REVIEW Open Access

Obstetric ultrasound use in low and middleincome countries: a narrative reviewEunsoo Timothy Kim1,2, Kavita Singh1,3* , Allisyn Moran4, Deborah Armbruster4 and Naoko Kozuki5,6

Abstract

Introduction: Although growing, evidence on the impact, access, utility, effectiveness, and cost-benefit of obstetricultrasound in resource-constrained settings is still somewhat limited. Hence, questions around the purpose and theintended benefit as well as potential challenges across various domains must be carefully reviewed prior to implementationand scale-up of obstetric ultrasound technology in low-and middle-income countries (LMICs).

Main Body: This narrative review discusses these issues for those trying to implement or scale-up ultrasound technology inLMICs. Issues addressed in this review include health personnel capacity, maintenance, cost, overuse and misuseof ultrasound, miscommunication between the providers and patients, patient diagnosis and care management,health outcomes, patient perceptions and concerns about fetal sex determination.

Conclusion: As cost of obstetric ultrasound becomes more affordable in LMICs, it is essential to assess the benefits,trade-offs and potential drawbacks of large-scale implementation. Additionally, there is a need to more clearly identify thecapabilities and the limitations of ultrasound, particularly within the context of limited training of providers, to ensure thatthe purpose for which an ultrasound is intended is actually feasible. We found evidence of obstetric uses ofultrasound improving patient management. However, there was evidence that ultrasound use is not associatedwith reducing maternal, perinatal or neonatal mortality. Patients in various studies reported to have both positiveand negative perceptions and experiences related to ultrasound and lastly, illegal use of ultrasound for determiningfetal sex was raised as a concern.

Plain English summaryPrior to the implementation and scale-up of obstetric ultra-sound technology in low- and middle-income countries,questions around the purpose, intended benefit, potentialtrade-offs and challenges must be carefully reviewed. Thisreview paper focuses on issues around health personnelcapacity, maintenance, cost, misuse and overuse, miscom-munication between patients and providers, patient diagno-sis and case management, health outcomes, perceptionsand concerns about fetal sex determination.We conclude that it is essential to assess the benefits,

trade-offs and potential drawbacks of large-scale imple-mentation as cost of obstetric ultrasound becomes more

affordable in low- and middle-income countries. Add-itionally, there is a need to more clearly identify the cap-abilities and the limitations of ultrasound, particularlywithin the context of limited training of providers, to en-sure that the purpose for which an ultrasound is intendedis actually feasible. We found evidence of obstetric uses ofultrasound improving patient management. However,there was evidence that ultrasound use is not associatedwith reducing maternal, perinatal or neonatal mortality.Patients in various studies reported to have both positiveand negative perceptions and experiences related to ultra-sound and lastly, illegal use of ultrasound for determiningfetal sex was raised as a concern.

BackgroundThere are numerous applications and indications for ultra-sound in low- and middle-income countries (LMICs).Ultrasound has been used to diagnose obstructed labor,non-cephalic presentation, single or multiple pregnancy, in-complete miscarriage, molar pregnancy, ectopic pregnancy,

* Correspondence: kavita_singh@unc.edu1Department of Maternal and Child Health, Gillings School of Global PublicHealth, University of North Carolina at Chapel Hill, 135 Dauer Dr, Chapel Hill,NC 27516, USA3MEASURE Evaluation, Carolina Population Center, University of NorthCarolina at Chapel Hill, Carolina Square, Suite 330, 123 West Franklin St,Chapel Hill, NC 27516, USAFull 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. The Creative Commons Public Domain Dedication waiver(http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.

Kim et al. Reproductive Health (2018) 15:129 https://doi.org/10.1186/s12978-018-0571-y

fetal abnormality, intrauterine growth restriction and pla-centa previa [1–5]. It has also been used to measure pelvicoutlet and estimate gestational age [5–7]. Although the de-gree of diagnostic accuracy may vary depending on whenpregnant women present themselves for an ultrasoundexam [7], its utility has been highlighted in many studiesnonetheless. Accurate assessment of gestational age for ex-ample is useful in distinguishing pre-term newborns fromnewborns who are low birth weight (but not pre-term),which is important because the needed interventions maydiffer [8].In LMICs, ultrasound is mainly used to diagnose ob-

stetric conditions [7, 9]. However, gynecological condi-tions have also been evaluated by ultrasound as well,particularly in emergency medicine [9, 10]. Kotlyar andMoore stated that 53% of ultrasound scans were per-formed either for obstetric or gynecological conditionsin a single center study from Monrovia, Liberia [11]. Inanother study in Cameroon, ultrasound determined that33% of patients had gynecological conditions and 15%had obstetric conditions [12]. Applications of ultrasoundalso extend to abdominal, musculoskeletal, cardiac, renal,pulmonary, trauma and soft tissue and vascular conditions[3, 7, 10, 13, 14] as well as HIV, tuberculosis, intussuscep-tion, Wilms’ tumor, Burkitt’s lymphoma, hepatitis C, Cha-gas’ disease, filariasis, myiasis and other protozoal,helminthic, viral and bacterial infections [7, 9]. For example,a trial in Rwanda reported that due to the high prevalenceof HIV related cardiomyopathies, TB-related pericardial ef-fusions and rheumatic disease, cardiac applications of ultra-sound were frequently used [14]. In Botswana, most of thedisease burden at the district level public hospitals areobstetric-related, followed by gynecological conditions andhepato-biliary conditions [4]. In Sudan where the preva-lence of schistosomiasis is high, applications of ultrasoundwere highest for liver and spleen disorders, followed byobstetric conditions [15]. As evidenced by past studies, ap-plications of ultrasound use are diversified based on geo-graphic and disease context. Such high adoptability andwide ranging applications of ultrasound make it suitable foruse in LMICs.In recent years, portable ultrasound machines have

become increasingly popular in LMICs due to theiraffordability, user-friendliness and adoptability to theharsh and restrictive conditions of resource-poor set-tings [3, 6, 7, 10, 16, 17]. For example, newer ultra-sound models are hand-held or hand-carried [17], and theprice has significantly decreased compared to that offull-size ultrasound machines [3]. The interface softwareon these devices have also been simplified and streamlined[6] so that they may be used more broadly by differentcadres of health workers, who have different capabilities[10]. As newer generations of ultrasound are becomingmore suitable for LMICs, manufacturers could be looking

to expand their markets and local healthcare staff may beweighing the benefits and disadvantages of introducingthese newer ultrasound models into their facilities. Multi-lateral and bilateral organizations as well as private do-nors, therefore, may want to comprehensively assess thefeasibility of introducing ultrasound in these settings be-fore appropriating funds.Although growing, evidence on the access, utility, ef-

fectiveness, and cost-benefit of obstetric ultrasound inresource-constrained settings is still somewhat limited[5, 18]. It is also important to note that there is no evi-dence of its impact on reducing maternal or perinatalmortality [9, 19, 20]. Hence, questions around the purposeand the intended benefit as well as potential challengesacross various domains must be carefully reviewed priorto implementation and scale-up of obstetric ultrasoundtechnology in LMICs.This narrative review discusses these issues for those try-

ing to implement or scale-up ultrasound technology inLMICs. Issues addressed in this review include healthpersonnel capacity, maintenance, cost, overuse and misuseof ultrasound, miscommunication between the providersand patients, patient diagnosis and care management, healthoutcomes, patient perceptions and concerns about fetal sexdetermination. Because obstetric conditions are the mostcommon indication for ultrasound in LMICs [7, 9], this re-view mainly focuses on obstetric uses of ultrasound.

MethodsThe existing literature on the use of ultrasound in LMICswas reviewed. The initial search was focused only on pub-lished review articles because of the short timeframe (May3rd – August 2nd, 2016) in which to prepare for a Tech-nical Consultation Meeting on Ultrasound Use in Low-and Middle-Income Countries which was held in Wash-ington D.C. on August 2nd, 2016. This narrative reviewwas initially prepared for the Technical ConsultationMeeting in Washington D.C. but was later expanded toinclude additional articles.Published review articles were searched in the follow-

ing databases: PubMed, Global Health, CINAHL, Webof Science and Embase. Exact search terms and filtersfor the review articles are included in the Appendix. Theinitial database search returned a total of 391 articlesand 6 review articles were retained for the literature re-view. After the initial electronic database search, 59 add-itional relevant articles and grey literature publicationswere identified through manual reference and Googlesearches. In this paper, data are synthesized from 65sources (See Fig. 1).Summary tables are presented for articles/studies in

the review (See Tables 1, 2 and 3). Study quality wasassessed using the following criteria: 1) study design –randomized or non-randomized; 2) presence of a control

Kim et al. Reproductive Health (2018) 15:129 Page 2 of 26

group. Information about the study location, study period,study description, key points, sample size, funding supportand type of ultrasound machine are also presented. Quali-tative studies or studies from which anecdotal accountswere retrieved were not included.

FindingsWhat are the effects of obstetric ultrasound?Care utilization and mortalityA recent cluster-randomized trial in Pakistan, Kenya,Zambia, the Democratic Republic of the Congo andGuatemala looked at the effects of antenatal ultrasound usein rural health center settings [20, 21]. After stratifying bycountry, each of the intervention and comparison clusterswere defined as a catchment area of a health center that re-cords about 500 births every year [21]. Pregnant women inthe intervention clusters were generally offered two ante-natal ultrasound examinations, first exam between 18 and22 weeks and the second exam between 32 and 36 weeks,and those with identified complications were referred tohigher-level health facilities [21]. At the first visit, gesta-tional age was determined and fetal number and position,

cervical length, potential amniotic fluid abnormalities andpotential congenital abnormalities were examined [21]. Atthe second visit, placental location and growth were ob-served in addition to all of the checks performed during thefirst visit [21]. The intervention package also includedhealth worker training for antenatal ultrasound, emergencyobstetric and neonatal care training at higher-level referralfacilities and community sensitization events where peoplewere informed about the diagnostic capabilities of ultra-sound as well as the antenatal ultrasound services being of-fered at intervention facilities [21]. This recent studyincluding five countries with a rigorous cluster-randomizeddesign found no difference in intervention and comparisonclusters in terms of antenatal care use, facility delivery, still-birth rate, neonatal mortality and maternal mortality [20].Another randomized study conducted in South Africa

found that routine second-trimester ultrasound scanningdid not result in a significant difference in perinatal mor-tality between the ultrasound scan group and the controlgroup, 4.3 and 4.1% respectively with a relative risk of1.05 and 95% confidence interval between 0.54 and 2.03[9, 19]. However, the study also noted that it had low

Fig. 1 Flowchart of search results

Kim et al. Reproductive Health (2018) 15:129 Page 3 of 26

Table

1QualityAssessm

entof

Stud

iesrelatedto

OutcomeMeasures

Autho

r,Year

Locatio

nStud

yPerio

dType

ofUltrasou

ndStud

yDescriptio

nKeyPo

ints

SampleSize

Fund

ingor

Supp

ort

Ross

AB,DeStig

ter

KK,Rielly

M,Sou

zaS,Morey

GE,Nelson

M,etal.2013[23]

Ugand

a2years

Low

cost,p

ortable

ultrasou

ndmachine

sQuantitativestud

y-no

controlg

roup

includ

ed.

Historicalcontrold

atawas

used

asacomparison

.Wom

enreceived

ultrasou

ndscan

atthefirstANCvisit

andat

32weeks

gestation.

Basedon

need

,ultrasou

ndscanswerepe

rform

edat

othe

rANCvisitsas

well.

Twomidwives

atthestud

yhe

alth

center

weretraine

dfor3days.Traininginclud

edclassroo

mandhand

s-on

sessions

with

acompe

tency

assessmen

t.Thefocusof

theprog

ram

was

ondiagno

sing

multip

lege

stations,seq

uelaeof

abortio

n,causes

ofob

structed

labo

rand

placen

taprevia.

Totalanten

atalcare

visits

sign

ificantlyincreased

from

133.5visitsto

230.3

visits.

