford:
National Perinatal Epidemiology Unit, 1994.
5 House of Commons Health Committee, (Chairman
N Winterton). Maternity services. Vol I, report. HC 29-I.
London: HMSO, 1992.
6 Department of Health. Maternity services.
The Government's response to the second report from the Health
Committee, Session 1991-92, Cm 2018. London: HMSO, 1992.
7 Department of Health. Changing childbirth.
Part I. Report of the Expert Maternity Group. London: HMSO,
8 NHS Management Executive. Woman-centred
maternity services. Executive letter EL(94)9. January 24 1994.
9 NHS. The patient's charter. Maternity
services. Leeds: Department of Health, 1994.
10 United Kingdom Central Council for Nursing,
Midwifery and Health Visiting. Supporting women who intend
to give birth at home. London: UKCC, 2001.
11 Department of Health. The NHS plan.
(Cm 4818-I). London: Stationery Office, 2000.
12 Maternity Care Working Party. Modernising
maternity care. London: Maternity Care Working Party, 2001.
13 Department of Health. Getting the
right start: National Service Framework for Children: emerging
findings. London: Department of Health, 2003.
14 Kenney N, Macfarlane A. Identifying problems
with data collection at a local level: survey of maternity units
in England. BMJ 1999; 319(7210) :619-22.
15 Office for National Statistics. Unpublished
analysis of birth registration data for 2002.
16 Macfarlane AJ, Mugford M, Henderson J,
Furtado A, Stevens J, Dunn A. Birth counts: statistics of pregnancy
and childbirth. Volume 2,Tables. Second edition. London: The
Stationery Office, 2000.
17 Office for National Statistics. Birth
and patterns of family building, England and Wales. Series
FM1. Published annually.
18 Office for National Statistics. Birth
and patterns of family building, England and Wales, 2001. Series
FM1, No 30. London: TSO, 2002; Table 8.2.
19 Office for National Statistics, VS1 tables,
2001. Titchfield: Office for National Statistics, 2002.
20 National Childbirth Trust. Home birth
in the United Kingdom. London: NCT, 2001.
21 Department of Health. NHS maternity
statistics, England: 2001-02. Statistical bulletin 2003/09.
London: Department of Health, 2003.
22 MIDIRS. Place of birth. In: MIDIRS
informed choice information pack. Bristol: MIDIRS, 2003.
23 Hundley VA, Cruickshank FM; Lang GD et
al. Midwife managed delivery unit: a randomised controlled
comparison with consultant led care. Br Med J 1994;309:1400-4.
24 Macvicar J, Dobbie G, Owen-Johnstone
L et al. Simulated home delivery in hospital: a randomised
controlled trial. Br J Obstet Gynaecol 1993;100:316-23.
25 Turnbull D, Holmes A, Shields N et al.
Randomised, controlled trial of efficacy of midwife-managed care.
26 Turnbull D, Shields N, McGinley M et
al. Can midwife-managed units improve continuity of care?
Br J Midwifery 1999;7:499-503.
27 Hundley VA, Milne JM, Glazener CMA et
al. Satisfaction and the three C's: continuity, choice and control.
Women's views from a randomised controlled trial of midwife-led
care. Br J Obstet Gynaecol 1997;104:1273-80.
28 Campbell R, Macfarlane A, Hempsall V
et al. Evaluation of midwife-led care provided at the Royal Bournemouth
Hospital. Midwifery 1999;15:183-93.
29 Saunders D, Boulton M, Chapple J et
al. Evaluation of the Edgware Birth Centre. Harrow: Northwick
Park Hospital, North Thames Perinatal Public Health, 2000.
30 Northern Region Perinatal Mortality Survey
Coordinating Group. Collaborative survey of perinatal loss in
planned and unplanned home births. Br Med J 1996;313:1306-9.
31 Chamberlain G, Wraight A, Crowley P eds.
Home births: the report of the 1994 confidential enquiry by
the National Birthday Trust Fund. Carnforth: Parthenon, 1997.
32 Office for National Statistics. Mortality
statistics: childhood, infant and perinatal. Series DH3. London:
TSO, published annually.
33 Campbell R, Davies IM, Macfarlane A,
Beral V. Home births in England and Wales, 1979: Perinatal mortality
according to intended place of delivery. Br Med J 1984;289:721-4.
34 Drife J. Data on babies' safety during
hospital births are being ignored. Br Med J 1999;319:1008.
35 Maternal and Child Health Research Consortium.
Confidential Enquiry into Stillbirths and Deaths in Infancy
[CESDI]. 5th annual report. London: Maternal and Child
Health Research Consortium, 1998.
36 Young G, Hey E. Choosing between home
and hospital delivery. Home birth in Britain can be safe. Br
Med J 1999;320:798.
37 Macfarlane A, McCandlish R, Campbell
R. Choosing between home and hospital delivery. There is no evidence
that hospital is the safest place to give birth. Br Med J 2000;320;798.
38 Chamberlain G. Choosing between home
and hospital delivery. Risk of home birth in Britain cannot be
compared with data from other countries. Br Med J 2000;320:798-9.
39 Campbell R. Review and assessment of
selection criteria used when booking pregnant women at different
places of birth. Br J Obstet Gynaecol 1999;106:550-6.
40 Reynolds JL, Yudkin P, Bull MJV. General
practitioner obstetrics: does risk prediction work? J R Coll
Gen Pract 1988;38:307-10.
41 Macfarlane A. The safest place of birthis
there a better analysis than meta-analysis? [commentary]. Birth
42 Olsen O. Meta-analysis of the safety
of home birth. Birth 1997. 24(1):4-13.