cost-effectiveness of continuity of midwifery care for ...... · centres and home birth), as...

Donnellan-Fernandez et al. Health Economics Review (2018) 8:32 https://doi.org/10.1186/s13561-018-0217-3

REVIEW Open Access

Cost-effectiveness of continuity ofmidwifery care for women with complexpregnancy: a structured review of theliterature

Roslyn E. Donnellan-Fernandez1* , Debra K. Creedy1 and Emily J. Callander2

Abstract

Background: Critical evaluation of the cost-effectiveness and clinical effectiveness of continuity of midwifery caremodels for women experiencing complex pregnancy is an important consideration in the review and reform ofmaternity services. Most studies either focus on women who experience healthy pregnancy or mixed risk samples.These results may not be generalised across the childbearing continuum to women with risk factors. This reviewcritically evaluates studies that measure the cost of care for women with complex pregnancies, with a focus onmethod and quality.

Aims / objectives: To critically appraise and summarise the evidence relating to the combined cost-effectiveness,resource use and clinical effectiveness of midwifery continuity models for women who experience complexpregnancies and their babies in developed countries.

Design: Structured review of the literature utilising a matrix method to critique the methods and quality of studies.

Method: A search of Medline, CINAHL, MIDIRS, DARE, EMBASE, OVID, PubMed, ProQuest, Informit, Science Direct,Cochrane Library, NHS Economic Evaluation Database (NHSEED) for the years 1994 – 2018 was conducted.

Results: Nine articles met the inclusion criteria. The review identified four areas of economic evaluation that relatedto women who experienced complex pregnancy and continuity of midwifery care. (1) cost and clinical effectivenesscomparisons between continuity of midwifery care versus obstetric-led units; (2) cost of continuity of midwiferycare and/or team midwifery compared to Standard Care; (3) cost-effectiveness of continuity of midwifery care forAustralian Aboriginal women versus standard care; (4) patterns of antenatal care for women of high obstetric riskand comparative provider cost.Cost savings specific to women from high risk samples who received continuity of midwifery care compared withobstetric-led standard care was stated for only one study in the review. Kenny et al. 1994 identified cost savings of AUS$29 in the antenatal period for women who received the midwifery team model from a stratified sub-set of high-riskpregnant woman within a mixed risk sample of 446 women. One systematic review relevant to the UK context, Ryanet al. (2013), applied sensitivity analysis to include women of all risk categories. Where risk ratio for overall fetal/neonataldeath was systematically varied based on the 95% confidence interval of 0.79 to 1.09 from pooled studies, theaggregate annual net monetary benefit for continuity of midwifery care ranged extremely widely from an estimatedgain of £472 million to a loss of £202 million. Net health benefit ranged from an annual gain of 15 723 QALYs to a loss(Continued on next page)

* Correspondence: [email protected] Maternity Care Collaborative, Nursing and Midwifery, GriffithUniversity, Logan campus, University Drive, Meadowbrook, Queensland 4131,AustraliaFull list of author information is available at the end of the article

© The Author(s). 2018 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, andreproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link tothe Creative Commons license, and indicate if changes were made.

http://crossmark.crossref.org/dialog/?doi=10.1186/s13561-018-0217-3&domain=pdfhttp://orcid.org/0000-0002-7011-8923mailto:[email protected]://creativecommons.org/licenses/by/4.0/

Donnellan-Fernandez et al. Health Economics Review (2018) 8:32 Page 2 of 16

(Continued from previous page)

of 6 738 QALYs. All other studies in this review reported cost savings narratively or within mixed risk samples where riskstratification was not clearly stated or related to the midwifery team model only.

Conclusions: Studies that measure the cost of continuity of midwifery care for women with complex pregnancyacross the childbearing continuum are limited and apply inconsistent methods of economic evaluation. The cost andoutcomes of implementing continuity of midwifery care for women with complex pregnancy is an important issuethat requires further investigation. Robust cost-effectiveness evidence is essential to inform decision makers, toimplement sustainable systems change in comparative maternity models for pregnant women at risk and to addresshealth inequity.

Keywords: Cost effectiveness, Midwifery care, Complex pregnancy, Continuity of midwifery care, Maternity models,Models of care, Health equity, Structured review

Introduction and backgroundReview and transformation of maternity service modelshave been on the policy agenda of the Australian Gov-ernment for the past decade [1, 2]. An important policygoal is to expand women’s access to midwifery caseloadcontinuity of care in both the public and private healthsectors [3, 4]. Continuity of midwifery care is where anamed midwife provides full antenatal, intrapartum andpostnatal care for a woman. The midwife provides phys-ical, emotional and social support, flexible individualisedcare and robust multi-agency liaison. This enhances highquality perinatal care for mother and baby and strongworking relationships with professionals [5]. Currently,only a small proportion of women are able to accesscontinuity of midwifery care during their pregnancy [6].Internationally, and in Australia, strong clinical and cost

evidence already exist to support systemic implementationof continuity of midwifery care for women with healthypregnancies [7–10]. However, in Australia the number ofwomen who experience complex pregnancy is increasing[11]. In this review, complex pregnancy is defined as iden-tified risk factors that place mother and/or baby at in-creased risk for adverse events. These can includebiomedical and/or psychosocial risks, as identified by thewoman and her care provider. Risk factors can be presentat the start of pregnancy or arise at any time during thecourse of childbearing [12] . Evidence also shows signifi-cant inequity, poorer outcomes and associated increasedhealthcare costs for women who experience complexpregnancy. Outcomes for these women and their babiesmay potentially improve by increasing public health accessto continuity of midwifery care models [13–17].No previous systematic reviews have focused on women

with complex pregnancies. To date, most systematic re-views of midwifery care have been conducted in theUnited Kingdom (UK) for low risk pregnancies. These re-views provide strong evidence for clinical and cost effect-iveness of continuity of midwifery care (including birthingcentres and home birth), as compared to obstetric - led

units, but discrete economic analysis of outcomes and costfor pregnant women with risk factors were not included[9, 18–23]. Further, maternity services in many countriesare not organised in the same configuration as in the UK,where clear delineation between continuity of midwiferycare and obstetric-led units are an established feature.Econometric models that applied productivity / effi-

ciency frontiers and standard international resource in-gredient approaches to develop predictive cost models,or other methods, for example, Net Benefit, were simi-larly limited [24, 25]. The implications of these studies isconsidered in the discussion in relation to costs of carefor women who experience high risk pregnancy along-side clinical health outcomes in the midwifery continuityof care studies considered in this review. The lack ofrigorous economic evaluation of different models of ma-ternity care for women at high risk of complications hasbeen emphasised in an integrated review examining costdata in relation to care provided in birth centres and athome with midwives [18]. This remains the case andprovides a strong justification for the present reviewgiven the current evidence that show increasing rates ofpregnancy complication and multiple complex maternalco-morbidity in Australia and elsewhere.Capacity to improve maternity services to women with

complicated pregnancy continues to pose a major chal-lenge for the Australian health system [26–28]. This isparticularly critical in rural and regional areas of the coun-try where service options are limited and outcomes aresignificantly poorer than they are for women and babies inmetropolitan areas [11, 29–31]. Critical evaluation of inte-grated evidence on the cost-effectiveness, resource useand clinical effectiveness of continuity of midwifery carefor women who experience complexity therefore is an im-portant consideration in quality review of maternity care.

Aims and objectivesThe aim of this review is to critically appraise availableliterature and summarise the evidence related to cost,


resource use, and clinical outcomes of care for womenwith complex pregnancies who received care in a con-tinuity of midwifery care model compared with othermaternity models.This structured, integrated review will examine the

available evidence for cost-effectiveness of antenatal,intrapartum and postnatal care in continuity of midwif-ery care. It will critically evaluate the methods of thesestudies. This includes their capacity to support publichealth policy through expanded implementation and ac-cess to continuity of midwifery care for women who ex-perience complications of pregnancy and childbearingand their babies.

MethodThis review used a stepped structured approach to docu-menting the search strategy [32]. The Matrix Methodwas then applied to ensure a systematic framework forarticle collection, organisation and analysis [33, 34]. Ap-plication of PRISMA guidelines strengthened credibilityand transparency of the reporting and assessmentprocess [35]. Use of eight quality appraisal questionsfrom the recommended checklist for appraising the costsand benefits of economic evaluation studies enabled ro-bust synthesis of the results of studies [36].

SearchTable 1 and Fig. 1 provide a summary of search details.

Table 1 Databases searched

Databases

Medline, CINAHL, MIDIRS, DARE, EMBASE, OVID, PubMed, ProQuest,Informit Science Direct, Cochrane Library, NHSEED

Published between 1994 and 2018

English language publications only

Article contained key search words or combined search terms:midwifery, midwife-led units, nurse-midwifery, birth centers, cost,cost-effectiveness, economic evaluation, economic outcomes,pregnancy risk classification, maternal outcomes, neonataloutcomes, clinical outcomes, maternity services

Primary research article or Systematic Review/Meta-analysis orIntegrative Review

Economic analysis secondary to RCT accepted

Peer-Reviewed Journals

Population sample of childbearing women and/or their babieswhere risk classification profile defined and/or includes womanwith high risk or complex pregnancy

Measurement of at least one economic outcome measurecombined with clinical and/or other outcome measures, inmidwifery care units or integrated midwifery continuitymodels that included antenatal, birthing and postnatalservices, compared to other maternity service models

Economic perspective is funder/health service

Inclusion criteriaInclusion criteria were primary research articles publishedin English language, peer-reviewed journals between theyears 1994–2018. The 24-year time - frame marked theemergence of studies on the cost-effectiveness of continu-ity of midwifery care, including the first Australian studies[10, 37, 38]. Non-English language papers were excluded,as were those that focused exclusively on low resourcecountries.

