JAN

JOURNAL OF ADVANCED NURSING

REVIEW PAPER

Comparing midwife-led and doctor-led maternity care: a systematic

review of reviews
Katy Sutcliffe, Jenny Caird, Josephine Kavanagh, Rebecca Rees, Kathryn Oliver, Kelly Dickson, Jenny
Woodman, Elaine Barnett-Paige & James Thomas
Accepted for publication 3 March 2012

Correspondence to K. Sutcliffe:
e-mail: k.sutcliffe@ioe.ac.uk
Katy Sutcliffe BA MSc PhD
Research Officer
Evidence for Policy and Practice Information
and Co-ordinating Centre (EPPI Centre),
Social Science Research Unit, Institute of
Education, University of London, UK
Jenny Caird BSc MA
Research Officer
Evidence for Policy and Practice Information
and Co-ordinating Centre (EPPI Centre),
Social Science Research Unit, Institute of
Education, University of London, UK
Josephine Kavanagh BA MA
Research Officer
Evidence for Policy and Practice Information
and Co-ordinating Centre (EPPI Centre),
Social Science Research Unit, Institute of
Education, University of London, UK
Rebecca Rees MA MSc FHEA
RCUK Academic Fellow
Evidence for Policy and Practice Information
and Co-ordinating Centre (EPPI Centre),
Social Science Research Unit, Institute of
Education, University of London, UK
Kathryn Oliver MSc
Post Graduate Research Student
Centre for Occupational and Environmental
Health, University of Manchester, UK

SUTCLIFFE K., CAIRD J., KAVANAGH J., REES R., OLIVER K., DICKSON K.,

Comparing midwifeled and doctor-led maternity care: a systematic review of reviews. Journal of
Advanced Nursing 68(11), 23762386. doi: 10.1111/j.1365-2648.2012.05998.x
WOODMAN J., BARNETT-PAIGE E. & THOMAS J. (2012)

Abstract
Aims. A report of a systematic review of reviews which examines the impact of
having midwives-led maternity care for low-risk women, rather than physicians.
Background. A rising birth rate, increasing complexity of births, and economic
constraints pose difficulties for maternity services in the UK. Evidence about the
most effective, cost-effective, and efficient ways to give maternity services is needed.
Data sources. Searches were carried out in AugustSeptember 2009 of ten electronic databases, 16 key nursing and research websites, and reference lists of 56
relevant reviews. We also contacted 38 experts for information. No date restrictions
were employed.
Review methods. A narrative review of systematic reviews or meta review was
conducted using transparent and systematic procedures to limit bias at all stages.
Systematic reviews that compared midwife-led care during pregnancy and birth with
physician-led care were eligible for inclusion.
Results. Three meta-analytic reviews were included. Midwife-led care for low-risk
women was found to be better for a range of maternal outcomes, reduced the
number of procedures in labour, and increased satisfaction with care. For some
maternal, foetal, and neonatal outcomes reviews found no evidence that care led by
midwives is different to that led by physicians. No adverse outcomes associated with
midwife-led care were identified.
Conclusions. For low-risk women, health and other benefits can result from having
their maternity care led by midwives rather than physicians. Moreover, there appear
to be no negative impacts on mothers and infants receiving midwife-led care.
Keywords literature review, maternity care, meta-review, midwifery, nursing,
systematic review

continued on page 2

2376

 2012 Blackwell Publishing Ltd

JAN: REVIEW PAPER

Kelly Dickson BSc MSc

Research Officer
Evidence for Policy and Practice Information
and Co-ordinating Centre (EPPI Centre),
Social Science Research Unit, Institute of
Education, University of London, UK
Jenny Woodman BA MA MSc
PhD Research Student
Institute of Child Health, University College
London, UK
Elaine Barnett-Paige BA MA
Research Officer
Evidence for Policy and Practice Information
and Co-ordinating Centre (EPPI Centre),
Social Science Research Unit, Institute of
Education, University of London, UK
James Thomas MA PhD MMus
Social Science Research Unit Assistant
Director
Evidence for Policy and Practice Information
and Co-ordinating Centre (EPPI Centre),
Social Science Research Unit, Institute of
Education, University of London, UK

Introduction
Nurses and midwives make up the largest group of healthcare
providers in virtually all countries. This vast and important
workforce is currently facing many challenges due to factors
such as increasing population growth rates, ageing populations, and a growing burden of long-term and non-communicable disease (ICN/FNIF 2006). For midwives in the UK,
challenges include a rising birth rate and increased numbers
of complex births (Royal College of Midwives 2011).
Difficulties across the sector are compounded by the twin
pressures of shortages in supply and increased demand for
nurses and midwives both in the UK (Royal College of
Midwives 2011) and in many health systems across the world
(ICN/FNIF 2006) and by the current global economic crisis
(WHO 2009). Governments in various countries are acting to
address these pressures. For example, the Initiative on the
Future of Nursing was set up in 2010 by the US government
to find solutions for the continuing challenges faced by the
nursing profession there. In the same year the Prime Ministers Commission on the Future of Nursing and Midwifery
was launched in the UK (Prime Ministers Commission on the
Future of Nursing and Midwifery 2010a). Recognizing
the economic constraints in which nursing is delivered the
commission had a remit to support nurses and midwives to
 2012 Blackwell Publishing Ltd

Comparing midwife-led and doctor-led maternity care

give the most effective, cost-effective, and efficient healthcare

to service users, families, and communities (Prime Ministers
Commission on the Future of Nursing and Midwifery
2010b). To inform the work of the UK commission we
undertook a synthesis of systematic review level evidence of
the benefits and costs of nursing and midwifery, both in the
healthcare system and wider society (Caird et al. 2010). This
article reports on a sub-set of findings from the meta-review
specific to the issue midwife-led care. Specifically, this aspect
of the review sought to understand the impact of having
midwives rather than physicians lead the care provided to
women during pregnancy and birth.