Thenu

mbe

rof

first

antenatalcarevisits

increasedfro

m82.2visits

to119.1visits.

Thenu

mbe

rof

second

antenatalcarevisits

increasedfro

m35.6visits

to60.4visits.

Thenu

mbe

rof

third

antenatalcarevisits

increasedfro

m11.6visits

to31.9visits.

Thenu

mbe

rof

fourth

antenatalcarevisits

increasedfro

m4.1visits

to19.0visits.

Theaveragenu

mbe

rof

deliveriesat

thehe

alth

center

sign

ificantly

increasedfro

m28.1

deliveriesto

45.6

deliveries.

Priorto

introd

uctio

nof

ultrasou

nd:

Anten

atalvisits

(n=41)

Atten

dedde

liveries

(n=23)

After

introd

uctio

nof

ultrasou

nd:

Anten

atalvisits

(n=23)

Atten

dedde

liveries

(n=23)

ImagingtheWorld

(received

fund

ingfro

mtheFine

berg

Foun

datio

n,theBillandMelinda

Gates

Foun

datio

n,Ph

illips

Health

Care,McKesson

Corpo

ratio

n,Peervue

Corpo

ratio

n).

Research

projectwas

also

supp

ortedby

anAlpha

Omeg

aAlpha

postgraduate

award.

Mbu

yita

S,TillyaR,

God

freyR,Kinyon

geI,Shaban

J,Mbaruku

G,2015[25]

Tanzania

1year

Portablehand

-held

Vscan

QuantitativeandQualitative

stud

y-Before

andAfte

rde

sign

with

interven

tion

andcontrolg

roup

s.Po

rtablehand

held

ultrasou

ndtechno

logy

(Vscan)was

introd

uced

inroutineANCservices.

Assistant

med

icalofficersat

thehe

alth

center

leveland

mid-levelp

rovide

rs(clinical

officers,nu

rsemidwives

and

med

icalattend

ants)at

the

dispen

sary

levelw

ere

traine

dto

usetheVscan

ultrasou

ndin

routineANC

services.

Betw

eenbaselineand

endlinepe

riods

inthe

interven

tionarea:

Thenu

mbe

rof

first

antenatalcarevisitsdid

notchange

sign

ificantly.

Thenu

mbe

rof

four

ormoreantenatalcarevisits

increasedsign

ificantly.

Thenu

mbe

rof

facility

deliveriesincreased

sign

ificantly.

Interven

tionarm:

Baseline(n

=381)

Endline(n

=423)

Con

trol

arm:

Baseline(n

=394)

Endline(n

=383)

GEHealth

care

East

Africa

Services

Limited

supp

ortedtheproject

VanDyk

B,Motto

JAandBu

chmannEJ,

2007

[19]

SouthAfrica

Not

Men

tione

dNot

Men

tione

dQuantitativestud

y-cluster-

rand

omized

controlledtrial.

Con

trol

grou

pof

noultra

soun

dwas

includ

ed.

Routinesecond

-trim

ester

ultrasou

ndscanswere

Routinesecond

-trim

ester

ultrasou

ndscanning

did

notresultin

asign

ificant

differencein

perin

atal

mortalitybe

tweenthe

ultrasou

ndscan

grou

p

Interven

tiongrou

p(n

=416)

Con

trol

grou

p(n

=388)

Not

Men

tione

d

Kim et al. Reproductive Health (2018) 15:129 Page 4 of 26

Table

1QualityAssessm

entof

Stud

iesrelatedto

OutcomeMeasures(Con

tinued)

Autho

r,Year

Locatio

nStud

yPerio

dType

ofUltrasou

ndStud

yDescriptio

nKeyPo

ints

SampleSize

Fund

ingor

Supp

ort

perfo

rmed

byan

expe

rienced

licen

sed

ultrason

ograph

erto

wom

enwith

low

risk

preg

nanciesat

18–23

weeks

byclinical

estim

ation.

andthecontrolg

roup

.How

ever,sam

plesize

was

insufficien

t.

GeertsLT,Brand

EJandTheron

GB,

1996

[22]

SouthAfrica

10mon

ths

Aloka

SSD-500,SSD

-620,SSD-650,w

ith3.5MHzcurviline

arelectron

icprob

es.

Quantitativestud

y-

rand

omized

controlled

trial.Con

trol

grou

pof

selectiveultrasou

ndcomparedto

the

interven

tiongrou

pof

routineultrasou

nd.

“Pregn

antpatientswith

out

riskfactorsforcong

enital

anom

aliesreferred

for

ultrason

ograph

ybe

tween

18and24

weeks

ofge

station”

participated

inthestud

y.Wom

enin

theinterven

tion

grou

preceived

“asing

lelevelo

neultrasou

ndexam

perfo

rmed

byob

stetric

registrarsor

med

ical

officersspecifically

traine

din

obstetric

ultrasou

nd.”

Thedifferencein

perin

atalmortality

betw

eentheinterven

tion

andcontrolg

roup

swas

notstatisticallysign

ificant.

Therewas

noten

ough

samplesize

tode

tect

differences

inpe

rinatal

mortalityalon

e.How

ever,w

hen

combine

dwith

major

morbidities,therewas

asign

ificant

increase

(25%

)in

totaladverse

perin

atal

outcom

esin

thecontrol

grou

pwith

outroutine

ultrason

ograph

y.

Interven

tiongrou

p(n

=496)

Con

trol

grou

p(n

=492)

Not

Men

tione

d

McClure

E,Golde

nbergR,

SwansonD,Saleem

S,Esam

aiF,Garces

A,etal.2017[20]

Pakistan,Ken

ya,

Zambia,Dem

ocratic

Repu

blicof

Con

go,

Guatemala

18mon

ths

Ultrasou

ndeq

uipm

ent

donatedby

GE

Health

care

Quantitativestud

y-cluster

rand

omized

trial.

Clusterswererand

omized

either

tothecontrolg

roup

ofusualcareor

tothe

interven

tiongrou

pof

receivingultrasou

ndat

16–22and32–36weeks

andapprop

riate

referrals.

Multip

leinterven

tion

compo

nentswereinclud

ed:

Two-weekob

stetric

ultrasou

ndtraining

for

varyingcadres

ofno

n-ph

ysicianhe

alth

workers.

3-mon

thpilotpe

riodand

follow-upqu

ality

assurance

system

toen

sure

successful

implem

entatio

n.Training

forreferralfacility

staffin

managem

entof

Therewas

nodifference

inprim

arycompo

site

outcom

e,stillbirthrate,

neon

atalmortalityrate,

near

missrate,m

aternal

death,at

leaston

eantenatalvisit,four

ormoreantenatalvisits

and

deliveryat

theho

spital

with

cesarean

section

capacity

betw

eenthe

interven

tionandcontrol

grou

ps.

78%

ofde

liveriesin

the

interven

tiongrou

preceived

oneor

more

ultrasou

ndserviceand

60%

received

two

ultrasou

ndservices.

9%of

wom

enwere

referred

basedon

Guatemala

(18clusters)

Zambia(10clusters)

Kenya(12clusters)

Pakistan

(10clusters)

Dem

ocratic

Repu

blic

oftheCon

go(8

clusters)

Trialfun

dedby

Bill&

Melinda

Gates

Foun

datio

n.The

ultrasou

ndtrial

cond

uctedby

Global

NetworkforWom

en’s

andChildren’sHealth

Research

(sup

ported

byEunice

Kenn

edyShriver

NationalInstituteof

Child

Health

andHum

anDevelop

men

t).O

bstetric

ultrasou

ndtraining

oversigh

tby

The

University

ofWashing

ton

Dep

artm

entof

Radiolog

ywith

supp

ortfro

mGE

Health

care

Kim et al. Reproductive Health (2018) 15:129 Page 5 of 26

Table

1QualityAssessm

entof

Stud

iesrelatedto

OutcomeMeasures(Con

tinued)

Autho

r,Year

Locatio

nStud

yPerio

dType

ofUltrasou

ndStud

yDescriptio

nKeyPo

ints

SampleSize

Fund

ingor

Supp

ort

major

obstetric

and

neon

atalcond

ition

s.Meetin

gswith

ministry

ofhe

alth

officialsand

hospitaladm

inistratorsin

orde

rto

improvehe

alth

system

managem

ent.

Com

mun

ityorientations

inorde

rto

raiseaw

aren

ess

abou

ttheavailabilityof

serviceandthepu

rpose

ofultrasou

nd.

ultrasou

nddiagno

sisand

71%

followed

upwith

the

referrals.

Ross

AB,DeStig

ter

KK,C

outin

hoA,

SouzaS,Mwatha

A,

MatovuA,etal.

2014

[26]

Ugand

a2years

Low

cost,p

ortable

ultrasou

ndmachine

sQuantitativestud

y-no

controlg

roup

includ

ed.

Historicalcontrold

atawas

used

asacomparison

.Wom

enreceived

ultrasou

ndscan

atthefirstANCvisit

andat

32weeks

gestation.

Basedon

need

,ultrasou

ndscanswerepe

rform

edat

othe

rANCvisitsas

well.

Twomidwives

atthestud

yhe

alth

center

weretraine

dfor3days.Traininginclud

edclassroo

mandhand

s-on

sessions

with

acompe

tency

assessmen

t.Thefocusof

theprog

ram

was

ondiagno

sing

multip

lege

stations,seq

uelaeof

abortio

n,causes

ofob

structed

labo

rand

placen

taprevia.

Com

parin

gpo

st-

interven

tionultrasou

ndgrou

pto

pre-interven

tion

historicalcontrol:

Therelativerate

ratio

sfor

anem

iatreatm

ent,

RR=1.26

(1.15–1.38);

dewormingtreatm

ent,

RR=1.21

(1.11–1.32);and

IPT2

treatm

ent,RR

=2.13

(1.35–3.35)wereall

sign

ificantlyhigh

erthan

1.Therelativerate

ratio

forHIV

testingwas

not

statisticallysign

ificant

andtherelativerate

ratio

forIPT1

treatm

ent,

RR=0.88

(0.79–0.98)w

assign

ificantlylower

than

1.

Priorto

introd

uctio

nof

ultrasou

nd:

Ane

mia–34

mon

ths

Dew

orming–

34mon

ths

HIV

testing–

24mon

ths

IPT1–40mon

ths

IPT2–36mon

ths

After

introd

uctio

nof

ultrasou

nd:

Ane

mia–22

mon

ths

Dew

orming–

23mon

ths

HIV

testing–

23mon

ths

IPT1–23mon

ths

IPT2–23mon

ths

Stud

ysupp

ortedby

Bill

andMelinda

Gates

Foun

datio

n,Sano

fi,Ph

ilips

Health

care,

McKessonCorpo

ratio

nandtheFine

berg

Foun

datio

n

Kim et al. Reproductive Health (2018) 15:129 Page 6 of 26

Table

2Summaryof

Stud

ieson

Ultrasou

nd’sIm

pact

onPatient

CareManagem

ent

Autho

r,Year

Locatio

nStud

yPerio

dType

ofUltrasou

ndStud

yDescriptio

nKeyPo

ints

SampleSize

Fund

ingor

Supp

ort

BussmannH,Koe

nE,

Arhin-Ten

korang

D,

Mun

yadzweGand

Troe

gerJ,2001

[4]

Botswana

18mon

ths

Real-tim

eultrasou

ndscanne

rSiem

ensSL-1,

with

a3.5MHz/5MHz

sector

prob

eanda

5–MHzlinearprob

e.

Quantitativestud

y-no

controlg

roup

includ

ed.

Clinicaldiagno

siscompared

with

ultrasou

nd-aided

diagno

siswith

outblinding

.Clinicians

madeaprovisional

clinicaldiagno

sis.Then

,the

sono

graphe

rmadean

ultrasou

nd-aided

diagno

sis

with

know

ledg

eof

the

clinicaldiagno

sis.