ResultsThe classification of included studies within the evidencehierarchy is documented in Table 2.

Appraisal of studiesThis review identified three systematic reviews that exam-ined the cost and clinical effectiveness of continuity of mid-wifery care and obstetric-led maternity models. All three ofthese reviews were undertaken in the UK (Table 3).A summary of the six primary studies included in this

review, including study design is provided in Table 4. Allstudies were completed in Australia.Results of these primary studies are reported in table 5.

Quality of studiesEconomic evaluations undertaken alongside randomisedcontrolled trials (RCT) constituted the most robust evi-dence for economic analysis of continuity of midwiferycare models available. Importantly, four RCTs includedin the systematic review by Sandall et al. [9] wereconducted in Australia and included women of mixedpregnancy risk classification. The four AustralianRCTs, were the only studies that examined cost re-sults for continuity of midwifery care models that alsoincluded women with identified pregnancy risk factors[10, 37–39]. Two of the RCT studies identified theireconomic evaluations as cost analyses, Homer et al.[39], Kenny et al. [37]. One other was identified as acost-effectiveness study on the NHS EED data base(Rowley et al. [38] and the remaining cost conse-quences analyses study, Tracy et al. [10], calculatedper woman cost of care based on DRGs as well asdirect and indirect costs for resource use.The quality of cost, resource use and clinical effect-

iveness evidence in the primary studies included inthis review therefore is high as they include mainlyRCT evidence and also incorporated results fromLevels III & IV of the evidence hierarchy. However,of the four RCT studies, three involved team midwif-ery models, as contrasted with continuity of care witha named midwife. In a team midwifery model a smallgroup of midwives (up to 6 and no more than 8) pro-vide care for identified women and the degree of con-tinuity is not as high as in continuity of care with a

Fig. 1 Flow Chart of study inclusion


named midwife. Other study designs considered costevidence with varying levels of quality. Based on theNHMRC evidence hierarchy, the studies, in order ofdecreasing quality included cost analysis [40], andcost consequences analyses based on retrospective re-cords audit [41].While a number of studies were identified that used

internationally validated ratios to model predictive costsfor mode of birth and other interventions, these studieswere excluded as they were not directly applied to con-tinuity of midwifery care or women with identified preg-nancy risk factors [42–45].Four areas of economic evaluation that relate to

women who experienced complex pregnancy were iden-tified from the review:

Table 2 Summary of Included Studies

Evidence hierarchy level Inclu

Level I:Systematic Review

Dev

Level IIRandomised Controlled Trial with Economic Evaluation

Hom

Levels III and IVQuasi-experimental Cost Studies (cohort, cross-sectional, casecontrol, non-randomised prospective, retrospective audit)

Gao

Econometric Studies – predictive cost, productivity, resourcemodels using datasets

No

a6 of 15 studies included in Sandall et al. 2016 review included cost/economic analin this review

1. Comparisons of midwife-led versus obstetricconsultant-led units for cost and clinical effectiveness

Two studies had data relevant to this theme. Whilecost models were based on trials that recruited womenwith low pregnancy risk, sensitivity analysis modelledcost for women of mixed pregnancy risk classificationbased on UK population data. In the two systematic re-views from the UK where continuity of midwifery careand obstetric consultant-led maternity models are com-mon, an estimated mean cost saving for each eligiblewoman of £12.38 was found in the continuity of midwif-ery care model overall. This provided aggregate healthsavings of £1.16 million per year for the health system ifonly half of all eligible women received continuity of

ded studies

ane et al., 2010, Ryan et al., 2013, Sandall et al., 2016a

er et al., 2001a,b, Kenny et al., 1994, Rowley et al., 1995, Tracy et al., 2013

et al., 2014, Jan et al., 2004

studies relevant to complex needs

yses – 4 of these cost studies included woman of mixed risk and were included

Table

3Summaryof

ThreeSystem

aticReview

sa

Aim

ofStud

ySample/Setting

Design

Major

CostFind

ings

Health

Outcomes

Streng

ths/Limitatio

ns

1.Sand

alletal.(2016)

Com

pare

effectsof

midwife-

ledcontinuity

mod

elswith

othe

rmod

elsfor

childbe

aringwom

enandtheirinfants

Prim

aryou

tcom

esantenatal,

birth&im

med

iate

postpartum

outcom

esSecond

aryou

tcom

esbirth

interventio

n,morbidity,som

easpectsof

resource

use&

costUnitedKing

dom

√

Includ

ed:15RC

Ts17

674

wom

en(Canada,Ireland

,Australia,U

K)Exclud

ed:22

stud

ies

Only6of

the15

RCTs

measuredcostof

mod

el;

only4of

the6RC

Tsthat

measuredcostsinclud

ed“m

ixed

risk”

preg

nant

wom

en/highriskpreg

nancy:

Kenn

y1994

Rowley1995

Hom

er2001

Tracy2013

System

aticreview

Cochrane

Preg

nancy&Childbirth

Group

TrialsRegister

+referencelistsof

retrieved

articles.

Selectioncriteria:p

ublishe

dandun

publishe

dtrials,

preg

nant

wom

enrand

omly

allocatedto

midwife-led

continuity

mod

elsof

care

orothe

rmod

elsof

care

for

preg

nancy&birth

Costtren

drepo

rted

narrativelyas

RCTcost

metho

dvaried,

e.g.

cost

analysis;C

EA;orno

tstated

Tren

dto

costsaving

effect

inmidwife-ledcontinuity

Costsaving

sintrapartum

care

–allstudies

Anten

atal:varied

Postnatal:1stud

yhigh

ercost/1

stud

yno

difference

Prim

ary0utcom

ein

midwife

–led

mod

els

(RR)

(CI)

↓region

alanalge

sia

(0.85,0.78

–0.92)

↓instrumen

talb

irth

(0.90,0.83

–0.97)

↓pre-term

<37

wk

(0.76,0.64

–0.91)

↓fetalloss<24

wk

(0.84,0.71

–0.99)

↑spon

tane

ousvaginalb

irth

(1.05,1.03

–1.07)

NodifferenceCSor

intact

perin

eum

Second

ary0utcom

emidwife

–ledmod

els:

↓am

niotom

y;↓ep

isiotomy;

↓fetal

loss

<24

wks;

Nolabo

uranalge

sia;long

erlabo

ur(M

D)0.50

hrs,

Nodifferencefor:

fetalloss>24

wks;

labo

urindu

ction;A/N

admission

;A/N

haem

orrhage;augm

ent

labo

ur;PPH

;low

birthw

eigh

t;5

min

Apg

ar<7;SC

BUadmission

;initiatebreastfeed

ing

Timeho

rizon

:RC

T(costinclud

ed)

1994

–2013

Wom

enreceivingmidwife

care

less

likelyto

have

epidural,

episiotomies,instrumen

talb

irth.

Spon

tane

ousvaginalb

irthrate

increased.

CSrate

nodifference.

Wom

enless

likelyto

have

pre-term

birth,lower

risk

oflosing

babies

<24wks,

Morelikelyto

becaredforin

labo

urby

aknow

nmidwife.

Noadverseeffectscompared

with

othe

rmod

els.

Con

clusion:mostwom

enshou

ldbe

offeredmidwife-led

continuity

ofcare

BUT

Eviden

cemay

notapplyto

wom

enwith

serio

uspreg

nancy

orhe

alth

complications

asthese

wom

enwereno

tspecifically

includ

edin

allstudies

/analysis

forclinicaleffectiven

essno

tstratified

Limitedeviden

ceCEA

for

wom

enwith

complex

preg

nancy

Com

bine

dresults:low

and

mixed

riskpreg

nant

wom

en4stud

iesused

different

econ

omicevaluatio

nmetho

ds-:narrativerepo

rtas

costassessmen

tinconsistent

Strong

eviden

cecost

improved

inmidwifery

mod

elsforlow

riskwith

,redu

cedinterven

tion+

increasedsatisfaction.

Mixed

riskstud

ies-

‘interpretwith

caution’

2.Ryan

etal.(2013)Analysis

ofeviden

ceon

cost–

effectiven

essof

midwife-led

care

comparedwith

consultant

–led

care

inUK

settings.