The review
Aims
Differences between midwife-led and physician-led care may
include variations in philosophy, relationship style, and care
setting. For example, some argue that philosophies of care
which privilege childbirth safety are diametrically opposed to
philosophies of care which privilege normalized or humanized
childbirth with minimal intervention, the latter being more
often associated with midwives (Sandall et al. 2010). Moreover, the relationship between pregnant women and different
care providers can also be very different, with midwives having a
continuing relationship with women throughout their pregnancy and labour (Sandall 1995). Care provided by midwives is
also more likely to be based in the home or other settings
compared with care provided by physicians which is more likely
to occur in medical settings. The aim of this meta-review was to
synthesize the available review level evidence comparing the
impact ofthese differentapproachesto maternity careprovision.

Design
Standard systematic reviews attempt to search exhaustively
for available research and to appraise and synthesize that
research in transparent and systematic ways. The ultimate
goal of many systematic reviews is to give a robust evidence
base that can inform policy, practice, and personal decisions.
To address the needs of the Commission on the Future of
Nursing this particular review needed to be both broad in its
scope, meaning a large body of literature needed to be
reviewed, and timely, meaning the evidence needed to be
synthesized relatively rapidly. Therefore, an approach was
chosen that limited the time needed to conduct the review
whilst still ensuring a robust and reliable synthesis. This
approach involved including only systematic reviews, making
this a review of reviews or a meta-review.
2377

K. Sutcliffe et al.

1992 research records

identified

1741 titles and

abstracts screened
for relevance

251 duplicates
removed

1435 records excluded

42 not in English
24 not OECD country
707 not systematic review
624 not nursing/midwifery AND
mental health/long term conditions/
role substitution

34 not comparative
4 wrong outcome

Research reports sought

for 306 included records

12 reports
unobtainable

294 reports
screened for
relevance

202 Reports excluded

0 not in English
3 not OECD country
69 not systematic review
88 not nursing/midwifery AND
mental health/long term conditions/
role substitution

33 not comparative
9 wrong outcome

46 reviews excluded

92 reports of 78 reviews
assessed for quality and
appropriate data
(5 midwifery reviews)

22 reviews do not meet quality

criteria
24 reviews found not to provide
usable data

32 systematic reviews
provide data for synthesis
(3 midwifery reviews)

Figure 1 Results of searching and screening for the broader review on nursing and midwifery 1992 Research records identified.
Reviewing systematic reviews is often a good option to
select when time is limited because (a) systematic reviews are
relatively simple to search for and rapid searches can
therefore be fairly comprehensive and (b) the findings of a
review of systematic reviews may be more comprehensive
than the findings of reviews of primary research because they
are taking advantage of the much more extensive searches
that were undertaken in the original systematic reviews.
Whilst reviews of reviews can be considered to be an
emerging methodology (Smith et al. 2011), there is a growing
body of this type of research in the midwifery, nursing, and
medical literature, for example Elliot et al. (2001) study of
2378

public health nursing; Bailey et al. (2010) study of managing

breathlessness; Swan et al. (2003) study of teenage pregnancy
and parenthood; and Smith et al. (2009) study on preventing
and treating pre-term birth.

Search methods
Search terms to identify literature on midwives were combined with terms for organization and delivery of care and
terms for systematic reviews. The specificity of the search
string was mitigated in part by applying it to a comprehensive
range of sources. The strategy was applied to three nursing
 2012 Blackwell Publishing Ltd

JAN: REVIEW PAPER

Comparing midwife-led and doctor-led maternity care

and midwifery databases including the British Nursing

Archive (BNI), the Cumulative Index to Nursing and Allied
Health Literature (CINAHL), and the Midwives Information
and Resource Service health databases (MIDIRS); to four
general health databases including PubMed, the Health
Management Information Consortium database, the Health
Technology Assessment (HTA), and the NHS EED (Economic evaluation Database); and to three databases of
systematic reviews, the Cochrane Library of Systematic
reviews, DARE (Database of abstracts of reviews of effectiveness), and DoPHER (Database of promoting health
effectiveness reviews). In addition, a search for grey literature was conducted using key nursing websites such as The
Academy of Nursing, Midwifery and Health Visiting
Research, health websites such as the Department of Health
website and research websites such as Google Scholar.
References from 56 relevant reviews or meta-reviews identified during searching were screened to identify further papers
and 38 expert informants and authors working in the field of
nursing and midwifery were contacted with a request for
published and unpublished reviews. Searches were carried out
in August and September 2009, no language or date
restrictions were employed.
Reviews were initially included if they met minimal criteria
for systematic reviews, that is, they described a search strategy
and criteria for including studies. In addition, reviews had to
include studies which compared outcomes of maternity
services in which midwives were the lead professional responsible for the planning, delivery, and organization of care, with
outcomes from services in which physicians were the lead.
Whilst this review was commissioned to inform policy
decisions in the UK, focusing on UK evidence only was felt to
be too restrictive. Included reviews were therefore limited to
those conducted in OECD countries to ensure a reasonable

level of comparability to the UK health system in terms of

available resources. The tight timelines for the review also
meant that there was no time for translation of reviews not
written in English. These inclusion criteria were initially
applied to titles and abstracts identified through searching.
Where no abstract was available from bibliographical database records, an attempt was made to retrieve the full paper.
Reviews initially included using information in the title and
abstract were subsequently re-screened using the full paper.