Patientswereen

rolled

basedon

apre-de

velope

dindicatio

nlist.Thelist

includ

edcond

ition

srelevant

tothestud

yho

spitalw

here

ultrasou

nddiagno

siscould

beuseful:obstetric,

cysticandsolid

masses

differentiatio

n,urinaryor

biliary

obstruction,de

tection

offluid

inpe

riton

eal,pleural

orpe

ricardialcavity

orventricular

brainspace,etc.

Diagn

osisby

ultrasou

ndwas

catego

rized

aseither

“sim

ilarim

proved

”,“differen

tim

proved

”or

“not

improved

”.“Sim

ilarim

proved

”indicated

ultrasou

ndprovidinga

diagno

sisof

high

certainty

whe

nclinicaldiagno

siswas

oflow

certainty.“Differen

tim

proved

”indicated

ultrasou

ndprovidinga

diagno

sisof

high

certainty

whe

nclinicaldiagno

sishad

notbe

enascertaine

d.“Not

improved

”indicated

ultrasou

nddiagno

sisbe

ingof

equalcertainty

with

clinical

diagno

sisor

ifno

patholog

ywas

foun

d.

Overall,334of

722(45%

)of

allp

regn

antcaseshad

improved

diagno

sisdu

eto

ultrasou

ndexam

ination.

Ofthe334im

proved

cases,

201(27%

)werecatego

rized

as“sim

ilarim

proved

”and133

(18%

)werecatego

rized

as“differen

tim

proved

”.For49

cases,care

managem

entchange

dim

med

iatelyafterthe

ultrasou

nddiagno

sis.

Thisstud

yshow

edthat

ultrasou

ndexam

inations

canaidclinicaldiagno

sis

andthereb

ycontrib

uteto

improvingthequ

ality

ofcare

inadistrictho

spital.

2309

patients

Life

Sciences

andTechno

logy

forDevelop

ingCou

ntries

prog

ram

oftheCom

mission

oftheEurope

anCom

mun

ities

supp

ortedthestud

y.

Kimbe

rlyHH,M

urray

A,M

ennickeM,

LiteploA,Lew

J,Bo

hanJS,etal.

2010

[6]

Zambia

6mon

ths

Sono

Site

180po

rtable

ultrasou

nd,w

ithan

extra

battery,acurved

array

abdo

minalprob

eand

softwareforim

age

storing

Quantitativestud

y-no

controlg

roup

includ

ed.

Observedexam

inations

cond

uctedat

midterm

and

attheen

d.“M

ostrecorded

scanswerepe

rform

edin

second

andthird

trim

ester.”

From

atotalo

f441

ultrasou

ndscansthat

werepe

rform

ed,74

ultrasou

ndscans(17%

)ledto

achange

inclinical

decision

-making.

Ofthecasesthat

hada

441scans

perfo

rmed

Sono

Site

donatedultrasou

ndmachine

s

Kim et al. Reproductive Health (2018) 15:129 Page 7 of 26

Table

2Summaryof

Stud

ieson

Ultrasou

nd’sIm

pact

onPatient

CareManagem

ent(Con

tinued)

Autho

r,Year

Locatio

nStud

yPerio

dType

ofUltrasou

ndStud

yDescriptio

nKeyPo

ints

SampleSize

Fund

ingor

Supp

ort

21midwives

unde

rwen

tthreetraining

perio

dswhich

were2to

3weeks

inleng

thandwereseparatedby

2to

3mon

thsof

inde

pend

ent

scanning

.Trainingtook

place

intheform

ofdidactic

sessions,p

racticalhand

son

sessions

andsupe

rvised

scanning

.Thetraining

conten

tinclud

ediden

tificationof

fetalp

resentation,fetal

heartrate,p

lacentallocatio

n,multip

lege

stations,and

estim

ationof

gestationalage

basedon

biparietald

iameter

andfemur

leng

th.

change

inclinical

managem

ent,no

n-vertex

presen

tatio

n,multip

lege

stations,lackof

fetalh

eart

rate

andlow

lyingplacen

taconsistedover

95%

ofthe

finding

s.

Kotlyar

SandMoo

reCL,2008

[11]

Libe

ria5weeks

Apo

rtableLO

GIQeUS

unit,with

a2.5-MHz

curviline

arprob

e,a

5-MHzph

ased

array

cardiacprob

e,a7-MHz

linearprob

e,anda

4.5-MHzen

dovaginal

prob

e.Dop

pler

and

colorflow

capabilities

wereavailable.

Quantitativestud

y-no

controlg

roup

includ

ed.

Pre-ultrasou

nddiagno

sis

comparedwith

Afte

r-ultrasou

nddiagno

sis.

“Ultrasou

ndim

ages

were

acqu

iredby

reside

ncy

traine

dem

erge

ncy

physicians

aspartof

routinepatient

care.U

pon

physicians’req

uests,

ultrasou

ndwas

perfo

rmed

oninpatientsin

themed

ical,

pediatric,surgicaland

obstetricalwards

aswellas

patientsin

theem

erge

ncy

room

.”

Ultrasou

ndexam

ination

change

dpatient

managem

entin

62%

ofcases.

Ofthecasesthat

hada

change

inpatient

managem

ent,thehigh

est

impact

was

seen

with

first

trim

esterob

stetric

cond

ition

s,focusedassessmen

tof

sono

graphy

intraumaexam

,cardiacapplications

and

second

andthird

trim

ester

obstetric

cond

ition

s.

102patients

GEMed

icalloaned

ultrasou

ndeq

uipm

ent

SteinW,Katun

daI

andBu

toto

C,2008

[27]

Tanzania

12mon

ths

Siem

ensSL-1,w

itha

3.5MHz/5MHzsector

prob

e,8MHzvaginal

prob

e.

Quantitativestud

y-no

controlg

roup

includ

ed.

Clinicaldiagno

siscompared

with

ultrasou

nd-aided

diagno

sis.

Thefirst-levelu

ltrasou

ndwas

rand

omlycond

uctedon

patientswith

absent

fetal

heartbeatandsuspectfetal

positio

nothe

rthan

ceph

alic.

Forsuspectedtw

insand

vaginalb

leed

ing,

itwas

regu

larly

re-examined

.The

second

-leveladvanced

Duringthestud

ype

riod,

542patientswereen

rolled.

Ofwhich,ultrasou

nddiagno

sisim

proved

clinical

managem

entin

212(39.1%

)of

thecases,conflictedwith

originalclinicaldiagno

sisin

81cases(14.9%

)andledto

achange

ofmanagem

entin

121cases(22.3%

).

542patients

Not

Men

tione

d

Kim et al. Reproductive Health (2018) 15:129 Page 8 of 26

Table

2Summaryof

Stud

ieson

Ultrasou

nd’sIm

pact

onPatient

CareManagem

ent(Con

tinued)

Autho

r,Year

Locatio

nStud

yPerio

dType

ofUltrasou

ndStud

yDescriptio

nKeyPo

ints

SampleSize

Fund

ingor

Supp

ort

ultrasou

ndwas

perfo

rmed

byaspecialistsono

graphe

rfor

patientswith

fund

alhe

ight

discrepancies,suspected

incompleteabortio

n,suspectedextrauterine

preg

nancyandabdo

minal

pain.

Midwives

received

two

mon

thsof

training

inbasic

obstetric

ultrasou

ndto

perfo

rmfirst-levelu

ltrasou

ndservices.

Shah

SP,Epino

H,

Bukhman

G,U

mulisaI,

Dushimiyim

anaJM

,Reichm

anA,etal.

2009

[14]

Rwanda

19weeks

+Sono

site

Micromaxx,with

endo

cavitary

prob

e,a

curved

arrayabdo

minal

prob

eandcardiacprob

es

Quantitativestud

y-no

controlg

roup

includ

ed.

Blinde

dreview

ofultrasou

ndscansforaccuracy

and

quality.

Datashe

etsforeach

ultrasou

ndscan

perfo

rmed

durin

groutineclinicalcare

werecollected

andanalyzed

.Localp

hysician

staff

unde

rwen

ta9-week

ultrasou

ndtraining

curriculum

that

includ

edlectures

andpractical

hand

s-on

sessions.

Lecturetopics

includ

edob

stetricalultrasou

nd,cardiac

ultrasou

nd,hep

ato-biliary

ultrasou

ndandothe

radvanced

uses

ofultrasou

nd.

Providersat

thestud

ysites

indicatedthat

ultrasou

ndscanschange

dtheirinitial

patient

managem

entplan

in43%

ofthecases.

Themostcommon

lyrepo

rted

change

afterthe

ultrasou

ndwas

the

considerationto

perfo

rma

surgicalproced

uresuch

asa

cesarean

section,biop

syor

minor

surgery.Cesarean

sections

werede

cide

difthe

patient

hadbreech

presen

tatio

n,placen

taprevia

ormultip

lege

stations.

345ultrasou

ndscans

Sono

Site

donatedultrasou

ndmachine

s

WylieBJ,Kalilani-Phiri

L,MadanitsaM,

Mem

beG,N

yirend

aO,M

awindo

P,et

al.

2013

[8]

Malaw

i4mon

ths+

Sono

Site

180po

rtable

ultrasou

ndQuantitativestud

y-no

controlg

roup

includ

ed.

Com

parison

ofultrasou

nd-

aide

ddatin

gandmen

strual

gestationalage

/postnatal

Ballard

estim

ation.

“61.8%

ofthesubjectswere

enrolledandim

aged

betw

een20

and27

weeks.

21.3%

wereim

aged

after

28weeks.16.3%

were

imaged

priorto

20weeks.”

8traine

es(fo

urresearch

staff,

onenu

rse,threemidlevel

clinicians)receivedan

intensiveon

e-week

Ultrasou

ndscanswereused

toconfirm

men

strual

gestationalage

in62.1%

ofthecases.

13%

wereno

taw

areof

their

lastmen

strualdatesandthus

hadtheirge

stationalage

calculated

solelyby

ultrasou

nd.

24.9%

ofthecaseshadto

have

theirge

stationalage

re-

datedby

ultrasou

nd.

Forover

athird

ofthe

patients,ultrasou

ndscans

played

acriticalrolein

improvingtheaccuracy

of

178patients

Ultrasou

ndmachine

sdo

nated

byVincen

tDep

artm

entof

ObstetricsandGynecolog

yat

theMassachusettsGen

eral

Hospitalw

ithamatching

grantfro

mtheSoun

dCaring

Prog

ram

(Son

osite).Research

supp

ortedby

theDorisDuke

CharitableFoun

datio

nandby

theNationalInstituteof

Health

.

Kim et al. Reproductive Health (2018) 15:129 Page 9 of 26

Table

2Summaryof

Stud

ieson

Ultrasou

nd’sIm

pact

onPatient

CareManagem

ent(Con

tinued)

Autho

r,Year

Locatio

nStud

yPerio

dType

ofUltrasou

ndStud

yDescriptio

nKeyPo

ints

SampleSize

Fund

ingor

Supp

ort

ultrasou

ndtraining

forfetal

biom

etry.Fou

rmon

thsof

additio

nalp

racticeandre

moteim

agereview

followed

.

theirge

stationalage

.BecauseMalariadu

ring

preg

nancyisariskfactor

for

low

birthweigh

t,accurate

datin

gby

ultrasou

ndcan

facilitateapprop

riate

clinical

managem

ent

Spen

cerJK

andAdler

RS,2008[28]

Ghana

1mon

thALO

GIQ

Book

portable

ultrasou

nd,w

ithalinear

L5–10MHzand

curviline

arC3–5MHz

broadb

andtransducers.

Quantitativestud

y-no

controlg

roup

includ

ed.

Determinationof

whe

ther

ultrasou

ndaide

dclinical

diagno

sis.

Patientswereseen

bya

radiolog

isttraine

din

gene

ral

abdo

minalultrasou

nd,

obstetric

ultrasou

nd,small

partsandmusculoskeletal

ultrasou

nd,and

othe

rproced

ures

byultrasou

ndin

avariety

ofclinicalsettings:a

surgeo

n’soffice,ho

spital

operatingroom

andclinics.