Estim

atepo

tentialcost

saving

sto

accrue

from

expansionof

midwife

–ledcare

inUK

UsedSection3CEof

Devaneet

al.2010SR

UnitedKing

dom

√

Econ

omicsynthe

sisof

3RC

Tsevaluatedagainst

guidelines

forecon

omic

review

Drummon

dand

Jefferson

(1996)

5796

wom

enHun

dley

1995

Youn

g1997

Begley

2009

Exclud

ed:

Flint1989

(sub

-group

costing49

of1001

wom

enon

ly)

System

aticreview

12electron

icdatabasesfor

costmidwife

ledmod

els:

CochraneMetho

dology

Register

NICEmetho

ds+multip

le1-way

sensitivity

analysisfor

econ

omicsynthe

sisof

costs

used

3RC

Tsappliedto

8scen

arios

CEmeasure

used

Increm

entalN

etBene

fit(IN

B):expressed

asNet

Mon

etaryBene

fit(NMB)

–£

value,andNet

Health

Meancostsaving

£12.38

per

wom

anmidwife

led(M

L)care

Expansionof

MLcare

to50%

ofalleligiblewom

enin

UK

projectedaggreg

ate

£1.16milcostsaving

/yr

Sensitivity

analysis:costchange

perwom

anvariedfro

msaving

£253.38(37.5QALYsgained

per

year)to

costincrease

£108.12

depe

nden

ton

assumptions

with

correspo

nden

taggreg

ateannu

alsaving

s£23.75

million,or

aggreg

ateannu

alcostincrease

£10.13

million

Timeho

rizon

:RC

T(costinclud

ed)1995

–2009

Threeecon

omicanalyses

used

insynthe

sisof

potentialcost

saving

from

increasing

midwife-

ledservices

foreligible

maternities.

Issues

iden

tifiedarou

ndge

neralizability

offinding

s.Highrate

oftransfer

from

MLto

med

ical-ledcare

instud

ies

demon

strates‘risk’assessm

ent

criteria

unableto

iden

tifyall

wom

enwho

willde

velop

complications

inpreg

nancy

Rigo

rous

health

econ

omic

assessmen

tmeasures:INB,NMB,QALYs

Limitedto

UKsystem

Exclud

edRC

Tsfro

mAustraliaandothe

rcoun

trieswhe

reno

comparison

with

consultant-ledmod

elMixed

riskpreg

nancy

profile;sub

-group

analysis

show

costresults

consistent

forgrou

psas

(RR)

fetalloss

andne

onatalde

athoverlap

with

1.00


Table

3Summaryof

ThreeSystem

aticReview

sa(Con

tinued)

Aim

ofStud

ySample/Setting

Design

Major

CostFind

ings

Health

Outcomes

Streng

ths/Limitatio

ns

Bene

fit(NHB)

–QALY,

Qualityadjusted

lifeyear

gain

andlabo

ur

3.Devane,D.etal.(2010).

Section3:assessed

CEof

midwife-ledcare

compared

with

consultant

–led

care.

Estim

ated

potentialcost

saving

sof

expand

ing

midwife-ledcare

inUK

(pp.

33–45)

UnitedKing

dom

√

Basedon

3of

4RC

TsSee2.above

Hun

dley

1995

2844

wom

en;

Youn

g1997

1299

wom

en;

Begley

2009

1653

wom

en

System

aticreview

see2.

above

Sensitivity

analysisx3based

on8scen

ario

SA1:System

aticallyvarying

estim

ated

costsaving

sSA

2:System

aticallyvarying

RRforoverallfetalloss

&ne

onatalde

athusinglow

riskand‘mixed

risk’cases

SA3:System

aticallyvarying

assumed

uptake

ofML

service

Aspu

blishe

din

Ryan,Revill

etal.2013

Timeho

rizon

:RC

T(costinclud

ed)

1995

–2009

Expand

ingmidwife

–led

maternitiesshow

:Red

uced

rate

ofinterven

tions

inML

continuity

ofcare,including

:<ANho

spitalization

Redu

ceduseof

region

alanalge

siain

birth,less

episiotomyandinstrumen

tal

delivery&greaternu

mbe

rsof

wom

enmorelikelyto

expe

riencespon

tane

ousvaginal

birthBU

Tmay

notextrapolate

towom

enwith

iden

tifiedrisk

factors

Cochranebias

assessmen

ttool

used

fortrialinternal

validity

Not

gene

ralisable,small

numbe

rof

stud

ies

CEvariedwith

unitsize,

locatio

nandvolume

a Articlespresen

tedin

reversechrono

logicorde

r;√de

notesaminim

umPR

ISMAscoreof

20ba

sedacross

apo

ssible

totalo

f27

check-listite

ms


Table

4Prim

aryArticlesReview

edaStud

yde

sign

Aim

ofStud

ySample/Setting

Design/Metho

dMod

elused

(link

costs&he

alth

outcom

es)

1.Gao,Y.etal.(2014).

Com

paredCEtw

omod

els,Midwifery

Group

Practice(M

GP)

againstbaselinecoho

rtof

Abo

riginalmothe

rs/infants.

Clinicalandcostanalysis

Australia√

Region

alho

spital,NorthernTerrito

ryMGPcoho

rt:7

commun

ities

MGPWom

en=310

MGPBabies

n=315

(Sep

t2009

–June

2011)

Baselinecoho

rt:2

commun

ities

BaselineWom

enn=412

Baselinebabies

n=416

(Jan

2004-Dec

2006)Allrisk

Econ

omicevaluatio

n-retrospe

ctive

recordsaudit(BaselineJan2004-

Dec2006)p

rospectivedata

collection

(MGPSept

2009-Jun

e2011)

Cost-conseq

uences

analysis:A

ustraliando

llars

Measured/calculated

direct

costspe

rgrou

pEstablishe

dcomparativecostandchange

spo

stestablishm

entMGPservicefro

mfirstantenatal

appo

intm

entto

6weeks

postpartum

for

Abo

riginalmothe

rsandbabies

2.Tracy,S.K.et

al.(2013).

Assessefficacy,safety

andcostof

caseload

midwifery

versus

standard

hospitalm

aternity

care

forwom

enof

mixed

obstetric

risk

Dec

2008

-May

2011

Australia√

Wom

enof

allp

regn

ancy

riskstatus

(not

stratified)

Sample1748

wom

en2tertiary

teaching

hospitalsites,

2states,N

SW/Queen

sland

2arm

RCTCaseloadcare,W

omen

with

anamed

midwife

n=871versus

Wom

enStandard

HospitalC

are

n=877

Intentionto

treatanalyses

Cost-conseq

uences

analysis:A

ustraliando

llars

Costof

care

perwom

anbasedon

DRG

separatio

nanddirect

andindirect

costsfor

resource

usecollected

from

hospitalfinancial

system

Prim

ary&second

aryclinical&costou

tcom

esUnivariate

logisticregression

,OR95%

CIsand

Pearsonχ2

test;p

values;non

-param

etric

bootstrappe

rcen

tileCIsinfersign

ificance

ofeffects

3.JanS.et

al.(2004).Holistic

econ

omic

evaluatio

nof

anAbo

riginalCom

mun

ityCon

trolledMidwifery

Prog

ram

inWestern

Sydn

ey1990-1996

Australia√

Sample:2grou

psof

Abo

riginalwom

en,

Western

Sydn

eybirthing

betw

eenOct

1990

–Dec

1996,N

epean&Blacktow

nho

spitals

n=834

Anten

atalcare

atDaruk

Abo

riginal

Com

mun

ityCon

trolledProg

ram,oreither

hospital

Costanalyses

estim

ated

Direct

Prog

ram

costsanddo

wnstream

saving

s.Retrospe

ctivecase

record

audit

Costanalysis:A

ustraliando

llars

Clinicalandcostdata

linkedfro

mcase

record

andNSW

Midwives

DataCollection1991–1996

with

hospitald

atalinkedwith

Australian

NationalD

RGcostweigh

ts;M

edication:PBS

(pharm

aceuticalbe

nefits)Diagn

ostic

tests:MBS

(med

icarebe

nefits)

Sensitivity

analysisused

tomod

elun

certainty

4.Hom

erC.S.etal.(2001).

Assessclinicalandcostdifference–team

commun

itymidwifery

care

-CMWCcompared

tocontrol/standard

hospitalcare-SH

C1997-1998

Australia√

Sampleof

wom

enof

mixed

preg

nancyrisk

n=1089

CMW

=550

SHC=539

One

Australianpu

blicho

spital

Stateof

NSW

RCT-Ze

lenDesign

Costanalysis:C

MW

vsSH

C2team

seach

with

6fulltim

emidwives

provided

care

for600wom

en/yr(25

births/m

th/team)

Calculatedmeancost/w

oman

for

9compo

nentsof

maternity

care

Costanalysis:A

ustraliando

llars

Meancost/w

oman/group

-standard

errorsand

95%

CIcalculatedusingbo

otstraptechniqu

eCom

pone

ntsof

care

andcostforresources

used

foreach

wom

an:anten

atalclinic;anten

atal

admission

;day

assessmen

tun

it;labo

urand

birth;ho

spital-b

ased

postnatalcare;do

miciliary

postnatalcare;and,

admission

ofne

onates

tothespecialcarenu

rsery(SCN),on

-callcosts.

Salariesandwages

calculated

atmarketprices

Sensitivity

analysisin

3areas:Neo

natal

admission

toSC

N;Efficien

cyof

ANclinics;

Prop

ortio

nof

electiveCS

5.Ro

wley,M.J.et

al.(1995).