Search outcome
For the broader review on both nursing and midwifery 1741
titles and abstracts and 294 full reports were screened.
Figure 1 above illustrates the flow of literature through the
broader review. A number of reviews were either irretrievable
or arrived too late to be included in the meta-review. Five
reviews with a focus on midwifery were found to meet the
initial inclusion criteria described above.

Quality appraisal
Once full reports of reviews had been retrieved further
quality criteria were applied. A quality assessment tool,
adapted from that used by Elliot et al. (2001), was employed
to assess whether reviews met a minimum quality threshold
or were of high quality on four dimensions. Reviews had to
meet the minimum quality threshold for each criterion to be
included. The specific criteria are described in Table 1 below.

Data abstraction and synthesis

A framework, developed specifically for this review, was used
to extract and record information from each review about

Table 1 Quality criteria for inclusion.

Quality parameter

A review was included if it

A review was rated as high quality if it also

Searching

Searched at least two electronic

databases including at least one of the following:
Medline
Embase
Cochrane Central Register of Controlled Trials
CINAHL (Cumulative Index to Nursing and
Allied Health Literature)
British Nursing Index
Applied inclusion criteria to studies in the review and
described the detail of those criteria
Formally assessed and described the quality of included
studies or applied quality inclusion criteria (e.g. RCTs only)
Grouped together the results of two or more studies and
report the direction of the findings from this pooled group

Conducted one of the following:

hand searching
searching for grey literature
contacting experts
citation chasing

Inclusion criteria
Quality assessment
Synthesis

 2012 Blackwell Publishing Ltd

Described application by two independent reviewers

Described application by two independent reviewers
Took account of study quality when
reporting synthesized findings

2379

K. Sutcliffe et al.

such items as the topic area, care providers, age of the review,
country of origin, and findings about clinical and social
outcomes. Several reviews were excluded at the data extraction stage due to the fact that data were not presented in a
format amenable to synthesis. Quality assessment and data
extraction were conducted separately by two reviewers who
then met to compare findings. Disagreements were resolved
through discussion and the arbitration of a third party where
required. To avoid double counting of evidence, reviewers
checked to see if individual studies were included in more
than one review.
A narrative synthesis of the included reviews was conducted; evidence was grouped according to a range of
maternal and neonatal outcomes. Effect sizes were not
subject to statistical synthesis.

Results
Of the five midwifery reviews, one was excluded at the
quality appraisal stage; it provided no information on the
quality of its included studies (Muthu & Fischbacher 2004).
Another review was excluded at the data extraction stage as
data were not presented in a format amenable to synthesis:
the authors of the review acknowledge that their findings
could not be generalized beyond the individual studies

(Walsh & Downe 2004). The remaining three reviews met

the minimum quality threshold for the four quality criteria
described in Table 1, and two were rated as high quality on
each criterion (Villar et al. 2001, Hatem et al. 2008).

Characteristics of included reviews

The three included reviews, summarized in Table 2, were
meta-analyses comparing obstetric outcomes according to
type of care provider, pooling results from trials with
sufficiently similar outcomes. Brown and Grimess (1995)
meta-analysis included a comparison of nurse-midwife and
physician-led care. Hatem et al. (2008) conducted a metaanalysis of midwife-led care versus other models of care for
pregnant women. Villar et al. (2001) reviewed patterns of care
for pregnant women and included three randomized controlled trials (RCTs) that evaluated the type of care provider.
The reviews included a total of 28 studies, only one of
which (Turnbull et al. 1996) was common to more than one
review (Villar et al. 2001 and Hatem et al. 2008). To avoid
double counting of evidence in this review, the findings of this
study are only counted as part of the Hatem et al. (2008)
findings and not the Villar et al. (2001) findings. The 28 trials
involved a total of 21,105 patient participants who
were primarily randomly allocated; 13 trials did not use

Table 2 Characteristics of included reviews.