Diagn

oses

includ

edbo

thinfectious

andno

n-infectious

diseases.

Ofthetotal67ultrasou

ndscanspe

rform

ed,54(81%

)of

them

wereconsidered

tobe

abno

rmalfinding

s.Allof

theabno

rmalfinding

sen

hanced

clinicaldiagno

sis

and40%

either

influen

ced

theou

tcom

eor

theclinical

decision

-makingregarding

treatm

ent.

67patients

GEHealth

care

provided

aLO

GIQ

Book

portable

ultrasou

nd

Steinm

etzJP

and

Berger

JP,1999[12]

Cam

eroo

n16

mon

ths

Multip

urpo

seALO

KA256,

with

twolinearprob

esof

3.5and5MHz

Quantitativestud

y-no

controlg

roup

includ

ed.

Com

parison

betw

eenthose

inwhich

thediagno

sisby

ultrason

ograph

ywas

confirm

edby

acertified

diagno

sisandthosein

which

thediagno

sisby

ultrason

ograph

ycouldno

tbe

confirm

ed.

Ultrasou

ndexam

swere

requ

estedby

anu

rseor

ado

ctor

andpe

rform

edby

onesurgeo

n-echo

graphist.

Indicatio

nsforultrasou

ndexam

sinclud

edgyne

cology,

liver,spleenandbiliary

tract,

gastrointestinaltract,

preg

nancy,renal-u

rinarytract

andothe

rs.

Intheretrospe

ctivereview

of1119

ultrasou

ndscans,78%

ofthescanswerede

emed

asabno

rmalfinding

s.67.8%

werejudg

eduseful

inclinicaldiagno

sisand

managem

entandon

ly4.6%

werejudg

edto

beerroneou

s.In

additio

n,forne

arlyahalf

ofthecasesthat

hadbe

enconfirm

edby

tissue

patholog

y,additio

nal

imagingtests,en

doscop

y,surgicalspecim

enor

labo

ratory

diagno

sis,

ultrasou

ndscanssugg

esteda

differentiald

iagn

osiswhich

hadno

tbe

enpreviously

considered

.

1119

ultrasou

ndscans

SamaritanHospitalp

rovide

dtheultrasou

ndmachine

.Foun

datio

nde

Jumelage

Oon

aChaplin

provided

fund

ingsupp

ort.

Kim et al. Reproductive Health (2018) 15:129 Page 10 of 26

Table

2Summaryof

Stud

ieson

Ultrasou

nd’sIm

pact

onPatient

CareManagem

ent(Con

tinued)

Autho

r,Year

Locatio

nStud

yPerio

dType

ofUltrasou

ndStud

yDescriptio

nKeyPo

ints

SampleSize

Fund

ingor

Supp

ort

ShorterM

andMacias

DJ,2012

[16]

Haiti

12days

Sign

oshand

held

ultrasou

nd,w

itha

3.5MHzanda7.0MHz

removableprob

e.

Quantitativestud

y-no

controlg

roup

includ

ed.

Retrospe

ctive,no

n-blinde

d,ob

servationalanalysisof

ultrasou

ndscans.

Ultrasou

ndwas

perfo

rmed

byon

eregistered

diagno

stic

med

icalsono

graphe

r-eligible

emerge

ncyph

ysician

sono

graphe

r(with

only

8hof

priorexpe

rience

with

thehand

held

device)

“Ultrasou

ndwas

perfo

rmed

forcomplaintsor

symptom

sindicativeof

illne

sses

that

couldpo

tentially

betriage

dor

diagno

sedby

ultrasou

nd(internalbleeding

,preg

nancy,shockandcardiac

dysfun

ction).”

Theuseof

portablehand

-he

ldultrasou

ndde

vicesled

toachange

in70%

ofpatient

managem

entplans.

Portableultrasou

ndim

aging

was

foun

despe

ciallyuseful

forno

n-traumaticabdo

minal

pain

andpreg

nancy-related

symptom

s.

50patients

Not

Men

tione

d

Kim et al. Reproductive Health (2018) 15:129 Page 11 of 26

Table

3Summaryof

theLiterature

onUltrasou

ndTraining

Prog

rams

Autho

r,Year

Locatio

nStud

yPerio

dType

ofUltrasou

ndStud

yDescriptio

nKeyPo

ints

SampleSize

Fund

ingor

Supp

ort

StantonKandMwanriL,

2013

[7]

Low

resource

settings

Not

App

licable

Not

Men

tione

d(Variesby

Stud

y)System

aticreview

.Midwives

andbirth

attend

antscanbe

traine

din

ultrasou

nd

32articles

Not

Men

tione

d

LaGrone

LN,Sadasivam

V,Ku

shne

rALandGroen

RS,2012[18]

Low-andmiddle-

incomecoun

tries

Not

App

licable

Not

Men

tione

d(Variesby

Stud

y)Review

article.

Gen

eralistandob

stetric

physicians

perfo

rmmost

ultrasou

ndscans.

Thetype

sof

ultrason

ograph

ytraining

rang

edfro

m“noform

altraining

toform

alcertificatio

nandreside

ncy

prog

rams.”

(p.808)

Allprog

ramshadcourses

that

consistedof

didactics

andhand

s-on

sessions.

Thetype

sof

follow-up

training

rang

edfro

m“

none

,totelemed

icine

case

review

,toform

alre-evaluations

andintensive

refre

sher

courses.”

(p.809)

Alth

ough

ultrasou

ndtraining

inLM

ICsoften

does

notmeettheWHO

criteria,selectprog

rams

repo

rted

high

levelsof

diagno

sticaccuracy

and

traine

es’kno

wledg

eretention.

41articlesforfinal

review

Not

Men

tione

d

Carrera

JM,2011[31]

Africa

Not

App

licable

Not

Men

tione

d(Variesby

Stud

y)Review

article.

Mostpractitione

rsin

Africa

areno

twell-trained

inultrason

ograph

y.40.4%

have

onlyreceived

atheo

reticalshortcourse.

38.3%

didno

thave

training

.14.9%

received

apractical

course.

Not

Men

tione

dNot

Men

tione

d

Kimbe

rlyHH,M

urrayA,

Men

nickeM,LiteploA,

Lew

J,Bo

hanJS,etal.

2010

[6]

Zambia

6mon

ths

Sono

Site

180po

rtable

ultrasou

nd,with

anextrabattery,acurved

arrayabdo

minal

prob

eandsoftware

forim

agestoring

Quantitativestud

y-no

controlg

roup

includ

ed.

Observedexam

inations

cond

uctedat

midterm

andat

theen

d.“M

ostrecorded

scans

werepe

rform

edin

second

andthird

trim

ester.”

(p.1269)

21midwives

unde

rwen

tthreetraining

perio

ds

96%

agreem

entbe

tween

expe

rtreview

ersand

traine

dmidwives

onInterpretatio

nof

fetal

heartrate.

91%

agreem

entbe

tween

expe

rtreview

ersand

traine

dmidwives

oniden

tificationof

placen

tallocation.

70%

disagreemen

t

441scans

perfo

rmed

Sono

Site

donated

ultrasou

ndmachine

s

Kim et al. Reproductive Health (2018) 15:129 Page 12 of 26

Table

3Summaryof

theLiterature

onUltrasou

ndTraining

Prog

rams(Con

tinued)

Autho

r,Year

Locatio

nStud

yPerio

dType

ofUltrasou

ndStud

yDescriptio

nKeyPo

ints

SampleSize

Fund

ingor

Supp

ort

which

were2to

3weeks

inleng

thandwere

separatedby

2to

3mon

thsof

inde

pend

ent

scanning

.Trainingtook

placein

theform

ofdidacticsessions,p

ractical

hand

son

sessions

and

supe

rvised

scanning

.Thetraining

conten

tinclud

ediden

tificationof

fetalp

resentation,fetal

heartrate,p

lacental

locatio

n,multip

lege

stations,and

estim

ationof

gestational

agebasedon

biparietal

diam

eter

andfemur

leng

th.

betw

eenexpe

rtreview

ers

andtraine

dmidwives

onmeasuremen

tof

biparietal

diam

eter.

Clinicalde

cision

-making

change

din

17%

ofthe

cases.

WylieBJ,Kalilani-PhiriL,

MadanitsaM,M

embe

G,

Nyirend

aO,M

awindo

P,et

al.2013[8]

Malaw

i4mon

ths+

Sono

Site

180po

rtable

ultrasou

ndQuantitativestud

y-no

controlg

roup

includ

ed.

Com

parison

ofultrasou

nd-aided

datin

gandmen

strual

gestationalage

/postnatal

Ballard

estim

ation.

“61.8%

ofthesubjects

wereen

rolledand

imaged

betw

een20

and

27weeks.21.3%

were

imaged

after28

weeks.

16.3%

wereim

aged

prior

to20

weeks.”

8traine

es(fo

urresearch

staff,on

enu

rse,three

midlevelclinicians)w

ent

throug

han

intensiveon

e-weekultrasou

ndtraining

forpe

rform

ingfetalb

iom

etry.Fou

rmon

thsof

additio

nalp

racticeand

remoteim

agereview

followed

.

Only5.7%

oftheultrasou

ndscanswith

abiom

etric

parameter

wereconsidered

unacceptable.

Ultrasou

ndhe

lped

improve

gestationalage

datin

gin

over

athird

oftheresearch

subjects(com

paredto

men

strualdatin

g).

Themed

iange

stationalage

determ

ined

byultrasou

nddatin

gandby

postnatal

Ballard

estim

ationwere

sign

ificantlydifferent.

Thetraine

esdidno

tbe

comeproficient

immed

iatelyafterthe

firston

e-weektraining

course.A

nadditio

nalfou

rmon

thsof

practiceand

remoteim

agereview

were

paramou

ntin

achievingthe

necessaryskill.

178patients

Ultrasou

ndmachine

sdo

natedby

Vincen

tDep

artm

entof

Obstetrics

andGynecolog

yat

the

MassachusettsGen

eral

Hospitalw

ithamatching

grantfro

mthe

Soun

dCaringProg

ram

(Son

osite).Research

supp

ortedby

theDoris

DukeCharitable

Foun

datio

nandby

the

NationalInstituteof

Health

.

Rijken

MJ,LeeSJ,Boe

lME,Papage

orgh

iouAT,

Visser

GH,D

wellSL,et

al.

2009

[32]

Thai-Burmesebo

rder

6mon

ths

ToshibaPo

wervision

7000

machine

with

a3.75-M

Hzconvex

prob

e

Quantitativestud

y-no

controlg

roup

includ

ed.

Intraobserverand

interobserver

variatio

nsmeasured.

Ultrasou

ndwas

Com

paredwith

theskilled

sono

graphe

r,thetraine

es’

fetalanthrop

ometric

measuremen

tsshow

edhigh

levelo

fagreem

ent.

349patients

Ultrasou

ndscanne

rdo

natedby

Ph.

Stou

tenb

eckandthe

Dep

artm

entof

Obstetrics

oftheUniversity

Med

ical

Cen

terUtrecht,The

Kim et al. Reproductive Health (2018) 15:129 Page 13 of 26

Table

3Summaryof

theLiterature

onUltrasou

ndTraining

Prog

rams(Con

tinued)

Autho

r,Year

Locatio

nStud

yPerio

dType

ofUltrasou

ndStud

yDescriptio

nKeyPo

ints

SampleSize

Fund

ingor

Supp

ort

perfo

rmed

oneveryfifth

preg

nant

wom

anwith

asing

letonpreg

nancy

betw

een16

and

40weeks’g

estatio

nand

who

hadun

dergon

ean

early

datin

gultrasou

ndscan,atten

ding

the

antenatalclinic.

Four

localh

ealth

workers

received

a3-mon

thcourse

consistin

gof

practicalandtheo

retical

training

inob

stetric

ultrasou

nd.

Thetraining

curriculum

was

develope

dbasedon

theWorld

Health

Organizationgu

idelines

andBritish

Med

ical

Ultrasou

ndSociety

recommen

datio

ns.