Exam

ined

cost/clinicaldifferences

forbirth

betw

een2grou

ps-Team

Midwifery

-6

midwives

vsroutineho

spitalcare

Australia√

Sampleof

wom

enof

mixed

preg

nancy

risk

n=814

Discretestratificationof

high

risk=

275wom

enTeam

midwifery

n=405

Hospitalcaren=409

One

Australianpu

blicho

spital

RCT:2grou

pscontinuity

team

(midwives)vs

routinecare

(hospital)

Costmeasured:

AustralianNational

CostWeigh

tsforDiagn

ostic

Related

Group

s(DRG

)pe

rbirth/de

livery

Intentionto

treat

Cost-effectiven

ess:Australiando

llars;d

irect

costs

Multip

leou

tcom

esmeasured.

Nosing

lemeasure

ofeffectiven

essde

rived

.Australiannatio

nalcostweigh

tsfordiagno

sis-

relatedgrou

ps(DRG

s)appliedto

outcom

esof

wom

enforwho

mcompleteresults

wereavail.

Perfo

rmed

retrospe

ctivelyby

clerkblinde

dto


Table

4Prim

aryArticlesReview

edaStud

yde

sign

(Con

tinued)

Aim

ofStud

ySample/Setting

Design/Metho

dMod

elused

(link

costs&he

alth

outcom

es)

Stateof

NSW

stud

y-basedon

med

icalrecords,covered

inpatient

costs.Costof

interven

tion&

comparativecare

estim

ated

byanalysing

midwives'salaries.

Nodiscou

ntingas

time-pe

riod<on

eyear.

Costsandqu

antitiesno

trepo

rted

separately.

Nosensitivity

analysisun

dertaken

.Nopricedatesgiven.

6.Kenn

y,P.et

al.(1994).Costanalyses:Team

Midwifery

VsStandard

hospitalcare.Includ

edclinicalou

tcom

esSept

1992

–July1993

Australia√

Samplen=446wom

enTeam

Midwifery

n=213

Standard

Caren=233

Westm

eadpu

blicho

spital

Stateof

NSW

RCT2Arm

Stud

yResource

costestim

ates:A

N,b

irth,

PNcare

Costestim

ated

whe

restatistically

sign

ificant

differencein

serviceuse

show

nInclud

ed:d

irect

costs,infrastructure,

staffsalaries-calculated

for‘low’and

‘high’riskwom

eneach

grou

p

Costanalysis(Drummon

d1987)

Costsestim

ated

basedon

resource

useat

AN,b

irthandPN

(includ

ingdo

miciliary)

stages

ofcare

separately

Costsbasedon

care

delivered

Nosensitivity

analysisun

dertaken

.Costin

gassumptions:costeffectiveifresource

costsof

midwifery

care

show

nto

beless

oreq

uivalent

toconven

tionalcareandhe

alth

bene

fitsof

midwife

care

relativeto

conven

tionalcareareshow

nto

bepo

sitive

a Studies

arepresen

tedin

reversechrono

logicorde

r;√de

notesaminim

umscoreof

6(from

possible

8)qu

ality

appraisalq

uestions;S

tudies

2,4,

5an

d6=rand

omised

controlledtrialw

ithlin

kedecon

omicevalua

tion


Table

5Prim

aryArticlesReview

edaStud

yresults

Stud

yMajor

CostFind

ings

Health

Outcomes

Streng

ths/Limitatio

ns

1.Gao,Y.etal.(2014)

Costsaving

AUS$703/mothe

r-infant

episod

efor

MGPcoho

rtwas

notstatisticallysign

ificant

(p=0.566)

MGP(m

idwifery

mod

el):

↓birthcost-$

411,p=

0.049

↓SC

Ncost–$1767,p=

0.144

↑ANcost+$272,p<

0.001

↑PNcost+$277,p<

0.001

↑infantreadmission

costs+$476,p=

0.05

↑travelcost=$115,p=

0.001

Timeho

rizon

:Midwife

coho

rt–allA

borig

inalmothe

rswho

gave

birthbe

tweenSept

2009

-June

2011

(and

their

infants)

Baselinecoho

rt–allA

borig

inalmothe

rswho

gave

birthbe

tweenJan2004

–Dec

2006

(and

theirinfants)

Wom

enwho

received

midwife

mod

elhad

moreantenatalcare,moreultrasou

nds,were

morelikelyto

beadmitted

toho

spitalin

antenatalp

eriod,

hadeq

uivalent

birth

outcom

es(i.e.mod

eof

birth;pre-term

birth;low

birthweigh

t)comparedwith

baselinecoho

rt.Babiesin

midwife

mod

eladmitted

toSpecialC

areNursery

had

sign

ificantlyredu

cedleng

thof

stay

Mixed

risk;sm

allsam

ple

Costassumptions

used

forecon

omic

analysis–expe

rtop

inionno

tprim

arydata

Missing

data

(3.7%

–24.5%);51%

all

cases=missing

data;

Timetren

dconfou

nding;

Hostelcosts&transportcostsno

tinclud

ed

2.Tracy,S.K.et

al.(2013).

Med

iancostsaving

of$566AUS/wom

anwith

Caseload/named

midwife

Timeho

rizon

:Dec

2008

–May

2011

Birthinterven

tions

redu

cedin

midwifery

mod

el30%

>spon

tane

ouson

setof

labo

r;↓analge

sia;

↓electivecaesarean;

Nosign

ificant

differenceforoverallrate

ofcaesareanbe

tweengrou

ps.

Similarsafe

outcom

esformothe

rsand

babies

betw

eengrou

ps

Registered

Trial:ACTRN12609000349246

Allpreg

nancyriskstatus

Nostratificationof

riskprofile

Defined

eligibility,inclusion

/exclusion

criteria

Stud

ysufficien

tlypo

wered

(80%

)and

Type

1error5%

Samplebias

challeng

edexternalvalidity

Cross-overs–didno

treceiveassign

edmod

elof

care

Non

-masking

ofgrou

pallocatio

nfro

mclinicians

3.JanS.et

al.(2004).

Net

costestim

ateAUS$1,200pe

rclient

–calculated

bysubtractingcostsaving

sto

othe

rcenters

Daruk

Anten

atalservicesaw

245wom

enfor339

preg

nanciesdu

ringstud

y

Timeho

rizon

:Wom

enbirthing

betw

een

Oct

1990

–Dec

1996

Nosign

ificant

differencein

servicebirth

weigh

tsor

perin

atalsurvival

Daruk

Anten

atalcare

=Gestatio

nalage

@1’stvisitlower;m

eannu

mbe

rANvisits

high

er;atten

danceforANtestsbe

tter

Wom

enstrong

lypo

sitivetowardmidwife

mod

elforrelatio

nship,

trust,accessibility,

flexibility,inform

ation,em

powermen

tandfamily-cen

teredcare

Mixed

riskpreg

nancy

Evaluatio

nframew

ork,bo

thqu

antandqu

almetho

dsFocusedon

antenatalcareattend

ance

and

access;costswerebroade

rthan

used

inconven

tionalecono

micanalyses

-includ

edbirthou

tcom

esandantenatalatten

dance

inasubseq

uent

preg

nancy

Assum

ptions

insensitivity

analyses

/estim

ated

downstream

health

costs

4.Hom

erC.S.etal.(2001).

Meancost/w

oman:C

MWCA$2

579vs

SHCA$3

483

Exclud

ingne

onatalcosts:

CMWCA$1

504(1449–1559;95%

CI)v

SHCA$1

643(1563–1729

95%CI)

Meancostsaving

9areasSH

C–CMCW:

Anten

atal+28.84

Day

Assessm

entUnit-5.42

Anten

atalinpatient

+38.74

On-callcost-21.81

Labo

ur/birth+68.83

HospitalP

ostnatalcare

43.85

Dom

iciliarycare

-11.06

Timeho

rizon

:1997–1998

(not

specific)

Caesarean

rate:C

MWM

13.3%

vsSH

C17.8%

(OR.0.6,95%

CI0.4±0.9,P=0.02)

Noothersign

ificant

differences

werede

tected

amon

gwom

enor

babies

forclinicalou

tcom

esor

even

tsdu

ringlabo

urandbirthbe

tweencare

mod

els

Costanalysisalon

gsideRC

T;10

000bo

otstrapreplications

Mixed

risksample;

Costsinclud

edresource

use,clinician

travel,neonate

care;

Noeq

uipm

ent,capitalo

rprog

ram

developm

entcosts;

Notransfer

rates;

Caseload/midwife

keyto

costsaving

;Not

possibleto

determ

ineop

timalcaseload

numbe

rs;unclear

ifdata

analyzed

byintentionto

treat


Table

5Prim

aryArticlesReview

edaStud

yresults

(Con

tinued)

Stud

yMajor

CostFind

ings

Health

Outcomes

Streng

ths/Limitatio

ns

SpecialC

areNursery

+2801.28

Total/w

oman

+904.09

5.Ro

wley,M.J.et

al.(1995).