Number and type of studies

Date range of included studies
Country of included studies
(number of UK studies)
Participants in included studies

Modes of care compared

Total number and range of

patient participants in
included trials

Brown & Grimes (1995)

Hatem et al. (2008)

Villar et al. (2001)

15 studies
13 CTs, 2 RCTs
19691988
Canada and USA

11 studies
RCTs only
19892003
Australia, Canada, UK (5)

3 studies
RCTs only
19921996
Australia, UK (2)

Preponderance of studies
indicate predominantly low
risk patients
Nurse-midwife (NMs)
compared with physician-led
care. The types of
organizational providers for
NMs and physicians were the
same or similar in all studies.
Care activities of the
providers were not
consistently reported.
(p. 337)
N = 7066
Range = 582507

Pregnant women classified as

low and mixed risk of
complications
Midwife-led care (midwives
providing care antenatally,
during labour and postnatally
compared with models of
medical-led care (where an
obstetrician or family
physician is primarily
responsible for care) and
shared care

Pregnant women considered to

be at low-risk of developing
complications
Two studies compared mid
wife-led care with
physician-led care. The third
compared midwife/GP led
care with physician-led care

N = 12,276
Range = 2003510

N = 1763*
Range = 891674

*This figure is for two of the three included trials only; the details of the third trial Turnbull et al. (1996) are captured in the figures for Hatem
et al. (2008).
2380

 2012 Blackwell Publishing Ltd

JAN: REVIEW PAPER

randomization. The trials were conducted between 1969

2003. All the trials were conducted in high income countries.
Seven were conducted in the UK, others were conducted in
Australia, Canada, and the USA. It should be noted that two
reviews focus on studies of low-risk women only (Brown &
Grimes 1995 and Villar et al. 2001) as did the majority of the
11 studies in Hatem et al.s 2008 review. However, the
Hatem review also included five studies with a mixture of low
and high risk obstetric patients. Therefore, the results from
these reviews vary in their applicability to low-risk populations of pregnant women.

Outcomes
The findings are reported below according to different
outcome types. The evidence supporting each finding is
provided by citing a number of factors. First, the review or
reviews supporting any given finding are cited. Second, the
number of studies synthesized in each review to support that
finding is recorded. Third, the type of evidence is indicated,
that is to say whether the evidence comes from randomized
controlled trials (RCTs) or controlled trials (CTs). Evidence
indicates that RCTs should be regarded as the most reliable
approach for assessing the effects of interventions (Oliver
et al. 2008). Fourth, the total number of patient participants
included in the relevant trials is provided to further illustrate
the strength of the available evidence.
Several points on interpreting our findings are important
to make here. First, the statement no evidence of difference
is used to describe many of the findings; this statement
neither indicates that there is no evidence available nor does
it indicate equivalence between comparison groups. Rather,
it indicates that statistical tests designed to identify between
group differences did not detect a statistically significant
difference between nurse/midwife-delivered interventions
and those provided by others. Most studies attempt to
demonstrate a difference between groups. Demonstrating
equivalence, or no difference, is more difficult and relatively
rare as this requires a much larger study. Second, the
statement no reports of evidence indicating harm should be
understood as indicating an absence of evidence of harm,
and not conclusive evidence that these interventions are risk
free. The reviews were only able to investigate the risks and
benefits in relation with the variables measured in the
included studies. Moreover, to be able to detect the risk of
rare outcomes, such as mortality, trials need to involve
sufficiently large numbers of participants. Third, it should be
noted that outcomes which indicated reductions in physical
harm or benefits for physical and mental health were,
naturally, categorized as positive outcomes, as were positive
 2012 Blackwell Publishing Ltd

Comparing midwife-led and doctor-led maternity care

perceptions of care. In addition, reduced needs for intervention were seen as conducive to a more naturalized or
normalized birth, and therefore were also categorized as
beneficial outcomes.
Neonatal and infant outcomes
Two of the reviews examined the impact of different types of
care (physician-led vs. midwife-led) on infant outcomes; by
and large the evidence provided no indication that care led by
midwives has a different impact on neonates than care led by
physicians. With respect to infant mortality a synthesis of 10
of the 11 RCTs (n = 11,806) in the review by Hatem et al.
(2008) found no evidence of a difference between midwifeled care and physician-led care for foetal loss and neonatal
death. Similarly with respect to infant physiological outcomes, there was no evidence of a difference between providers in terms of:
admission to a neonatal intensive care unit (Hatem et al.
(2008): 10 RCTs n = 11,782),
preterm birth (Hatem et al. (2008): 5 RCTs n = 7516),
neonatal convulsions (Hatem et al. (2008): 1 RCT n = 1216),
foetal distress (Brown and Grimes (1995): 3 CTs n = 341),
and
physical condition immediately following delivery (5-minute Apgar score of seven or less) (Hatem et al. (2008): 8
RCTs n = 6780).
The only outcome for which evidence was mixed was low
birth-weight. Brown and Grimes (1995) meta-analysis of
three CTs (n = 4429) found that women receiving midwifeled care gave birth to fewer low birth-weight babies whilst
Hatem et al.s (2008) meta-analysis of five RCTs (n = 8009)
found no evidence to indicate a difference between providers.
Overall, no evidence was reported in these three reviews that
indicates any risks to neonates by having care led by
midwives as opposed to physicians.
Maternal outcomes
Across the reviews a variety of maternal outcomes were
reported; including physiological outcomes, psychosocial
outcomes, patient satisfaction and organization, and delivery
of care. All three reviews examined physiological outcomes
and these types of outcomes were the most commonly
reported.
Physiological outcomes
Midwife-led care was found to have a statistically significant
positive effect on some physiological outcomes for women
when compared with physician-led care, whilst no evidence
of a difference between providers was found for other physiological outcomes. None of the three reviews reported any
2381