Nethe

rland

s.NIHR

Biom

edicalResearch

Prog

ram

fund

edA.T.P.

KozukiN,2015[5]

KozukiN,M

ullany

LC,

Khatry

SK,G

himire

RK,

Paud

elS,Blakem

oreK,

etal.2016[52]

Nep

al7mon

ths

Sono

site

Nanom

axx

system

,and

aC60n

(obstetric)prob

e

Quantitativestud

y-

controlg

roup

includ

edforcomparin

gfacility

deliveryratesandrate

ofadverseou

tcom

es.

Preg

nant

wom

enwho

werege

stationalage

32weeks

ormoreand

consen

tedto

participate

inthestud

ywerevisited

intheirho

mes.

Threeauxiliary

nurse

midwives

received

two

one-weekultrasou

ndtraining

sessions

which

wereseparatedby

amon

th.Trainingtook

placein

theform

oflectures,d

emon

stratio

nsandpractice.

Thetraining

conten

tinclud

ediden

tification

offetalp

resentation,

fetalh

eartbe

at,m

ultip

lege

stationandplacen

tal

positio

n.

The“canno

tde

term

ine”

selectionby

tworeview

erteam

sforfetalp

resentation

was

0.1and0.3%

ofthe

ultrasou

ndexam

s.The“canno

tde

term

ine”

selectionby

tworeview

erteam

sformultip

lege

stationwas

0.9and6.6%

oftheultrasou

ndexam

s.The“canno

tde

term

ine”

selectionby

tworeview

erteam

sforplacen

taprevia

was

34and44%

ofthe

ultrasou

ndexam

s.

804wom

enChildren’sPrize,the

NationalInstitutes

ofHealth

/NationalInstitute

ofChild

Health

and

Hum

anDevelop

men

t,Bill

andMelinda

Gates

Foun

datio

n.Ultrasou

ndmachine

sdo

natedby

Sono

Site

Soun

dcaring

Prog

ram.

Kim et al. Reproductive Health (2018) 15:129 Page 14 of 26

Table

3Summaryof

theLiterature

onUltrasou

ndTraining

Prog

rams(Con

tinued)

Autho

r,Year

Locatio

nStud

yPerio

dType

ofUltrasou

ndStud

yDescriptio

nKeyPo

ints

SampleSize

Fund

ingor

Supp

ort

SteinW,Katun

daI

andBu

toto

C,2008

[27]

Tanzania

12mon

ths

Siem

ensSL-1,w

itha

3.5MHz/5MHzsector

prob

eanda8MHz

vaginalp

robe

Quantitativestud

y-no

controlg

roup

includ

ed.

Clinicaldiagno

sis

comparedwith

ultrasou

nd-aided

diagno

sis.

Thefirst-levelu

ltrasou

ndwas

rand

omlycond

ucted

onpatientswith

absent

fetalh

eartbe

atand

suspectfetalp

osition

othe

rthan

ceph

alic.For

suspectedtw

insand

vaginalb

leed

ing,

itwas

regu

larly

re-examined

.Thesecond

-levelad

vanced

ultrasou

ndwas

perfo

rmed

byaspecialist

sono

graphe

rforpatients

with

fund

alhe

ight

discrepancies,suspected

incompleteabortio

n,suspectedextrauterine

preg

nancyand

abdo

minalpain.

Midwives

received

two

mon

thsof

training

inbasicob

stetric

ultrasou

ndto

perfo

rmfirst-level

ultrasou

ndservices.

Traine

dmidwives

iden

tified

twins,fetalh

eartrate

andfetalp

osition

with

100%

agreem

entas

the

sono

graphe

r.Theresults

forvaginal

bleeding

agreed

in76.6%

ofthecasesbe

tween

midwives

andthe

sono

graphe

r.

542patients

Not

Men

tione

d

BellG,W

achira

Band

Den

ning

G,2016[33]

Kenya

15mon

ths

Not

Men

tione

dQuantitativestud

y-no

controlg

roup

includ

ed.

Traine

es’m

eanscores

werecomparedacross

sessions.

Traine

esweregiven

materialsto

stud

ybe

fore

theworksho

p.They

then

participated

inthree

all-d

ayworksho

pshe

ldevery3to

5mon

ths.

Training

includ

edreview

ofconten

tforpo

int-of-

care

ultrasou

nd(“the

abdo

minal,p

leuraland

cardiacassessmen

tfor

freefluid,the

thoracic

exam

forpn

eumotho

rax,

anob

stetric

exam

for

intrauterin

epreg

nancy,

After

theworksho

p,participantsde

mon

strated

increase

inknow

ledg

eand

acqu

isition

ofpracticalskills.

Average

scores

forpractical

skillswereno

tsign

ificantly

different

amon

gvarying

cadres

ofhe

althcare

workers.

81traine

esTheChristianHealth

Associatio

nof

Kenya

provided

training

facilities,

DAKFoun

datio

nprovided

ultrasou

ndeq

uipm

ent,

TheEm

erge

ncyMed

icine

KenyaFoun

datio

n,The

Emerge

ncyMed

icine

Dep

artm

entat

the

University

ofIowaand

TheAga

Kahn

University

Hospitalp

rovide

dfund

ingsupp

ort.

Kim et al. Reproductive Health (2018) 15:129 Page 15 of 26

Table

3Summaryof

theLiterature

onUltrasou

ndTraining

Prog

rams(Con

tinued)

Autho

r,Year

Locatio

nStud

yPerio

dType

ofUltrasou

ndStud

yDescriptio

nKeyPo

ints

SampleSize

Fund

ingor

Supp

ort

cul-d

e-sacfluid,fetal

heartactivity

and

positio

n”-p.

2),

demon

stratio

nsand

practice.Profession

alcadres

oftraine

esin

clud

edclinicalofficers,

doctorsandnu

rses.

Green

woldN,W

allace

S,ProstAandJaun

iaux

E,2014

[34]

Mozam

biqu

e12

mon

ths

Sono

site

2M-Turbo

portableultrasou

nd,

with

integrated

obstetric

biom

etric

charts

Quantitativestud

y-no

controlg

roup

includ

ed.

Detectio

nratesfor

traine

esun

der

supe

rvisionand

with

outsupe

rvision

werecompared.

Allwom

enattend

ing

pren

atalcare

betw

een

11weeks

andterm

were

offeredultrasou

ndexam

s.9nu

rses

andclinical

officersun

derw

entan

8-weektraining

course

followed

by10

mon

thsof

remotesupe

rvision.

Training

includ

ed1

weekof

form

allectures

and7weeks

ofpractical

hand

s-on

sessions.

Thetraining

conten

tinclud

edbasicuses

ofultrasou

nd,first-trim

ester

ultrasou

nd,estim

ationof

gestationalage

and

iden

tificationof

fetal

presen

tatio

n,placen

tal

positio

n,multip

lepreg

nanciesand

uterinefib

roids.

Com

parison

ofthe

detectionratesforbasic

ultrasou

ndfinding

sbe

tweenthetw

ogrou

psrevealsthat

fetalano

malies

aretheon

lycond

ition

that

theparticipantsun

der

direct

supe

rvisionde

tected

sign

ificantlyhigh

er.

Thede

tectionratesfor

preg

nancyloss/in

trauterin

efetald

eath,twin

preg

nancies,fib

roids,

placen

taprevia,b

reech

presen

tatio

nandtransverse

presen

tatio

nwere

statisticallysimilarbe

tween

thetw

ogrou

ps.

1744

preg

nant

wom

en,804

scanne

dim

ages

bytraine

esun

der

direct

supe

rvision,

940scanne

dim

ages

bytraine

esalon

e

Med

icalAid

Film

s,MaM

AMozam

biqu

eand

Sono

site

provided

financialsupp

ort

Shah

SP,Epino

H,

Bukhman

G,U

mulisaI,

Dushimiyim

anaJM

,Reichm

anA,etal.2009

[14]

Rwanda

19weeks

+Sono

site

Micromaxx,

with

endo

cavitary

prob

e,acurved

array

abdo

minalprob

eand

cardiacprob

es

Quantitativestud

y-no

controlg

roup

includ

ed.

Blinde

dreview

ofultrasou

ndscansfor

accuracy

andqu

ality.

Datashe

etsforeach

ultrasou

ndscan

perfo

rmed

durin

groutineclinicalcare

werecollected

and

analyzed

.

Betw

eentheparticipants

andtheultrasou

nd-trained

emerge

ncymed

icine

physicians,the

rewas

96%

agreem

entin

their

interpretatio

nsof

thescans.

Participantscontinuedto

useultrasou

ndafterthe

training

.

345ultrasou

ndscans

Sono

Site

donated

ultrasou

ndmachine

s

Kim et al. Reproductive Health (2018) 15:129 Page 16 of 26

Table

3Summaryof

theLiterature

onUltrasou

ndTraining

Prog

rams(Con

tinued)

Autho

r,Year

Locatio

nStud

yPerio

dType

ofUltrasou

ndStud

yDescriptio

nKeyPo

ints

SampleSize

Fund

ingor

Supp

ort

Localp

hysician

staff

unde

rwen

ta9-week

ultrasou

ndtraining

curriculum

that

includ

edlectures

andpractical

hand

s-on

sessions.

Lecturetopics

includ

edob

stetricalultrasou

nd,

cardiacultrasou

nd,

hepato-biliaryultrasou

ndandothe

radvanced

uses

ofultrasou

nd.

Sipp

elS,Murug

anandan

K,Levine

AandShah

S,2011

[10]

Develop

ingworld

Not

App

licable

Not

Men

tione

d(Variesby

Stud

y)Review

article.

“Nostandardized

approaches

availablefor

theleng

thof

training

,curriculum

forge

neral

practitione

rs,q

ualifications

oftraine

rsor

mechanism

oftraining

.”(p.4)

Theleng

thof

ultrasou

ndtraining

rang

edbe

tween

4days

andseveralm

onths.

Thecadres

oftraine

esrang

edfro

mclinical

officers,nu

rses

andnu

rse

midwives

toph

ysicians.

Traine

rsrang

edfro

mreside

ntph

ysicians,

emerge

ncyph

ysicians,

radiolog

ists,cardiolog

iststo

ultrasou

nd-fe

llowship

traine

dem

erge

ncy

physicians.

Thecurriculum

and

metho

dof

training

varied

basedon

thespecificgo

als

oftheprog

rams.

Overall,“a

shortbu

tintensivetraining

perio

dissufficien

tforprep

aring

clinicalofficers,nu

rses

and

physicians

aliketo

perfo

rmbasicultrasou

ndexam

s”(p.4),espe

ciallywhe

nbo

thdidacticsandpractical

sessions

areinclud

edand

skillsmainten

ance

isen

suredthroug

hrefre

shers.

Not

Men

tione

dNot

Men

tione

d

Kim et al. Reproductive Health (2018) 15:129 Page 17 of 26

Table

3Summaryof

theLiterature

onUltrasou

ndTraining

Prog

rams(Con

tinued)

Autho

r,Year

Locatio

nStud

yPerio

dType

ofUltrasou

ndStud

yDescriptio

nKeyPo

ints

SampleSize

Fund

ingor

Supp

ort

Becker

DM,TafoyaCA,

Becker

SL,Kruge

rGH,

Tafoya

MJandBecker

TK,

2016

[17]

Low-andmiddle-

incomecoun

tries

Not

App

licable

Variesby

Stud

ySystem

aticreview

.“Sho

rt-trainingcoursesmay

lead

tosign

ificant

know

ledg

eretentionand

improvepracticalskills,

even

whe

npriorultrasou

ndexpe

riencewas

minim

al.”

(p.307)

36manuscripts

includ

edforfinal

review

Not

Men

tione

d

SwansonJO

,Kaw

ooya

MG,SwansonDL,Hippe

DS,Dun

gu-M

atovuPand

NathanR,2014

[35]

Ugand

a10

mon

ths

GELogiqE

Ultrasou

ndandGE

LogicBo

okXP

,with

wide-band

(2.0to

5.5MHz)convex

arraytransducers

Quantitative–no

control

grou

pinclud

ed.