Meancost↓4.5%

perbirth:

Team

MW

vRo

utinecare

A$3

324vs

A$3

475

Timeho

rizon

:May

1991

–June

1992

Includ

edfirstANvisitto

6weeks

afterbirth

Team

MW

wom

en:highe

rANclassattend

ance

OR1.73;95%

CI:1.23-2.42

↓birthinterven

tions

36%

vs24%

OR;1.73

(1.28–2.34);p<

0.001

↓pe

thidineuse0.32

(0.22–0.46)

↓ne

wbo

rnresuscitatio

n0.59

(0.41–0.86)

Maternalsatisfactionwith

team

care

was

greateron

3elem

ents:informationgiving

;participationin

decision

-making,

andrelatio

nships

with

caregivers.

Less

costthan

routinecare

andfewer

adverse

maternaland

neon

atalou

tcom

es

Coststud

yalon

gsideRC

TInclud

edwom

enof

allp

regn

ancy

riskstatus

Mod

elwas

team

midwifery

care,not

caseload

continuity

Costsbasedon

lyon

DRG

s;i.e.top

–do

wncoston

ly/no

tde

tailed.

Unableto

compare

with

othe

recon

omicevaluatio

ns

6.Kenn

y,P.et

al.(1994).

Team

Midwifery

vsStandard

Care:Avg

costs

ANcost/w

oman

Highrisk

$427vs

$456

Low

risk

$135vs

$133

Average

additio

nalcostpe

rbirth/wom

an$4.21

vs$9.36

PNcost/w

oman:

Hospitalstay$356.64

vs$397.26

(earlierdischarge)

Dom

iciliary$45.45

vs$45.80

Timeho

rizon

:Sep

t1992

–July1993

Sign

ificant

differences:m

anipulativevaginal

birth,ep

isiotomy&pe

rinealtears.

Wom

enin

team

midwife

care

repo

rted

high

erlevelsof

satisfactionover

3pe

riods

ofantenatal,birthandpo

stnatalcarewith

inform

ation,commun

icationandmidwife

attitud

eandskill

RCTLevel1

eviden

ce;

Allriskpreg

nancyinclud

ed;

Discretecosts:

AN,b

irthandPN

Robu

st,b

ottom-upcosting;

Team

midwife

mod

el,not

caseload;

Low

riskof

bias,alth

ough

blinding

notstated

;Loss

tofollow

up-19

inTM

vs22

inSH

C

a Studies

arepresen

tedin

reversechrono

logicorde

r;√de

notesaminim

umscoreof

6(from

possible

8)qu

ality

appraisalq

uestions;S

tudies

2,4,

5an

d6=rand

omised

controlledtrialw

ithlin

kedecon

omicevalua

tion



midwifery care [20]. However, it is of note that these re-sults are highly sensitive to assumptions, particularlychanges in the rate of fetal loss and neonatal death, aswell as the midwife’s caseload. When sensitivity analysiswas applied to include women of all risk categories andthe risk ratio for overall fetal/neonatal death was system-atically varied based on the 95% confidence interval of0.79 to 1.09 from pooled studies, the aggregate annualnet monetary benefit for continuity of midwifery careranged extremely widely. This varied from an estimatedgain of £472 million to a loss of £202 million. Net healthbenefit ranged from an annual gain of 15 723 QALYs toa loss of 6 738 QALYs. Additionally the midwife’s case-load needs to be sufficiently large to attain operationalefficiencies, otherwise the cost per maternity increases.The conclusion therefore is that the evidence base forcost-effectiveness of continuity of midwifery care forwomen with pregnancy risk is limited [19, 20]. As stated,these findings are limited to the UK context wheremidwifery-led and obstetric-led units are an establishedfeature of the health system. This is not the case in othercontexts, including Australia.

2. Cost of continuity of midwifery care and/or teammidwifery compared to Standard Care (medical)

Over the past two decades economic evaluations con-ducted alongside RCTs in several Australian states havedemonstrated cost saving and clinical effectiveness ofcontinuity of midwifery care models compared withstandard hospital care in the same setting [10, 37–39,46]. Some of these studies focused on ‘team midwifery’[37–39, 46] while others evaluated continuity of midwif-ery care models. In a ‘team model’ there is no primarycare provider and the level of continuity is variable,compared to continuity of midwifery care models wherenamed midwives provide services for women across thefull continuum of antenatal, birth and postnatal care [9].The cost evaluations of team midwifery and one costevaluation of continuity of midwifery care in Australiahave included pregnant women of mixed-risk status[10]. However, all these studies, with the exception ofKenny et al. [37] did not stratify results specific towomen with high-risk pregnancy.The most recent mixed-risk Australian trial identified

a median cost saving of A$566 for women who receivedcontinuity of midwifery care compared to standard hos-pital care services, these savings cannot be generalised tohigh risk groups [10]. This trial identified safe outcomesfor mothers and babies but no significant difference be-tween continuity of midwifery care and standard care forprimary outcomes of epidural analgesic use duringlabour, number of CS, instrumental vaginal births or un-assisted vaginal births. Earlier rigorous cost analysis of

community-based continuity of midwifery care modelfor all-risk women in Australia also identified mean costsavings per woman of A$804 in the continuity of mid-wifery care model. This included a significant differencein the rate of CS [39]. After neonatal costs were ex-cluded in this study, mean cost savings continued tofavour women and babies in the continuity of midwiferycare model by A$139 [39]. While it was not possible todetermine optimal service volume based on caseloadnumbers, the number of women booked for care in thecontinuity of midwifery care model was one of the im-portant keys to cost-effectiveness. The reason for this re-lates to efficiency and savings generated by the volumeof women able to be allocated to a maternity model inrelation to the staff ratio required to provide maternityservices [20, 47].Earlier team midwifery RCT studies identified reduced

levels of birth intervention in addition to modest costsavings for women of all-risk. One study identified as acost-effectiveness study used Australian Diagnostic Re-lated Groups ‘top-down costing’ that showed a meancost reduction for birth of 4.5% for women in the mid-wifery group, [38]. The other study, a cost analysis, ana-lysed discrete costs (‘bottom-up costing’) for eachepisode of service (i.e. antenatal, birth, and postnatalcare) in the midwifery model versus standard hospitalcare [37]. Specific cost for high- and low-risk pregnancyepisodes of care is shown in Table 4. Kenny et al. [37] isthe only study identified that separated the risk stratifi-cation profile of women in their all-risk pregnancy sam-ple in relation to costs. All the studies suggested a costsaving in intrapartum care in the midwifery model. Onestudy suggested higher cost and one study showed nodifference in cost of postnatal care in the midwiferymodel compared with the medical-led model. Cost re-sults for postnatal care also were not stratified as specificto women with pregnancy risk.

3. Cost-effectiveness of continuity of midwifery care forAboriginal women versus standard care

Two studies attempted to measure the cost of continu-ity of midwifery care in identified Australian populationswith higher pregnancy risk status. Gao et al. [41] used aretrospective baseline cohort measured against a pro-spective cohort of pregnant Aboriginal women (all-riskstatus) to identify cost changes from the first antenatalvisit through to six weeks postpartum after introductionof continuity of midwifery care. While there was a trendfor cost savings of A$703 for women at 6 weeks, thesewere not significantly different from baseline costs. Lim-itations of the study included small sample size, cost as-sumptions (hostel and transport were not included), andmissing data (51% of all cases). While no significant


difference in major birth outcomes was identified ante-natal attendance and hospital admissions increased, andaverage length of special care nursery stay for the babiesof the women decreased.An earlier cost analysis of a metropolitan,

Aboriginal-controlled, continuity of midwifery care ser-vice (all-risk) estimated direct program costs and down-stream savings in the health sector of A$1,200 perwoman [40]. Downstream savings projected longer -term cost benefits that were gained, for example, fromreductions in resource use experienced by associatedservices. The study used Australian National DRG costweights [48] and cost data from Medicare and thePharmaceutical Benefits Scheme [49] and sensitivity ana-lysis to model uncertainty. Costs included were broaderthan those used in conventional economic analyses.Among the additional cost considerations were clinicaloutcomes for birth, antenatal attendance in a subsequentpregnancy, and subtraction of cost savings to other cen-tres. While more recent clinical evaluation of midwiferymodels of care for Aboriginal women have demonstratedsignificant improvement in infant birthweight and peri-natal survival, specific cost analysis of these benefits havenot yet been undertaken as part of the studies [15, 50].

4. Patterns of antenatal care for women of highobstetric risk and comparative provider costs

In this review antenatal care provided by midwives forhigh risk and mixed risk samples showed reduced cost inthree RCT studies [20, 37, 39] and increased cost in twoothers (non-RCT) [40, 41](as shown in Tables 3 and 4).This is consistent with a cochrane review of patterns

of antenatal care which showed, among different pro-viders of antenatal care (midwife, general practitioner,obstetrician), primary outcome measures of low birth-weight, pre-eclampsia/eclampsia, severe postpartum an-aemia, and treated urinary tract infection (all high riskfactors for pregnancy complications including pre-termbirth) demonstrate similar clinical effectiveness [51, 52].