K. Sutcliffe et al.

evidence of risks to womens physical health by having

midwives lead care rather than physicians.
In terms of physiological outcomes, spontaneous vaginal
birth was significantly more likely with midwife-led than
physician-led care (Brown and Grimes (1995): 5 CTs; Hatem
et al. (2008): 9 RCTs n = 11143). Also found to be significantly better for midwife-led care were reduced needs for a
number of interventions including several pain relief interventions. These included avoidance of:
vacuum extraction and/or forceps deliveries [Brown and
Grimes (1995): 5 CTs; Hatem et al. (2008): 10 RCTs n =
12,497],
episiotomies [Brown and Grimes (1995): 4 CTs; Hatem et al.
(2008): 11 RCTs n = 12,296]
regional analgesia/anaesthesia [Hatem et al. (2008): 11 RCTs
n = 11,892],
intrapartum analgesia/anaesthesia [Hatem et al. (2008):
5RCTs n = 7039],
analgesia [Brown and Grimes (1995): 3 CTs n = 292],
anaesthesia [Brown and Grimes (1995): 3 CTs n = 300]
and
opiate analgesia [Hatem et al. (2008): 9RCTs n = 10,197].
No evidence of a difference between providers was
identified for a total of ten different maternal physiological
outcomes across the three reviews. These outcomes included
caesarean section [Brown and Grimes (1995) 4 CTs; Hatem
et al. (2008): 11 RCTs n = 12,701]; antepartum haemorrhage [Hatem et al. (2008): 4 RCTs n = 3655]; postpartum
haemorrhage [Hatem et al. (2008): 7 RCTs n = 8454];
induction of labour [Hatem et al. (2008): 10 RCTs
n = 11,711]; augmentation/oxytocin during labour [Hatem
et al. (2008): 10 RCTs n = 11,709]; mean length of labour
[Hatem et al. (2008): 2 RCTs n = 1614]; manual removal of
the placenta [Brown and Grimes (1995): 3CTs n = 306];

use of intravenous fluids [Brown and Grimes (1995): 3

CTs n = 236]; anaemia [Villar et al. (2001): 1 RCT n = 1674]
and malpresentation [Villar et al. (2001): 1RCT n =
1674].
Findings for pregnancy induced hypertension, use of
amniotomy and perineal injuries were, however, mixed. Of
the two trials with an appropriate comparator in Villar et al.
(2001), one small trial (n = 89) was not statistically significant. However, a much larger trial (n = 1674) demonstrated
a statistically significant positive effect of midwife-led care on
pregnancy induced hypertension compared with standard
care. For amniotomy, Brown and Grimes (1995) metaanalysis of three CTs (n = 292) saw fewer amniotomies
amongst women receiving midwife-led care, whilst Hatem
et al.s (2008) meta-analysis of three RCTs (n = 1543) found
no evidence of a difference between providers. Perineal
lacerations were more likely in women receiving midwife-led
care in Brown and Grimess (1995) meta-analysis of three
CTs (n = 339). However, another two meta-analyses in
Hatem et al. (2008) found no evidence of a difference
between providers for perineal laceration requiring suturing
(7 RCTs n = 9439) and women with an intact perineum (8
RCTs n = 9706). Table 3, below identifies the full range of
physiological outcomes for women and the direction of
evidence for these outcomes.
Other maternal outcomes
Midwife-led care was also found to have positive benefits for
a number of other non-physical outcomes. It should be noted,
however, that whilst evidence is available from a large
number of trials for many of the physical outcomes, there is
less extensive evidence for many of the non-physical outcomes evidence; for some outcomes evidence comes from
only one or two trials.

Table 3 Maternal physiological outcomes by direction of evidence.

Evidence of improved outcomes
with midwife-led care

No evidence of a difference
between care providers

Outcomes with mixed evidence

Avoidance of vacuum extraction and/or

forceps deliveries
Avoidance of episiotomies
Avoidance of regional analgesia/anaesthesia
Avoidance of intrapartum analgesia/anaesthesia
Avoidance of both analgesia and anaesthesia
Avoidance of opiate analgesia

Caesarean section

Pregnancy induced hypertension

Antepartum haemorrhage
Postpartum haemorrhage
Induction of labour
Augmentation/oxytocin during labour
Mean length of labour
Manual removal of the placenta
Use of intravenous fluids
Anaemia
Malpresentation

Use of amniotomy
Perineal injuries

2382

 2012 Blackwell Publishing Ltd

JAN: REVIEW PAPER

In terms of patient satisfaction, Villar et al. (2001) found

that women were more satisfied with their experience of
getting questions answered (1 RCT n = 89) and had a higher
confidence in midwife-led care (1 RCT n = 89). Hatem et al.
(2008) found women receiving midwife-led care had a higher
perception of control during labour (1 RCT n = 471).
In terms of organization and delivery of care, women
receiving midwife-led care had less foetal monitoring [Brown
and Grimes (1995): 4CTs n = 409], and were likely to
experience attendance at birth by a known midwife [Hatem
et al. (2008): 6 RCTs n = 5225]. Women receiving midwifeled care were also significantly less likely to experience
antenatal hospitalization [Hatem et al. (2008): 5 RCTs
n = 4337] and breastfeeding initiation was significantly more
likely in women assigned to midwife-led groups [Hatem et al.
(2008): 1 RCT n = 405]. However, there was no evidence of
a difference between providers in terms of postpartum
depression [Hatem et al. (2008): 1 RCT n = 1213], mean
number of antenatal visits [Hatem et al. (2008): 1 RCT
n = 405], or duration of postnatal stay [Hatem et al. (2008):
2 RCTs n = 1944].