Expe

ctingmothe

rsreceived

ultrasou

ndexam

sas

partof

the

routineantenatal

care

visit.

14midwives

from

the

stud

yhe

alth

centers

unde

rwen

ta6-week

ultrasou

ndtraining

course

onlim

ited

obstetric

ultrasou

nd.13

ofwhich

hadno

prior

ultrasou

ndtraining

.Thetraining

curriculum

includ

ed“ultrasou

ndph

ysics,relevant

anatom

yandph

ysiology,

instrumen

tatio

nandbasic

mainten

ance.”(p.509)

Training

metho

dsinclud

ed“lectures,

small-g

roup

tutorials,

audiovisualm

aterialsand

supe

rvised

clinical

scanning

.”(p.509)

Obstetricultrasou

ndprovided

bymidwives

change

dtheclinical

diagno

sisin

upto

12%

ofthecases.

Thequ

ality

assurance

review

ofmidwives’scans

diagno

sing

gestational

numbe

rshow

ed100%

sensitivity

andspecificity.

Thequ

ality

assurance

review

ofmidwives’

scansdiagno

sing

fetal

presen

tatio

nshow

ed90%

sensitivity

and96%

specificity.

939patients

GEFoun

datio

n,University

ofWashing

tonRadiolog

yHealth

Services

Research

Seed

Grant

Prog

ram

and

Seattle

International

Foun

datio

nsupp

orted

thestud

y.

Kim et al. Reproductive Health (2018) 15:129 Page 18 of 26

power due to the small sample size [19]. Other studieseither did not have enough statistical power to detectsignificant differences in mortality outcomes [22] or sim-ply speculated that increases in antenatal care visits as aresult of ultrasound or postpartum uses of ultrasoundwill likely reduce mortality outcomes [23, 24].Some studies that employed a weak study design (lim-

ited randomization and absence of control groups) re-ported that ultrasound services generally appeared to beassociated with an increase in the number/proportion ofantenatal care use and facility delivery, which is contraryto the findings of a recent large-scale cluster-randomizedtrial [20, 21].Ross et al. conducted a study at a rural community

health center in Uganda with portable ultrasounds [23].Exams were offered to women during their first ante-natal visit and at 32 weeks gestation, and additionalscans were performed when they were deemed appropri-ate [23]. The study noted that ultrasound exam feeswere affordable to all patients at about $2 USD [23].Lacking a control group, the study utilized a historicalcontrol to assess changes in outcome measures pre andpost-intervention [23]. Total antenatal care visits at thehealth center significantly increased from a monthlyaverage of 133.5 visits before the program to 230.3 visitsafter the program [23]. Over a period of 64 months, theaverage number of deliveries at the health center alsosignificantly increased from 28.1 deliveries before theprogram to 45.6 deliveries after the program [23].Mbuyita et al. conducted a study in Tanzania to deter-

mine if mid-level providers were capable of using port-able ultrasound machines after training [25]. Comparedto baseline data, the number of first antenatal care visitsdid not change significantly [25]. However, the numberof women receiving four or more antenatal care visits in-creased significantly between baseline (27.2%) and end-line (60.3%) periods in the intervention area [25]. Thenumber of facility deliveries also increased significantlybetween baseline and endline periods in the interventionarea [25].Lastly, Ross et al. examined the components of routine

antenatal care services before and after the introductionof routine ultrasound [26]. The study showed signifi-cantly increased rates of anemia treatment, deworming,and two doses of intermittent preventive treatment ofmalaria (IPT2) after the ultrasound intervention [26].

Patient diagnosis and care managementAlthough recent evidence points to ultrasound having nosignificant effect on mortality, facility delivery and ante-natal care use [20, 21], there were studies reporting thatultrasound in LMICs helped improve the quality of carefor both obstetric and non-obstetric conditions [10].Muller-Rockstroh, in her narrative description about a

hospital in Northwest Tanzania, stated that ultrasoundscans helped midwives determine the timing and appro-priate mode of delivery. Midwives were able to delay oraccelerate timing of delivery and determine appropri-ate management plans based on ultrasound findings[24, 25]. The ability of ultrasound to confirm clinic-ally suspected obstetric complications and even im-prove patient management has been reported in manyother studies [4, 6, 8, 11, 12, 14, 27, 28].A study in Malawi reported that despite having a

learning curve, mid-level providers were able to confirmor improve determination of gestational age using ultra-sound [8]. Correct determination of gestational age isnecessary to distinguish premature newborns from new-borns who are growth-restricted as the required careand interventions are different [8]. Correct determin-ation of gestational age also has implications for prevent-ing mother-to-child transmission of HIV [7].The utility of ultrasound has been documented in med-

ical emergencies and disaster settings which include obstet-ric uses as well. Understandably, portable hand-held orhand-carried ultrasound devices have been employed inthese situations where there is unreliable access to powerand a high need for rapid triage [17]. During a 12-day disas-ter relief effort in Haiti after the earthquake in 2010, theuse of portable hand-held ultrasound devices led to achange in 70% of patient management plans [16]. Portableultrasound imaging was found especially useful for diagnos-ing non-traumatic abdominal pain and pregnancy-relatedsymptoms [16].

What are key issues related to health personnel?Training of health personnelThe training of health personnel is a major factor forquality implementation of obstetric ultrasound services,as appropriate patient management relies on the abilityof health personnel to use the machine proficiently andinterpret findings accurately [7]. In many LMICs, thereis a paucity of health personnel who are capable of pro-viding ultrasound services [29].A WHO Study Group recommended in its 1998 ultra-

sound manual that candidates for ultrasound trainingshould have at least two to three years of prior healthcaretraining [30]. Qualified candidates are then recommendedto complete at least 6 months of training in a recognizedtraining center including 50 first trimester pregnancyexams and 200 s and third trimester exams for obstetricapplications of ultrasound [30]. It is important to note,however, that recommendations for the different numbersof examinations were given as a guide. A recent reviewarticle, which documented training opportunities forultrasonography in LMICs, concluded that the majority ofhealth personnel using ultrasound in LMICs did not meetthe minimum WHO training standards [18]. In Africa,

Kim et al. Reproductive Health (2018) 15:129 Page 19 of 26

40.4% of ultrasound service providers completed only ashort theoretical course, 14.9% had practical courses withonly 2.1% occurring in a hospital environment, and 38.3%had no training at all [7, 31]. Ultrasonography traininglargely targeted lower- and mid-level practitioners such aslocal health workers, nurses and midwives as well asnon-radiologist physicians [18]. Although the literaturesuggests that past trainings were not in compliance withWHO’s recommendations, many studies demonstratedthat lower- and mid-level providers were capable of pro-viding ultrasound services resulting in improved clinicalmanagement [5, 6, 8, 14, 27, 29, 32–35]. There is alsosome evidence that general physicians at rural health facil-ities can adequately provide ultrasound services in termsof correct use and interpretation of findings [14].When adequate training materials and methods are

used, short intensive training courses have been shown toprovide significant acquisition of knowledge and practicalskills for all levels of health workers in LMICs [10, 17].There is also evidence that follow-up refresher trainingscan be effective for retention of knowledge and skills[10, 17] though the sustainability of these coursesafter donor funding is expended is not adequately docu-mented. In addition to training, however, it is importantto note that there are other issues which are essential inobtaining good clinical readings. These issues include en-suring that the machine is set for the proper application(i.e. obstetric uses) and that the use setting has appropriatelighting, temperature, electrical supply and IT require-ments [36]. It is also important to understand the limita-tions of different ultrasound models and devices. Forexample, the image quality of portable ultrasound maynot be sufficient to identify fetal anomalies and early ges-tational age.

Concerns about misuse, overuse and miscommunicationMisuse, overuse and lack of communication between healthproviders and patients have been frequently cited as con-cerns in the literature. A study in Uganda demonstratedthat over half of all obstetric scans performed were consid-ered inappropriate because they were either scans for gesta-tional age estimation outside of the recommended timeperiod, routine monitoring despite no sign of intrauterinegrowth restriction or re-assessment of the placental pos-ition because of the technical inability to determine it witha previous scan [37]. In another study in Botswana, healthproviders noted that conventional methods such as historytaking and physical examination were neglected because ofeasy access to ultrasound services [2]. There were indica-tions that even in situations where conventional methodswould have sufficed, ultrasound diagnosis was recom-mended for convenience [2]. Such inappropriate use raisesconcerns because health providers should not neglect con-ventional methods of providing care [2, 3].

Even when ultrasound services are available, they mustbe complementary to routine care [4, 7]. This is espe-cially true for obstetric use of ultrasound. For example,estimation of gestational age is more accurate early inpregnancy [8]. As pregnancy progresses, the variation offetal size increases and it becomes difficult to accuratelyestimate gestational age by ultrasound alone [8]. Sincepregnant women in LMICs often present themselves forantenatal care later in pregnancy, ultrasound datingmust be used in conjunction with conventional methodsto accurately estimate gestational age [8].As for cases of misuse, financial incentives for provid-

ing unnecessary ultrasound services have been reportedas one reason for misuse in LMICs. After the initialultrasound scan, multiple follow-up scans were oftenscheduled in private clinics as a way to increase revenue[38, 39]. Using ultrasound for financial gain or any otherreason that is not clinically-based, can significantly re-duce the cost-effectiveness of services [1] and drive upthe costs paid by women and their families.Lastly, there is evidence for lack of communication be-

tween health providers and patients. A study inBotswana reported that the average time of interactionbetween providers and patients was 15 min, of whichonly 15% were devoted to communicating with the pa-tient [2]. Health providers were preoccupied with thetechnical aspects of performing an ultrasound scan andpatients received very little attention throughout theprocess [2]. In a study in Iran, none of the study patientsreported receiving written information about the pur-pose of the ultrasound exam, 48% reported that theultrasound operator did not answer any questions and90% reported that they were never shown the image ofthe fetus on the ultrasound screen [40]. These findingsare of concern as patients’ perception of and experiencewith ultrasound are mainly determined by the quality oftheir interaction with the health providers [2].

Concerns about fetal sex determinationPrior studies in different countries document that ultra-sound has been illegally used for fetal sex determination.In India, the male-dominant sex ratio for children under5 years is thought to be associated with fetal sex determin-ation and sex-selective abortions [1]. A population-basedstudy in Delhi, India found that 56% of respondents eitherdid not know that fetal sex determination was illegal orthought that it was lawful [9, 41]. In the same study, 2.6%of respondents had ever requested fetal sex determinationand when the female sex of the fetus was disclosed, over63% of them were aborted [9, 41]. In a study in Nepal,6.8% of surveyed pregnant women received ultrasoundexams for fetal sex determination despite it being unlawful[42]. The study also found that compared to women whohad at least one live born son prior to the recent

Kim et al. Reproductive Health (2018) 15:129 Page 20 of 26

pregnancy, women with no live born sons and three ormore live born daughters had 1.55 higher odds of receiv-ing an ultrasound exam, which may suggest use for fetalsex determination [42]. Other popular reasons for deter-mining fetal sex were reported to be curiosity and prepar-ation for the baby [43, 44].Negative consequences of fetal sex determination go

beyond obvious legal and ethical violations. Incorrect as-sessment of the fetal sex can also result in negative expe-riences for women, especially in settings where sonpreference is high. Incorrect assessment of the fetal sexis not uncommon. In a study in Ghana, fetal sex was ac-curately determined in only 86.5% of the cases [45]. Astudy in Nigeria reported that incorrect assessment ofthe fetal sex, particularly for mistaking female fetuses asmales, resulted in marital conflict, physical violence fromhusbands, regret of undergoing tubal ligation, negativeperception towards ultrasound and negative feelings to-wards the newborns [46]. Such negative consequencescan be exacerbated when accreditation and regulationfor ultrasound use are limited [9, 47]. Ethical training forhealth providers on appropriate uses of ultrasound andregulations for both public and private clinics are im-portant [7].