DiscussionIncreasingly health services need to justify quality out-comes as well as value for money [53–56]. Quality ma-ternity care is especially important for women whoexperience high risk pregnancy as inequitable healthoutcomes for these mothers and babies pose additionalpolicy and service implementation challenges for gov-ernment [57, 58]. Moreover, decision-makers often grap-ple to determine the most effective and sustainablemodels of care to close these gaps [28, 59, 60]. In highresource settings such as Australia, many women withthe most significant health inequities also experiencepregnancy complications with long-term comorbidity

[11, 30]. The public health burden, including the cost ofchronic disease for these women, their babies and thehealth system is higher and often lifelong [61]. Economicevaluation to inform decision-making regarding thecomparative cost-effectiveness of different maternitymodels across the continuum of childbearing thereforeshould be a high priority.This review demonstrated that there are few studies

specific to evaluating cost-effectiveness of midwiferycontinuity of care models for women who experiencehigh-risk pregnancy relative to other models of mater-nity care, including standard and traditional obstetric ledmodels. Of the studies included, significant limitationsand caveats apply. Inter-country comparison of cost andmodels of maternity care between health systems that donot share the same features prohibit comparative gener-alisability of both outcomes and models of care, includ-ing the costs attributable to different models andsystems. The costs and outcomes may vary widely ac-cording to structural factors such as funding model andworkforce arrangements and the influence of demo-graphic features and characteristics of woman who ex-perience high - risk pregnancy within the study samples.A strong international evidence base supports

woman’s early access to antenatal care, pre-natal educa-tion and health promotion strategies provided by mid-wives as an effective intervention to improve maternaland neonatal outcomes when integrated with other spe-cialised health and social support services [57, 58, 60].Poor access to antenatal care, including delayed attend-ance for the first visit is associated with higher rates ofpre-term birth and low birthweight infants and increasedinterventions in late pregnancy, all of which have beenfound to negatively impact cost [11, 62]. This reviewfound a small limited evidence-base to support the deliv-ery of cost -effective antenatal care by midwives towomen with identified pregnancy risk factors that deliverequivalent and/or improved health outcomes for themand their babies when compared to standard or trad-itional models of obstetric care [10, 37–39]. Additionally,while earlier systematic review has shown that low-riskpregnant women who receive midwifery-led care requirefewer antenatal visits, generating significant short-termcost savings for services [52], this is not always the casewhere women have identified medical and psychosocialrisk factors.Two studies included in this review identified higher

antenatal costs associated with increased frequency ofvisits for women identified with higher pregnancy riskswho may otherwise experience increased morbidity andmortality in pregnancy and childbearing [40, 41]. Con-sideration of overall ‘downstream’ savings’ within mid-wifery continuity of care models for women with riskfactors therefore is a relevant consideration. It is


recommended that future analyses include measures andmethods broader than those used in conventional eco-nomic analyses, for example, longer term modelling ofdisutility costs associated with onset of chronic diseasestates [63]. Downstream savings have been demonstratedto be important in estimating both program and healthsector costs accurately, particularly where access andsignificant health inequities have been identified [40].The limitations of the current studies in measuring theseeffects could be assessed by applying different methodsin health economics. Discrete choice experiment (DCE),for example, has been proposed as a more reliablemethod for eliciting women’s preferences for maternitycare [64]. DCE assesses and measures the costs associ-ated with consumer preferences for health care by askingpregnant women what they want.With respect to intrapartum care, while studies show

resource inputs and cost ratios for mode of birth to berelatively consistent among countries over time, recentcomparison of the costs of childbirth show significantcross-country variation. Factors that have been associ-ated with inter-country cost increases relate specificallyto workforce salary rates and provider charges infee-based health systems [24]. Further, overuse andunderuse of birth interventions, for example surgicalbirth, which may be more prevalent in women who ex-perience high risk pregnancy also demonstrate signifi-cant variation and remain subject to multiple influences,including health provider, health system, and fundingmodel [25, 65]. Data from all-risk pregnancies also showthat birth by caesarean section (CS) costs substantiallymore than vaginal birth [66]. International cost ratios formode of birth validated in Scotland, England, andAustralia indicate the incremental equivalent cost ratiosas: vaginal birth = 1; instrumental birth = 1.3; caesarean= 2.5 [44]. However, in this review no studies applied ormodelled these cost variations for intervention norlinked health outcomes specific to women with preg-nancy risk factors in comparative models of maternitycare. Despite this, outcomes in some of the studies in-cluded in this review show that costly intrapartum inter-ventions, including surgical birth in women withpregnancy risk factors may be safely reduced in intrapar-tum care for some women who receive continuity ofmidwifery care thereby also resulting in some cost saving[37–39].Place of birth is strongly associated with cost [22, 23,

67]. Cost is increased in hospital settings and com-pounded in facilities with fragmented models of care[68]. However, women with a complex pregnancy cur-rently access many fragmented maternity models and asignificant amount of their care in hospitals [11, 31].The model of maternity care therefore is an importantissue when considering cost. The recent introduction of

a national maternity care classification system (MaCCs)by the Australian government will enable improvedcomparison of outcomes and cost between midwiferycontinuity of care and other maternity models [69, 70] .In different maternity models and among different

provider groups, increased rates of surgical birth, espe-cially caesarean section, and other routine medical prac-tices associated with the cascade of intervention inchildbirth increase cost and morbidity [9, 43]. Longerbed stay associated with over intervention for womenand their infants results in increased rate and length ofhospitalisation, including admission and readmissions,and additional cost in the antenatal, intrapartum, andpostpartum periods [31, 69, 71]. The potential savingsfrom improved clinical outcomes generated throughmidwifery continuity of care across the childbearingcontinuum should be further evaluated in woman whoexperience high risk pregnancy and should include thepostnatal period [62].Most published studies have focused on women con-

sidered low risk for developing complications and re-ceiving midwifery-led care. Robust evidence frominternational and Australian studies demonstrates im-proved cost and clinical outcomes for these women andtheir babies across a number of key areas, notablyphysiological vaginal birth [9, 20].Significantly, midwifery models for low risk women

have shown a trend to variable cost saving in health ser-vice models where volume is sufficient to achieve effi-ciency and economies of scale [19, 20]. Savings accruewhere caseloads are maintained at an upper threshold of40 women per midwife per annum [20]. High-volumeinstitutional settings may optimise savings in thesemodels when antenatal hospitalisation rates are keptlow, vaginal birth rate is maximised, women and infantsundertake early discharge, and receive postnatalfollow-up at home or in the community [7, 10, 20, 72,73]. However, whether these clinical and cost benefitscan be extended through greater use of midwifery con-tinuity of care for women who experience pregnancy riskfactors require further evaluation. Discrete economicevaluation of midwifery continuity of care in the postna-tal period for women with pregnancy risks, as comparedto other maternity models including obstetric and stand-ard care was identified as significantly lacking.In the limited studies examined in this review, diver-

sity in study design and variation in the quality of the re-sults generated often negate reliable comparison of costresults. Where studies include women of mixed preg-nancy risk, variation and inconsistency of both study de-sign and the methods applied precluded reliable,comprehensive cost comparisons across the maternitycare continuum for woman with pregnancy risk factors.Robust economic evaluations conducted alongside a


RCT were considered to have high validity and reliabil-ity. None, however, focused exclusively on women withcomplicated pregnancy. This was in contrast tonon-randomised retrospective audit studies [40, 41].Studies in which a variety of statistical imputationmethods or expert opinion or estimates were used to ac-count for missing data created further challenges for re-liability in establishing the cost accuracy results of theeconomic evaluation [41, 74].Methodological challenges were identified in this re-

view. The first of these was selective risk sampling. Ofstudies included in this review some used “mixed risk”pregnancy samples that did not stratify clinical results orcost specific to the high-risk sub-set within the sample,thereby limiting generalisability of results. A secondchallenge included the variables selected for measure-ment in each study. The variables selected showed sig-nificant variation. Accurate measurement of variablesdepended on the data available and the reliability of thedata sources. The data sources of studies included in thisreview demonstrated wide fluctuation in reliability andquality across different time horizons making compari-son of outcomes and cost unreliable.Inconsistencies that compounded the methodological

challenges outlined above also were identified in relationto the various type of economic evaluations of maternitycare identified in this review – cost-effectiveness, costconsequences analysis, and economic synthesis. Theseincluded the use of varying cost methodology and studyassumptions. For example, ‘top-down’ costing approachesthat used diagnostic related groups cost weightsreflected activity-based funding models [10, 38], as con-trasted with ‘bottom-up’ costing that incorporated meas-urement of specified resource components – forexample equipment, consumables, staff salaries, caseloadnumbers, infrastructure costs [37, 46]. Moreover, sensi-tivity analysis was included in some of the economicevaluations and not in others. Incomplete or significantamounts of missing data replaced with estimates alsocalled into question the reliability and transferability ofcost estimate results. Even synthesis of results fromRCTs that applied the most rigorous health economicmeasures of INB, NMB, NHB and QALYs in the system-atic review conducted by Ryan et al [20] estimated pro-jected costs for midwife continuity of care thatfluctuated from significant aggregate saving and QALYgains to significant aggregate loss and QALY reductionswhen assumptions were challenged.