Discussion
This systematic review of reviews indicates that midwife-led
care for low-risk women has the capacity to meet the needs of
the mothers and infants it serves, and that for some outcomes
it may serve their needs better than physician-led care. These
findings are corroborated by other research which calls for
midwife-led care to be used much more widely. A 2011
scoping review on safety in maternity services from the Kings
Fund concludes that deploying midwife-led models of care
across the service for low- and medium-risk women, would
improve a range of outcomes, be a more cost-effective model
of service delivery and release obstetricians to focus on
women with more complex needs (Sandall et al. 2011).
Similarly the Birthplace in England study (Hollowell 2011)
which collected data on almost 65,000 low-risk births
between 20082010 concluded that midwifery units could
see an increase in their number in future as they appear to be
safe, offer benefits to both mothers and babies and lower
costs.

Strengths and limitations

Conducting a review of reviews is a very complex and
challenging undertaking. This is primarily due to the fact that
as reviews attempt to summarize and synthesize a great deal
of research evidence in to one report, detail about the
included primary studies is necessarily limited. This was
 2012 Blackwell Publishing Ltd

Comparing midwife-led and doctor-led maternity care

compounded in this instance by the fact that the included

reviews had slightly different objectives to those of the metareview. For example, of the three included reviews, one
compared physicians with both nurse midwives and nurse
practitioners (Brown & Grimes 1995), another compared
midwife-led care with other models of care, not necessarily
physicians (Hatem et al. 2008), and the third focused on
patterns of care rather than explicitly comparing different
providers (Villar et al. 2001). We were, therefore, reliant
upon review authors to give sufficient detail on the studies
they included about providers of interventions and forced to
exclude reviews where this had not been adequately reported.
Due to these limitations, absence of evidence, especially of
evidence of harm, does not equate to evidence of absence
(Altman & Bland 1995), thus interpretation needs to be
cautious.
Another of the many challenges of conducting a review of
reviews is that different reviews often include the same
studies. To avoid bias from double counting of findings it was
necessary to check for overlap between reviews to ensure that
each studys findings were only represented at one point in
the review. In this meta-review only one study was found to
be common to more than one review. However, by only
counting this study once [it was only counted in the findings
of the Hatem et al. (2008) review and not the Villar et al.
(2001) review] difficulties arose in the interpretation of metaanalyses in the Villar review that included this particular
study.
Assessing cost-effectiveness of interventions is also a
challenge for this type of review. Although authors of
primary studies often report costs or cost-effectiveness, it is
rarely the case that they give data in a format which can be
used in systematic reviews. Therefore, the presence of reliable
cost-effectiveness data in reviews of reviews, including this
one, are rare. Both Dierick-van Daele et al. (2008) and
Brown and Grimes (1995) have discussed the challenges in
obtaining cost-effectiveness data for systematic review. The
former of these studies gives a comprehensive description of
the difficulties experienced in obtaining data for a review of
economic evaluations of substitution of skills between health
professionals.
Despite these challenges this meta-review was able to give a
clear and robust synthesis of the review level evidence about
the impact of having care provided to pregnant and birthing
women led by midwives as opposed to physicians. However,
by selecting only research evidence that has already been
reviewed, reviews of reviews are not able to access the most
recent and up-to-date evidence. The searches for this metareview were carried out in 2009, yet the most recent article
included in any of the three reviews was published in 2003
2383

K. Sutcliffe et al.

What is already known about this topic

Increased birth rates, and increasing numbers of
complicated births are some of the challenges faced by
the maternity workforce in the UK.
These factors are compounded by shortages in supply
and increased demand for midwives both in the UK and
in many health systems across the world and by the
current global economic crisis.
Initiatives in countries such as the UK and the US have
been developed to support nurses and midwives to give
the most effective, cost-effective and efficient healthcare
to service users, families, and communities.

What this paper adds

A rigorous and complex meta-review approach provides
a robust evidence-base on the impact of having maternity
care led by midwives as opposed to physicians.
Midwife-led care for low-risk women compared with
physician-led care appears to improve a range of
maternal outcomes, to reduce the number of procedures
in labour, and increase satisfaction with care.
The pooled evidence provides no indication that
maternity care led by midwives poses any greater risks
than care led by physicians in predominantly low risk
women.

Implications for practice and/or policy

Research to establish the cost-effectiveness of replacing
physician-led care with midwife-led care and more
research on non-physical outcomes such as satisfaction
and breastfeeding initiation is needed.
More research is also needed to explore variation in
midwife-led approaches to care, to enable an
understanding of what are the effective mechanisms in
midwife-led care.
Though evidence about cost effectiveness is unavailable,
midwife-led maternity care should be considered by
practitioners and policy makers seeking to improve the
health and satisfaction needs of women and infants.

(Hatem et al. 2008). The Cochrane review by Villar et al.

(2001) has been updated to identify more recent studies
(Dowswell et al. 2010). However, no further findings on the
impact of midwife-led care are available as the updated
review focuses exclusively on the impact of reduced schedules
of antenatal visits and no longer includes studies looking at
2384

alternative care providers. Nonetheless, as noted above, more

recent evidence appears to corroborate the findings of this
review (Hollowell 2011, Sandall et al. 2011).