What issues are important to patients?Patient perception on ultrasound servicesIn the studies reviewed, women in LMICs generally heldpositive views about ultrasound services [23]. Rijken etal. found that on the Thai-Burmese border, ultrasoundwas considered a tool that could increase safety duringpregnancy and childbirth [23, 48]. In a rural Botswanadistrict hospital, pregnant women showed signs of trust-ing the ultrasound results more than their own bodilysensations to confirm a live fetus [2]. The womenexpressed relief that “there was life in the baby” (p. 697)[2]. These women also referred to the ultrasound expertsas the “Whites” (p. 697) and regarded the same ultra-sound services provided by African health providers assubstandard [2].Cultural resistance to ultrasound services was not re-

ported to be a major problem. A study in Zambia re-ported that patients would often wait in line forultrasound exams, even after clinic hours were over [6].In addition, many patients who came for antenatal visitsstated that availability of ultrasound was their motivationfor attendance [6]. A qualitative study in Tanzania re-ported that the majority of women desired to receiveultrasound exams despite the lack of understandingabout the benefits and the procedures [49]. Those whodid not view ultrasound favorably thought that ultra-sound would cause harm to the fetus [49].Negative perceptions or misconceptions about ultrasound

were reported in other studies as well [40, 50]. Women

interviewed in a study in rural Kenya perceived that under-going an ultrasound exam meant there was an emergencyor a problem with the pregnancy [50]. However, all inter-viewed women perceived ultrasound to add significantvalue in reassuring the health and progress of the baby [50].In another study in Iran, 39% of women chose not toundergo an ultrasound exam due to the following reasons:ultrasound can be harmful (especially with regards to fetalmalformation); results are not important; exams are expen-sive; and busy schedules [40].Motivation for seeking ultrasound services as well as

access to ultrasound services have been documented todiffer by social class. A study in Nepal found that the ad-justed odds of receiving an ultrasound exam were 3.4times higher for women who had 7 to 10 years of educa-tion and 10.28 times higher for women who had morethan 10 years of education compared to women with noformal education [42]. In sub-Saharan Africa, access toobstetric ultrasound in rural areas has shown to be aslow as 6% [7, 31].Providers’ explanations or lack of explanations about

the ultrasound scan play a role in terms of patient per-ception and satisfaction. Women who received ultra-sound services in a rural setting in Botswana describedthat turning the light off before the examination was anunusual experience [2]. Patients normally expect exam-ination rooms to be “brightly lit” (p. 694) [2]. In a studyin southeast Nigeria, 70% of study participants reportedthat they did not interact with the sonographers, 24% re-ported being afraid prior to the scan, 11% reported beingafraid during the scan and 4% reported being afraid afterthe scan [51].

What are potential issues with the ultrasound device?MaintenanceAlthough there are no previous studies rigorously testingthe durability of ultrasound machines in LMICs, somestudies provide anecdotal accounts. Kozuki et al. describeshaving to replace refurbished portable ultrasound machinesfour times in the span of 17 weeks in a study in Nepal dueto hardware and software errors [5, 52]. In another studyby Kimberly et al., portable ultrasound machines performedwell even in excess heat, humidity, dust and long travels ina 6-month study [6]. At the 1-year follow-up visit, however,38% of the midwives who responded to a follow-up surveycited problems with the ultrasound machine [6]. Theseproblems included depletion of ultrasound gel and flicker-ing image on the ultrasound screen [6]. Another study re-ported that maintenance was not an issue during the5-month study period [14]. In fact, a study conducted in ex-treme conditions of the Amazon jungle reported that port-able ultrasound machines functioned well for twotwo-week trips [53]. However, there is limited evidence formaintenance beyond the research study period. One study

Kim et al. Reproductive Health (2018) 15:129 Page 21 of 26

describes that establishing a maintenance protocol withinroutine health systems can be long and difficult [24]. In thisstudy, such difficulty left many ultrasound machines unre-paired [24].

CostIn high-income countries, there has been many studiesassessing the cost effectiveness of using ultrasound toidentify fetal abnormalities, as demonstrated by a 2002 re-view [54]. Cost effectiveness studies are less common forLMICs. Only a few studies provided an overview of pro-gram costs and the realized benefits. Bussmann and col-leagues reported that in a district hospital in Botswana,initial capital and recurrent costs of the ultrasound wereconsidered affordable, relative to the overall unspent hos-pital budget in the study period [4]. They also found that“the marginal cost of providing an ultrasound was lessthan a quarter of providing an X-ray examination” (p.1030) [4]. The average and marginal costs of an ultra-sound diagnosis in improving patient management ap-peared to be affordable as well, 56.58 USD and 0.64 USDrespectively [4]. Lastly, the average and marginal costs ofpotentially improving a health outcome through an actualchange in therapy, were considered affordable, 260.79USD and 2.93 USD respectively [4]. Overall, the studyconcluded that providing ultrasound scans in Botswanawas financially feasible [4]. Another study in Nepal esti-mated the impact of ultrasound on the potential to avertperinatal deaths attributable to non-cephalic birth, mul-tiple birth and placenta previa [5, 52]. The study estimatedthat with an early ultrasound diagnosis, a total of 160 po-tential perinatal deaths could be averted, which translatesto about 65 USD saved per life [5, 52]. This estimationwas derived based on the assumption that all early diagno-ses will lead to the prevention of perinatal deaths [5, 52],likely overestimating the true impact. However, it also didnot adjust for potential disability life years averted by pre-venting maternal mortality or morbidity [5, 52]. Hence,the estimated impact on averting potential perinataldeaths does not capture the full range of benefits thatearly ultrasound diagnosis offers.In summary, there is a lack of high quality cost effective-

ness studies on obstetric ultrasound in the literature [54].The majority of published studies are from high-incomecountries, and they did not include discussions about lon-ger term costs or cost incurred to women [54].

DiscussionThis review examined various factors associated withintroducing obstetric ultrasound in LMICs. A recentcluster-randomized trial found that use of ultrasound inrural health centers did not impact antenatal care attend-ance, facility delivery, maternal mortality, neonatal mortalityand stillbirths [20]. A related case study further suggested

that scale-up of routine antenatal ultrasound is not war-ranted [55]. Other studies with less rigorous designs, how-ever, found that ultrasound use was associated with theincrease of antenatal care attendance [23, 25] and facilitydelivery [6, 7, 23, 25, 49]. Although antenatal ultrasounduse did not affect mortality measures, there is evidence sug-gesting that ultrasound can confirm and improve patientmanagement for both obstetric and non-obstetric condi-tions [4, 6, 8, 10–12, 14, 27, 28] . One potential concernwith the introduction of ultrasound is whether it will dis-rupt existing services routinely provided at the health facil-ity. Muller-Rockstroh describes that select district hospitalsin Tanzania decided to train nurse-midwives for ultrasounduse instead of training radiographers or radiography assis-tants [24]. This is because use of ultrasound by radiogra-phers would have disrupted the X-ray services alsoprovided by them [24]. Even when disruption of other ser-vices is not an issue, there is still a disagreement in the lit-erature about routine versus selective provision ofultrasound services. Kongnyuy and van den Broek arguethat ultrasonography should be routinely performed for allwomen because it is regarded as safe and affordable [1]. Itcould also save costs by detecting abnormalities early inpregnancy [1, 9]. Papp and Fekete support routine ultra-sound screening on the basis that 85 to 90% of congenitalmalformations occur without maternal or family ante-cedents and therefore, selective ultrasound screening maymiss a lot of cases that cannot be deducted based on pre-vious medical history [56]. Tautz and colleagues offer acounter-argument, however, that there is insufficientevidence for ultrasound to be recommended as a rou-tine screening tool [2]. Instead, they argue that selectiveuse of ultrasound during the antenatal period can com-plement diagnoses that remain uncertain after otherclinical tests have been performed [2]. Hofmeyr alsoadds that routine ultrasound services for patients withalready confirmed pathologies may risk wasting humanresources that could be allocated more efficiently else-where [9, 57].While there is a lack of consensus in the literature, the

WHO issued a brief which recommends that pregnantwomen receive only one ultrasound scan before 24 weeksgestation for accurate determination of gestational age,early identification of fetal anomalies and multiple preg-nancies, appropriate preparation and management ofpreterm and post-term births and helping create a posi-tive pregnancy experience [58]. For those who did notreceive a scan during this period, a later scan may beconsidered to determine fetal number, presentation andplacental location [58]. However, ultrasound may beused more than once depending on the specific patientcondition.Additional guidelines and considerations are discussed in

the literature for large-scale implementation of ultrasound.

Kim et al. Reproductive Health (2018) 15:129 Page 22 of 26

Health personnel trainingMid- and lower-level health providers demonstrated com-petence in using ultrasound and in making accurate diag-nosis with only a short intensive training [5, 6, 8, 10, 14, 17,27, 29, 32–34]. Although past training programs werelargely successful, a few points have been consistentlyhighlighted in the literature for improving future endeavors.First, because past training programs varied greatly in

duration and content, there is a need for an internation-ally recognized standard of training or a certificate pro-gram [59, 60]. In addition, an ideal training modulewould minimize the interruption to local health workers’schedules and allow for practical hands-on sessions andapplications [59, 61].Second, training programs should incorporate manage-

ment lessons such as short-term and long-term mainten-ance of the ultrasound machine, image storage, imagereview and quality assurance as a part of the curriculum[18, 59]. Attention to these management topics will in-crease the likelihood that ultrasound machines are con-tinuously being used after the training has concluded,particularly in rural and isolated locations [18, 59].Third, training programs should ensure that continuing

education is available for participants after the short inten-sive training courses conclude [59]. This may be in theform of follow-up refresher sessions, direct supervision,review of donated textbooks, review of publicly accessibleor affordable journals, presentations and teleconferencing[3]. Remote learning and supervision via teleconferencingcould potentially reduce costs and have a positive impacton the quality of care [62]. Providers in LMICs have alsoshown receptiveness to the possibility of distance learningmodules [7, 60]. This option is especially appealing forrural health workers because commitment to lengthytraining programs and continuous in-person follow upsessions may be unrealistic [61].Fourth, training programs should sensitize health pro-

viders in legal and ethical conduct regarding ultrasounduse and towards patients’ desire for being informedabout their care [2]. This is because overuse and misuse[2, 37–39], determining fetal sex [1, 5, 9, 41] and thelack of communication between health providers and pa-tients were found to be major concerns regarding ultra-sound use in LMICs [2].In the long term, a more sustainable solution would be

to involve local radiology societies and training institu-tions to increase support for locally trained specializedsonographers [7, 63]. This trend is already underway asa number of training institutions and associations havebeen established in Africa in recent years [7].

Ultrasound acquisition and maintenanceSeveral factors must be considered when acquiring ultra-sound machines and setting up maintenance protocols

[59, 64]. First, various features of the ultrasound ma-chine – image quality, level of radiation and safety, easyand robust operability, transducers, ultrasound gel, sup-plies for cleaning, maintenance service and storage mustbe thoroughly considered [59, 64]. Second, and of keyimportance is the selection of an ultrasound machinethat serves the purpose for which it is intended (e.g. thequality of the image resolution determines what can bediagnosed with a particular machine). Third, there mustbe a comprehensive assessment of the practice environ-ment in which ultrasound will be used [64]. The reliabil-ity of electricity supply, the volume of patients receivedat the health facility, and the intended use/types of diag-noses expected may determine the local preference forfull-size versus portable ultrasound machines [59]. Forexample, a high volume hospital with a more reliablesource of electricity might prefer a full-size, plug-inultrasound because of the larger size rendering protec-tion against theft. For smaller health clinics operating inrural areas, using a battery-operated portable ultrasoundmachine may be more practical. Even with portableultrasound machines, security measures can be taken toguard against theft or damage. These measures includestoring the machines in a locked room or placing itunder constant observation when there is a high demandfor use [55, 59]. For health facilities that already ownfull-size ultrasound machines, however, purchasing port-able ultrasound machines might be redundant [3]. Third,durability of ultrasound machines as well as the manu-facturer’s local capacity for maintenance should be con-sidered. It may also be important to train in-housemechanics to take primary responsibility over minor re-pairs [59].