ConclusionRobust evaluation and comparative cost performance ofalternative models of maternity care is an importantconsideration in the provision of safe, quality maternityservices for women who experience complicated

pregnancy. While it is well known that poor outcomes atstart to life contribute to long-term chronic diseasestates that is costly for the health system, optimisingclinical effectiveness outcomes and cost efficiency forcare of women who experience complex pregnancy re-quires higher prioritisation. This review found limitedevidence to support the cost-effectiveness of midwiferycontinuity of care for women with complex pregnancy.Further evaluation of cost, resource use and clinical out-comes comparative to other models of maternity care iscritical. Further, this review shows that those studies thathave attempted to measure these costs demonstrate arange of inconsistencies. The application of inconsistentmethod undermines valid cost comparison of maternitymodels in developed countries. This remains an ongoingchallenge for policy makers and service providers inimplementing system change.Equitable access to continuity of midwifery care is an

important issue for women with pregnancy complica-tion. Evidence on the comparative cost-effectiveness re-source use and clinical outcomes delivered through newmaternity services is essential to the development of sus-tainable maternity models. This issue has relevance inan increasing number of settings in Australia and otherhigh resource countries in which services that addresshealthy start to life are critical to reduce current mater-nal newborn health inequity, and to meet the needs andexpectations of women and their families.

AbbreviationsAUS: : Australia; CI: Confidence interval; CS: Caesarean section; DCE: Discretechoice experiment; DRG: Diagnostic Related Group; INB: Incremental NetBenefit; NHB: Net Health Benefit; NHMRC: National Health Medical ResearchCouncil; NHS: National Health Service; NICE: National Institute ClinicalExcellence; NMB: Net Monetary Benefit; QALY: Quality Adjusted Life Year;RCT: Randomised controlled trial; UK: United Kingdom

AcknowledgementsDoctoral research of the first author was supported through a MidwiferyFellowship provided by WCH Foundation SA.

FundingThe authors declare that no funding was received to prepare this review.

Authors’ contributionsRDF, DKC & EC conceptualized the manuscript. RDF searched data bases anddrafted the manuscript. RDF, DKC & EC each contributed to criticallyreviewed drafts of the manuscript. RDF & DKC reviewed and refined Tables.RDF coordinated revision of the manuscript. All authors read and approvedthe final manuscript.

Competing interestsThe authors declare that they have no competing interests.

Publisher’s NoteSpringer Nature remains neutral with regard to jurisdictional claims inpublished maps and institutional affiliations.

Author details1Transforming Maternity Care Collaborative, Nursing and Midwifery, GriffithUniversity, Logan campus, University Drive, Meadowbrook, Queensland 4131,


Australia. 2Menzies Health Institute Queensland, Griffith University, GoldCoast, Queensland 4222, Australia.

Received: 9 January 2018 Accepted: 22 November 2018

References1. Australian Government. Improving Maternity Services in Australia: The

report of the Maternity Services Review. Canberra: Department of Health &Ageing, Commonwealth of Australia; 2009. p. 1-68.

2. Commonwealth of Australia. National Maternity Services Plan. Canberra: TheAustralian Health Ministers' Conference; 2011.

3. Newnham E. Midwifery directions: The Australian maternity services review.Health Sociology Review. 2010;19(2):245–59.

4. Wilkes L, Gamble J, Adam G, Creedy D. Reforming maternity services inAustralia: Outcomes of a private practice midwifery service. Midwifery. 2015;31(10):935–40.

5. England NHS. Implementing Better Births: Continuity of Carer. London: NHS;2017.

6. Dawson K, McLachlan H, Newton M, Forster D. Implementing caseloadmidwifery: exploring the views of maternity managers in Australia – Anational cross-sectional survey. Women and Birth. 2016;29(3):214–22.

7. McLachlan H, Forster DA, Davey MA, Farrell T, Gold L, Biro M, et al. Effects ofcontinuity of care by a primary midwife (caseload midwifery) on caesareansection rates in women of low obstetric risk: the COSMOS randomisedcontrolled trial. British J Obstet Gynaecol. 2012;119(12):1483–92.

8. Homer C. Models of maternity care: evidence for midwifery continuity ofcare. Med J Aust. 2016;205(8):370–4.

9. Sandall J, Soltani H, Gates S, Shennan A, Devane D. Midwife-led continuitymodels versus other models of care for childbearing women. CochraneDatabase Syst Rev. 2016;4:CD004667.

10. Tracy S, Hartz D, Tracy M, Allen J, Forti A, Hall B, et al. Caseload midwiferycare versus standard maternity care for women of any risk: M@NGO, arandomised controlled trial. The Lancet. 2013;382:1723 - 32, Publishedonline September 17, 2013 https://doi.org/10.1016/S0140-6736(13)61406-3.

11. Australian Institute of Health and Welfare. Australia’s mothers and babies2016—in brief. Canberra: AIHW; 2018.

12. Eunice Kennedy Shriver National Institute of Child Health and HumanDevelopment. What are the factors that put a pregnancy at risk? UnitedStates: US Department of Health and Human Services; 2016 [Available from:https://www.nichd.nih.gov/health/topics/high-risk/conditioninfo/factors.

13. McNeill J, Lynn F, Alderdice F. Public health interventions in midwifery: asystematic review of sytematic reviews. BMC Public Health. 2012;12:955.

14. Small R, Roth C, Raval M, Shafiei T, Korfker D, Heaman M, et al. Immigrantand non-immigrant women's experiences of maternity care: a systematicand comparative review of studies in five countries. BMC Pregnancy andChildbirth. 2014;14:152 2014.

15. Bertilone C, McEvoy S. Success in Closing the Gap: favourable neonataloutcomes in a metropolitan Aboriginal Maternity Group Practice Program.Med J Aust. 2015;203(6):262 e1 - .e7.

16. Rayment-Jones H, Murrells T, Sandall J. An investigation of the relationshipbetween the caseload model of midwifery for socially disadvantagedwomen and childbirth outcomes using routine data - A retrospective,observational study. Midwifery. 2015;31:409–17.

17. de Jonge A, Sandall J. Improving research into models of maternity care toinform decision making. PLoS Med. 2016;13(9):e1002135.

18. Henderson J, Petrou S. Economic Implications of Home Births and BirthCenters: A Structured Review. Birth. 2008;35:136–46.

19. Devane D, Brennan M, Begley C, Clarke M, Walsh D, Sandall J, et al.Socioeconomic Value of the Midwife: A systematic review, meta-analysis,meta-synthesis and economic analysis of midwife-led models of care.London: Royal College of Midwives; 2010.

20. Ryan P, Revill P, Devane D, Normand C. An assessment of the cost-effectiveness of midwife-led care in the United Kingdom. Midwifery. 2013;29:368–76.

21. Sutcliffe K, Caird J, Kavanagh J, Rees R, Oliver K, Dickson K, et al. Comparingmidwife-led and doctor-led maternity care: a systematic review of reviews. JAdv Nurs. 2012;68(11):2376–86.

22. Schroeder E, Patel N, Keeler M, Rocca-Ihenacho L, Macfarlane A. The EconomicCosts Of Intrapartum Care In Tower Hamlets: A Comparison Between The Cost

Of Birth In A Freestanding Midwifery Unit And Hospital For Women At LowRisk Of Obstetric Complications. Midwifery. 2017;45:28–35.

23. Schroeder E, Petrou S, Patel N, Hollowell J, Puddicombe D, Redshaw M,et al. Cost effectiveness of alternative planned places of birth in woman atlow risk of complications: evidence from the Birthplace in England nationalprospective cohort study. Br Med J. 2012;344(e2292):1–13.

24. Bellanger M, Or Z. What can we learn from a cross-country comparison ofof the costs of child delivery? Health Econ. 2008;17(Suppl. 1):S47–57.

25. Gibbons L, Belizan J, Lauer J, Betran A, Merialdi M, Althabe F. The GlobalNumbers and Costs of Additionally Needed and Unnecessary CaesareanSections Performed per Year: Overuse as a Barrier to Universal Coverage.Geneva: World Health Organisation; 2010.

26. Homer C, Biggs J, Vaughan G, Sullivan E. Mapping maternity services inAustralia: Location, classification and services. Aust Health Rev.2011;35:222–9.

27. Yelland J, Riggs E, Small R, Brown S. Maternity services are not meeting the needsof immigrant women of non-English speaking background: Results of twoconsecutive Australian population based studies. Midwifery. 2015;31(7):664–70.

28. Kildea S, Tracy S, Sherwood J, Magick-Dennis F, Barclay L. Improvingmaternity services for Indigenous women in Australia: moving from policyto practice. Med J Aust. 2016;205(8):375–9.

29. Pilcher J, Kruske S, Barclay L. A review of rural and remote health serviceindexes: are they relevant for the development of an Australian rural birthindex? BMC Health Serv Res. 2014;2014(14):548.

30. Monk A, Harris K, Donnolley N, Hilder L, Humphrey M, Gordon A, et al.Perinatal deaths in Australia 1993-2012 (AIHW Perinatal deaths series no. 1.Cat. no. PER 86). Australian Institute of Health and Welfare: Canberra; 2016.

31. National Health Performance Authority. Healthy Communities: Child andmaternal health in 2009 - 2012. Sydney: National Health PerformanceAuthority; 2014.

32. Kable A, Pich J, Maslin-Prothero S. A structured approach to documenting aserach strategy for publication: A 12 step guideline for authors. Nurse EducToday. 2012;32:878–86.