Conclusion
Implications for research and practice
The review level evidence synthesized here indicates that
women who receive care led by midwives rather than
physicians may access a range of physical and other benefits.
Furthermore, there is no indication in the reported evidence
of any risks associated with opting for midwife-led care for
(predominantly) low-risk women. Whilst it was not possible
to establish the cost-effectiveness of replacing physician-led
care with midwife-led care, practitioners and policy makers
seeking to address the health and satisfaction needs of women
and infants receiving maternity services may wish to consider
this option.
However, as noted at the beginning of this article developing an efficient service, and an effective service, is necessary
for addressing the current economic crisis (WHO 2009) and
nursing shortages across the globe (Buchanan & Aiken
2008). Therefore, whilst indicators of the lower costs of
midwife-led care are emerging (Hollowell 2011, Sandall et al.
2011) robust review level evidence of the cost effectiveness of
adopting such approaches in the UK is vital. Other research
still needed in this area would be that which extends the
evidence base for a number of outcomes, particularly nonphysical outcomes such as satisfaction and breastfeeding
initiation. Qualitative research with women experiencing
maternity care led by different providers could also help to
identify the relative importance for them of the different
outcomes measured in the reviewed literature, and could
identify further salient outcomes for measurement in future
trials.
More research is also needed to explore the variation in
approaches in midwife-led care. As described above the
dimensions on which different models of care may differ are
numerous. Therefore, future research in this area would
benefit from exploring the theoretical underpinnings of these
complex interventions and their associations with processes
and outcomes. That is to say, it will be beneficial to
understand more precisely what the effective mechanisms in
midwife-led care are. The rigorous and complex approach
used to pool review level evidence on the effectiveness of
midwife-led maternity care, compared with physician-led
care, was able to give a robust evidence-base showing that it
may have some benefits for women and infants. Moreover,
the pooled evidence provides no indication that maternity
 2012 Blackwell Publishing Ltd

JAN: REVIEW PAPER

care led by midwives poses any greater risks than care led by
physicians in predominantly low-risk women.

Acknowledgements
We are grateful to Ginny Brunton and Mark Newman, from the
EPPI-Centre, who gave the project valuable advice throughout.

Funding
This work was undertaken by the EPPI-Centre, which
received funding from the Department of Health. The views
expressed in the publication are those of the authors and not
necessarily those of the Department of Health.

Conflict of interest
No conflict of interest has been declared by the authors.

Author contributions
All authors meet at least one of the following criteria
(recommended by the ICMJE: http://www.icmje.org/ethical_1author.html) and have agreed on the final version:
substantial contributions to conception and design, acquisition of data, or analysis and interpretation of data;
drafting the article or revising it critically for important
intellectual content.

Supporting Information Online

There is no Supporting Information associated with this
article.

References
Altman J. & Bland D. (1995) Statistics notes: absence of evidence is
not evidence of absence. British Medical Journal 311, 485.
Bailey C., Wagland R., Dabbour R., Caress A., Smith J. & Molassiotis A. (2010) An integrative review of systematic reviews related
to the management of breathlessness in respiratory illnesses. BMC
Pulmonary Medicine 10(63), 113.
Brown S. & Grimes D. (1995) A meta-analysis of nurse practitioners
and nurse midwives in primary care (structured abstract). Nursing
Research 44, 332339.
Buchanan J. & Aiken L. (2008) Solving nursing shortages: a common
priority. Journal of Clinical Nursing 17, 32623268.
Caird J., Rees R., Kavanagh J., Sutcliffe K., Oliver K., Dickson K.,
Woodman J., Barnett-Page E. & Thomas J. (2010) The Socioeconomic Value of Nursing and Midwifery: A Rapid Systematic
Review of Reviews. EPPI Centre, Social Science Research Unit,
Institute of Education, University of London, London.

 2012 Blackwell Publishing Ltd

Comparing midwife-led and doctor-led maternity care

Dierick-van Daele A., Spreeuwenberg C., Derckx E., Metsemakers J.
& Vrijhoef B. (2008) Critical appraisal of the literature on economic evaluations of substitution of skills between professionals: a
systematic literature review. Journal of Evaluation in Clinical
Practice 14, 481492.
Dowswell T., Carroli G., Duley L., Gates S., Gulmezoglu A.M.,
Khan-Neelofur D. & Piaggio G.G. (2010). Alternative versus
standard packages of antenatal care for low-risk pregnancy.
Cochrane Database of Systematic Reviews 10, CD000934 (DOI:
10.1002/14651858.CD000934. pub2).
Elliot L., Crombi I., Irvine L., Cantrell J. & Taylor J. (2001) The
Effectiveness of Public Health Nursing: A Review of Systematic
Reviews. The Stationery Office, London.
Hatem M., Sandall J., Devane D., Soltani H. & Gates S. (2008)
Midwife-led versus other models of care for childbearing women.
Cochrane Database of Systematic Reviews 4, CD004667 (DOI:
10.1002/14651858.CD004667.pub2).
Hollowell J. (2011) Birthplace programme overview: background,
component studies and summary of findings. Birthplace in England
Research Programme. Final report part 1. NIHR Service Delivery
and Organisation programme.
International Council of Nurses, Florence Nightingale International
Foundation (2006) The Global Nursing Shortage: Priority Areas
for Intervention. International Council of Nurses and Florence
Nightingale International Foundation, Geneva.
Muthu V. & Fischbacher C. (2004) Free-standing midwife-led
maternity units: a safe and effective alternative to hospital delivery
for low-risk women? Evidence-Based Healthcare and Public
Health 8, 325331.
Oliver S., Bagnall A., Thomas J., Shepherd J., Sowden A., White I.,
Dinnes J., Rees R., Colquitt J., Oliver K. & Garrett Z. (2008)
RCTs for Policy Interventions? A Review of Reviews and MetaRegression. University of Birmingham, Birmingham.
Prime Ministers Commission on the Future of Nursing and Midwifery
(2010a) Prime Ministers Commission on the Future of Nursing and
Midwifery: Summary Vision. Retrieved from http://cnm.independent.gov.uk/second-phase/summary-vision/ on 18 January 2010.
Prime Ministers Commission on the Future of Nursing and Midwifery (2010b) Front Line Care: the Future of Nursing and Midwifery in England. Produced by COI for the Prime Ministers
Commission on the Future of Nursing and Midwifery in England.
First published March 2010, London.
Royal College of Midwives (2011) State of Maternity Services Report
2011. Royal College of Midwives, London. Retrieved from http://
www.rcm.org.uk/college/policy-practice/government-policy/stateof-maternity-services/ on 31 January 2011.
Sandall J. (1995) Choice, continuity and control: changing midwifery,
towards a sociological perspective. Midwifery 11, 201209.
Sandall J., Devane D., Soltani H., Hatem M. & Gates S. (2010)
Improving quality and safety in maternity care: the contribution of
midwife-led care. Journal of Midwifery & Womens Health 55,
255261.
Sandall J., Homer C., Sadler E., Rudisill C., Bourgeault I., Bewley S.,
Nelson P., Cowie L., Cooper C. & Curry N. (2011). Staffing in
Maternity Units: Getting the Right People in the Right Place at the
Right Time. The Kings Fund, London. Retrieved from http://
www.kingsfund.org.uk/publications/maternity_unit_staff.html on
31 January 2012.

2385

K. Sutcliffe et al.
Smith V., Devane D., Begley C., Clarke M. & Higgins S. (2009) A
systematic review and quality assessment of systematic reviews of
randomised trials of interventions for preventing and treating
preterm birth. European Journal of Obstetrics & Gynecology and
Reproductive Biology 142, 311.
Smith V., Devane D., Begley C. & Clarke M. (2011) Methodology in
conducting a systematic review of systematic reviews of healthcare
interventions. BMC Medical Research Methodology 11, 15.
Swan C., Bowe K., McCormick G. & Kosmin M. (2003) Teenage
Pregnancy and Parenthood: A Review of Reviews, Evidence Briefing. National Health Service Health Development Agency, London.
Turnbull D., Holmes A., Shields N., Cheyne H., Twaddle S., Johnstone I., Geer I., McIlwaine G. & Lunan C. (1996) Randomised,

controlled trial of efficacy of midwife managed care. Lancet 348,

213218.
Villar J., Carroli G., Khan-Neelofur D., Piaggio G. & Gulmezoglu M.
(2001) Patterns of routine antenatal care for low-risk pregnancy.
Cochrane Database of Systematic Reviews 2001(4): CD000934.
DOI: 10.1002/14651858.CD000934.
Walsh D. & Downe S. (2004) Outcomes of free-standing, midwifeled birth centers: a structured review. Birth 31, 222229.
World Health Organization (2009) The Financial Crisis and Global
Health: Report of a High-Level Consultation. World Health
Organization, Geneva 19 January 2009. World Health Organization, Geneva.

The Journal of Advanced Nursing (JAN) is an international, peer-reviewed, scientific journal. JAN contributes to the advancement of
evidence-based nursing, midwifery and health care by disseminating high quality research and scholarship of contemporary relevance
and with potential to advance knowledge for practice, education, management or policy. JAN publishes research reviews, original
research reports and methodological and theoretical papers.
For further information, please visit JAN on the Wiley Online Library website: www.wileyonlinelibrary.com/journal/jan
Reasons to publish your work in JAN:
High-impact forum: the worlds most cited nursing journal and with an Impact Factor of 1477 ranked 11th of 95 in the 2011 ISI
Journal Citation Reports (Social Science Nursing).
Most read nursing journal in the world: over 3 million articles downloaded online per year and accessible in over 10,000 libraries
worldwide (including over 3,500 in developing countries with free or low cost access).
Fast and easy online submission: online submission at http://mc.manuscriptcentral.com/jan.
Positive publishing experience: rapid double-blind peer review with constructive feedback.
Rapid online publication in five weeks: average time from final manuscript arriving in production to online publication.
Online Open: the option to pay to make your article freely and openly accessible to non-subscribers upon publication on Wiley
Online Library, as well as the option to deposit the article in your own or your funding agencys preferred archive (e.g. PubMed).

2386

 2012 Blackwell Publishing Ltd

This document is a scanned copy of a printed document. No warranty is given about the accuracy of the copy.
Users should refer to the original published version of the material.