33. Torraco R. Writing integrative literature reviews: Guidelines and examples.Hum Resour Dev Rev. 2005;4:356–67.

34. Denyer D, Pilbeam C. Doing a literature review in business andmanagement. Presentation to the British Academy of Managment DoctoralSymposium, Liverpool: UK; 2013.

35. Moher D, Liberati A, Tetzlaff J, Altman D. The PRISMA Group. PreferredReporting Items for Systematic Reviews and Meta-Analyses: The PRISMAStatement. PLoS Med. 2009;6(6).

36. National Health and Medical Research Council. How to compare the costsand benefits: evaluation of the economic evidence. Australian GovernmentPublishing Service; 2001.

37. Kenny P, Brodie P, Eckermann S, Hall J. Westmead Hospital Team MidwiferyProject Evaluation Final Report. Sydney: Centre for Health EconomicsResearch and Evaluation, University of Sydney; 1994.

38. Rowley M, Hensley M, Brinsmead M, Wlodarczyk J. Continuity of care by amidwife team versus routine care during pregnancy and birth: arandomised trial. Med J Aust. 1995;163(6):289–193.

39. Homer C, Matha DV, Jordan LG, Wills J, Davis GK. Community-basedcontinuity of midwifery care versus standard hospital care: a cost analysis.Aust Health Rev. 2001a;24(1):85–93.

40. Jan S, Conaty S, Hecker R, Bartlett M, Delaney S, Capon T. An holisticeconomic evaluation of an Aboriginal community - controlled midwiferyprogram in Western Sydney. J Health Serv Res Policy. 2004;9(1):14–21.

41. Gao Y, Gold L, Josif C, Bar-Zeev S, Steenkamp M, Barclay L, et al. A cost-consequences analysis of a Midwifery Group Practice for Aboriginal mothersand infants in the Top End of the Northern Territory Australia. Midwifery.2014;30(4):447–55.

42. Stewart M, McCandlish HJ, Brocklehurst P. Report of a structured review ofbirth centre outcomes: review of evidence about psychosocial andeconomic outcomes for women with straightforward pregnancies who planto give birth in a midwife-led birth centre, and outcomes for their babies.Oxford: National Perinatal Epidemiology Unit; 2005.

43. Tracy S, Welsh A, Hall B, Hartz D, Lainchbury A, Bisits A, et al. Caseloadmidwifery compared to standard or private obstetric care for first timemothers in a public teaching hospital in Australia: a cross sectional study ofcost and birth outcomes. BMC Pregnancy Childbirth. 2014;14:46.

44. Tracy S, Tracy M. Costing the cascade: estimating the cost of increased obstetricintervention in childbirth using population data. BJOG. 2003;110:717–24.

https://doi.org/10.1016/S0140-6736(13)61406-3https://www.nichd.nih.gov/health/topics/high-risk/conditioninfo/factors


45. Petrou S, Glazener C. The economic costs of alternative modes of deliveryduring the first two months postpartum: results from a Scottishobservational study. Br. J Obstet Gynaecol. 2002;109(2):214–7.

46. Homer C, Davis G, Brodie P, Sheehan A, Barclay L, Wills J, et al. Collaboration inmaternity care: a randomised controlled trial comparing community-basedcontinuity of care with standard hospital care. BJOG. 2001b;108:16–22.

47. Sandall J, Page L, Homer C, Leap N. Chapter 2 Midwifery continuity of care:what is the evidence? In: Homer C, Brodie P, Leap N, editors. MidwiferyContinuity of Care: A Practical Guide. Chatswood NSW: Elsevier Australia;2008. p. 25–46.

48. Commonwealth Department of Health and Ageing. National Hospital CostData Collection. Canberra; 2010.

49. Australian Government Department of Human Services. Medicare Online2014 [Available from: https://www.humanservices.gov.au/individuals/subjects/medicare-services.

50. Hickey S, Roe Y, Gao Y, Nelson C, Carson A, Currie J, et al. The IndigenousBirthing in an Urban Setting study: the IBUS study. A prospective birth cohortstudy comparing different models of care for women having Aboriginal andTorres Strait Islander babies at two major maternity hospitals in urban SouthEast Queensland, Australia. BMC Pregnancy Childbirth. 2018;18:431.

51. Villar J, Ba'aqeel H, Piaggio G. WHO randomised controlled trial for theevaluation of a new model of routine antenatal care. The Lancet.2001;357:1551–64.

52. Patterns of routine antenatal care for low risk women (Cochrane Review)[Internet]. John Wiley & Sons. 2007.

53. Renfrew MJ, McFadden A, Bastos MH, Campbell J, Channon AA, Cheung NF,et al. Midwifery and quality care: findings from a new evidence-informedframework for maternal and newborn care. The Lancet. 2014;384(9948):1129–45.

54. Berwick D, Nolan T, Whittington J. The Triple Aim: Care, Health And Cost.Health Aff. 2008;27(3):759–69.

55. Porter M. The Strategy to Transform Health Care and the Role of Outcomes.OECD Policy Forum, People at the Centre: The Future of Health; 2017.

56. All Party Parliamentary Group on Global Health. Triple Impact Report.London; 2016.

57. Miller S, Abalos E, Chamillard M. Beyond too little, too late and too muchtoo soon: a pathway towards evidence-based, respectful maternity careworldwide. Lancet 2016; published online September 15. 2016.

58. Van Lerberghe W, Matthews Z, Achadi E, Ancona C, Campbell J, Channon A,et al. Country experience with strengthening of health systems anddeployment of midwives in countries with high maternal mortality. TheLancet. 2014;384(9949):1215–25.

59. COAG. Health Council. National Framework for Maternity Services Project.2017.

60. ten Hoope-Bender P, de Bernis L, Campbell J, Downe S, Fauveau V, FogstadH, et al. Improvement of maternal and newborn health through midwifery.The Lancet. 2014;384(9949):1226–35.

61. Moore T, Arefadib N, Deery A, West S. The First Thousand Days: AnEvidence Paper. Centre for Community Child Health, Murdoch Children'sResearch Institute: Parkeville, Victoria; 2017.

62. Homer C, Friberg IK, Dias MAB, ten Hoope-Bender P, Sandall J, Speciale AM,et al. The projected effect of scaling up midwifery. The Lancet. 2014;384(9948):1146–57.

63. Drummond MF, Sculpher MJ, Torrance GW, O'Brien BJ, Stoddart GL.Methods for the Economic Evaluation of Health Care Programmes, 3rd ed.3rd ed. Oxford, UK: Oxford University Press; 2005.

64. Petrou S, McIntosh E. Commentary: Using Stated Preference Discrete ChoiceExperiments to Elicit Women's Preferences for Aspects of Maternity Care.Birth. 2011;38(1):47–8.

65. Betrán A, Temmerman M, Kingdon C, Mohiddin A, Opiyo N, Torloni M, et al.Interventions to reduce unnecessary caesarean sections in healthy womenand babies. The Lancet. 2018;392:1358–68.

66. Petrou S, Henderson J, Glazener C. Economic aspects of caesarean sectionand alternative modes of delivery. Baillieres Best Pract Res Clin ObstetGynaecol. 2001;15(1):145–63.

67. Birthplace in England Collaborative Group BP, Hardy, P, Hollowell, J et al. .Perinatal and maternal outcomes by planned place of birth for healthywomen with low risk pregnancies: the Birthplace in England nationalprospective cohort study. British Medical Journal. 2011;343:d7400.

68. Scarf V, Catling C, Viney R, Homer C. Costing Alternative Birth Settings forWomen at Low Risk of Complications: A Systematic Review. PLoSOne. 2016;11(2):e0149463.

69. National Health Performance Authority. Hospital Performance: Costs ofacute admitted patients in public hospitals in 2011 - 2012 TechnicalSupplement. Canberra 2015; 2015.

70. Donnolley N, Butler-Henderson K, Chapman M, Sullivan E. The developmentof a classification system for maternity models of care. Health Inf Manag J.2016;45(2):64–70.

71. National Health Performance Authority. Hospital Performance: Length ofstay in public hospitals in 2011 - 2012. 2013.

72. Kenny C, Devane D, Normand C, Clarke M, Howard A, Begley C. A Cost-Comparison of Midwife-led Compared with Consultant-led Maternity Carein Ireland (The MIDU Study). Midwifery. 2015;31(11):1032–8.

73. Henderson J, McCandlish R, Kumiega L, Petrou S. Systematic review ofeconomic aspects of alternative modes of delivery. Br J Obstet Gynaecol.2001;108:149–57.

74. Gelman A, Hill J. Chapter 25 Missing-data imputation. Data Analysis UsingRegression and Multilevel / Hierarchical Models. UK: Cambridge UniversityPress; 2006. p. 529-43.

https://www.medicareaustralia.gov.au/provider/business/online/medicare-online.jsphttps://www.medicareaustralia.gov.au/provider/business/online/medicare-online.jsp

AbstractBackgroundAims / objectivesDesignMethodResultsConclusions

Introduction and backgroundAims and objectivesMethodSearchInclusion criteriaResultsAppraisal of studiesQuality of studiesDiscussionConclusionAbbreviationsAcknowledgementsFundingAuthors’ contributionsCompeting interestsPublisher’s NoteAuthor detailsReferences

cost-effectiveness of continuity of midwifery care for ...... · centres and home birth), as...

Documents