Cost-effectivenessRecent studies that included descriptions about theirprogram’s costs and the realized benefits concluded thatproviding ultrasound services was financially feasibleand cost-effective [4, 5]. Yet, there is still a lack of highquality cost-effectiveness studies conducted in LMICs[54]. This calls for more studies to be conducted withrigorous designs [54], so that a true comparison of theadded value of the service and the intervention costsversus the counterfactual can be made. It is also essentialto define the package of services in which effectivenessis evaluated. Key questions to consider are how recom-mended antenatal ultrasound services [58] would be in-tegrated into existing maternity services most effectivelyand what potential trade-offs exist with other essentialservices in resource-constrained settings. A recent casestudy from the Democratic Republic of the Congo re-ported that the success of implementing structuralchanges in the health system would rely on the level ofstakeholder effort, motivation, political will and financial

Kim et al. Reproductive Health (2018) 15:129 Page 23 of 26

and human resources available [55]. In the absence ofthese elements, streamlined integration into the healthsystem would be challenging.

Promoting cultural competenceAlthough there was a tendency of overestimating the diag-nostic capabilities of ultrasound [2, 7], patients generallyseemed to hold positive views about ultrasound services[23]. Cultural resistance to ultrasound was also not amajor issue [6, 49]. In certain cultures, however, viewingthe insides of a pregnant mother or projecting the imageof a fetus might be considered offensive [3]. Introductionof ultrasound services in these settings may require invest-ment into community engagement about the benefits ofultrasound during pregnancy [3]. As one strategy, educa-tion efforts could highlight areas where conventionalmethods are similar to the procedures of ultrasound, sothat any uneasiness related to trying new technology maybe mitigated. In Tanzania, for example, the resemblancebetween applying ultrasound gel and rubbing local medi-cine on the women’s belly for diagnosing problematicpregnancies seemed to help establish women’s trust [24].Including husbands in the intervention may also be a cul-turally appropriate strategy in some cases as one studyfound that they played a role in encouraging women toseek antenatal ultrasound exams [23].

ConclusionThis literature review focused on obstetric uses of ultra-sound in LMICs. As cost of obstetric ultrasound be-comes more affordable in LMICs, it is essential to assessthe benefits, trade-offs and potential drawbacks oflarge-scale implementation. Additionally, there is a needto more clearly identify the capabilities and the limita-tions of ultrasound, particularly within the context oflimited training of providers, to ensure that the purposefor which an ultrasound is intended is actually feasible(e.g. the image quality of portable ultrasound is not suffi-cient to identify fetal anomalies and early gestationalage). We found evidence of obstetric uses of ultrasoundimproving patient management. However, there was evi-dence that ultrasound use is not associated with redu-cing maternal, perinatal or neonatal mortality. Patientsin various studies reported to have both positive andnegative perceptions and experiences related to ultra-sound and lastly, illegal use of ultrasound for determin-ing fetal sex was raised as a concern.

FundingThis study was carried out with support provided by the United StatesAgency for International Development (USAID) through MEASURE Evaluation(cooperative agreement AID-OAA-L-14-00004). We are grateful to the CarolinaPopulation Center and its NIH Center grant (P2C HD050924) and its traininggrant (T32 HD007168) for general support.

The views and opinions expressed in this paper are those of the authors andnot necessarily the views and opinions of the United States Agency forInternational Development.

Authors’ contributionsStudy concept and design: ETK, KS, AM, DA. Acquisition of studies to review:ETK, KS, AM, DA, NK. Analysis and interpretation of the studies: ETK, KS, AM,DA. Drafting of the manuscript: ETK. Critical revision of the manuscript forimportant intellectual content: ETK, KS, AM, DA, NK. All authors read andapproved the final manuscript.

Ethics approval and consent to participateNot applicable.

Consent for publicationNot applicable.

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 details1Department of Maternal and Child Health, Gillings School of Global PublicHealth, University of North Carolina at Chapel Hill, 135 Dauer Dr, Chapel Hill,NC 27516, USA. 2Carolina Population Center, University of North Carolina atChapel Hill, Carolina Square, Suite 210, 123 West Franklin St, Chapel Hill, NC27516, USA. 3MEASURE Evaluation, Carolina Population Center, University ofNorth Carolina at Chapel Hill, Carolina Square, Suite 330, 123 West FranklinSt, Chapel Hill, NC 27516, USA. 4US Agency for International Development,1300 Pennsylvania Avenue, NW, Washington, DC 20523, USA. 5InternationalRescue Committee, 1730 M St. NW Suite 505, Washington, DC 20036, USA.6Department of International Health, Johns Hopkins Bloomberg School ofPublic Health, 615 N. Wolfe St., Baltimore, MD 21205, USA.

Received: 16 April 2018 Accepted: 13 July 2018

References1. Kongnyuy EJ, van den Broek N. The use of ultrasonography in obstetrics in

developing countries. Trop Dr. 2007;37:70–2.2. Tautz S, Jahn A, Molokomme I, Görgen R. Between fear and relief: how rural

pregnant women experience foetal ultrasound in a Botswana district hospital.Soc Sci Med. 2000;50:689–701.

3. Harris RD, Marks WM. Compact ultrasound for improving maternal andperinatal Care in low-Resource Settings: review of the potential benefits,implementation challenges, and public health issues. J Ultrasound Med.2009;28:1067–76.

4. Bussmann H, Koen E, Arhin-Tenkorang D, Munyadzwe G, Troeger J. Feasibilityof an ultrasound service on district health care level in Botswana. Tropical MedInt Health. 2001;6:1023–31.

5. Kozuki N. Epidemiology, Diagnosis, and Care-seeking related to Risk Factorsfor Intrapartum-related Fetal and Neonatal Death in Rural Nepal [Doctoraldissertation]; 2015.

6. Kimberly HH, Murray A, Mennicke M, Liteplo A, Lew J, et al. Focusedmaternal ultrasound by midwives in rural Zambia. Ultrasound Med Biol.2010;36:1267–72. https://doi.org/10.1016/j.ultrasmedbio.2010.05.017.

7. Stanton K, Mwanri L. Global maternal and child health outcomes: the role ofobstetric ultrasound in low resource settings. World journal of preventivemedicine; 2013. Available: http://pubs.sciepub.com/jpm/1/3/3/.

8. Wylie BJ, Kalilani-Phiri L, Madanitsa M, Membe G, Nyirenda O, et al.Gestational age assessment in malaria pregnancy cohorts: a prospectiveultrasound demonstration project in Malawi. Malar J. 2013;12:183. https://doi.org/10.1186/1475-2875-12-183.

9. Groen RS, Leow JJ, Sadasivam V, Kushner AL. Review: indications forultrasound use in low- and middle-income countries. Tropical Med IntHealth. 2011;16:1525–35. https://doi.org/10.1111/j.1365-3156.2011.02868.x.

Kim et al. Reproductive Health (2018) 15:129 Page 24 of 26

10. Sippel S, Muruganandan K, Levine A, Shah S. Review article: use ofultrasound in the developing world. Int J Emerg Med. 2011;4:72. https://doi.org/10.1186/1865-1380-4-72.

11. Kotlyar S, Moore CL. Assessing the utility of ultrasound in Liberia. J EmergTrauma Shock. 2008;1:10–4.

12. Steinmetz JP, Berger JP. Ultrasonography as an aid to diagnosis andtreatment in a rural African hospital: a prospective study of 1,119 cases. AmJ Trop Med Hyg. 1999;60:119–23.

13. Ahmad S, Kampondeni S, Molyneux E. An experience of emergencyultrasonography in children in a sub-Saharan setting. Emerg Med J. 2006;23:335–40. https://doi.org/10.1136/emj.2005.028761.

14. Shah SP, Epino H, Bukhman G, Umulisa I, Dushimiyimana JM, et al. Impactof the introduction of ultrasound services in a limited resource setting: ruralRwanda 2008. BMC Int Health Hum Rights. 2009;9:4. https://doi.org/10.1186/1472-698X-9-4.

15. Doehring-Schwerdtfeger E, Abdel-Rahim IM, Dittrich M, Mohamed-Ali Q,Franke D, et al. Ultrasonography as a diagnostic aid for a district hospital inthe tropics. Am J Trop Med Hyg. 1992;46:727–31.

16. Shorter M, Macias DJ. Portable handheld ultrasound in austereenvironments: use in the Haiti disaster. Prehosp Disaster Med. 2012;27:172–7. https://doi.org/10.1017/S1049023X12000611.

17. Becker DM, Tafoya CA, Becker SL, Kruger GH, Tafoya MJ, et al. The use ofportable ultrasound devices in low- and middle-income countries: asystematic review of the literature. Tropical Med Int Health. 2016;21:294–311. https://doi.org/10.1111/tmi.12657.

18. LaGrone LN, Sadasivam V, Kushner AL, Groen RS. A review of trainingopportunities for ultrasonography in low and middle income countries.Tropical Med Int Health. 2012;17:808–19. https://doi.org/10.1111/j.1365-3156.2012.03014.x.

19. Van Dyk B, Motto JA, Buchmann EJ. Routine second-trimester ultrasound forlow risk pregnancies in a south African community. Int J Gynaecol Obstet.2007;98:257–8. https://doi.org/10.1016/j.ijgo.2007.03.034.

20. Mcclure E, Goldenberg R, Swanson D, Saleem S, Esamai F, et al. 3. Routineantenatal ultrasound in low/middle income countries: a cluster randomizedtrial. Am J Obstet Gynecol Oral Plenary Session I. 2017. Available: https://www.ajog.org/article/S0002-9378(16)30983-8/pdf.

21. McClure EM, Nathan RO, Saleem S, Esamai F, Garces A, et al. First look: acluster-randomized trial of ultrasound to improve pregnancy outcomes inlow income country settings. BMC Pregnancy Childbirth. 2014;14:73. https://doi.org/10.1186/1471-2393-14-73.

22. Geerts LT, Brand EJ, Theron GB. Routine obstetric ultrasound examinationsin South Africa: cost and effect on perinatal outcome--a prospectiverandomised controlled trial. Br J Obstet Gynaecol. 1996;103:501–7.

23. Ross AB, DeStigter KK, Rielly M, Souza S, Morey GE, et al. A low-costultrasound program leads to increased antenatal clinic visits and attendeddeliveries at a health care clinic in rural Uganda. PLoS One. 2013;8:e78450.https://doi.org/10.1371/journal.pone.0078450.

24. Müller-Rockstroh B. Appropriate and appropriated technology: lessonslearned from ultrasound in Tanzania. Med Anthropol. 2012;31:196–212.https://doi.org/10.1080/01459740.2011.639105.

25. Mbuyita S, Tillya R, Godfrey R, Kinyonge I, Shaban J, et al. Effects ofintroducing routinely ultrasound scanning during ante Natal Care (ANC)clinics on number of visits of ANC and facility delivery: a cohort study. ArchPublic Health. 2015;73:36. https://doi.org/10.1186/s13690-015-0086-8.

26. Ross AB, DeStigter KK, Coutinho A, Souza S, Mwatha A, et al. Ancillarybenefits of antenatal ultrasound: an association between the introduction ofa low-cost ultrasound program and an increase in the numbers of womenreceiving recommended antenatal treatments. BMC Pregnancy Childbirth.2014;14:424. https://doi.org/10.1186/s12884-014-0424-9.

27. Stein W, Katunda I, Butoto C. A two-level ultrasonographic service in amaternity care unit of a rural district hospital in Tanzania. Trop Dr. 2008;38:125–6. https://doi.org/10.1258/td.2007.070045.

28. Spencer JK, Adler RS. Utility of portable ultrasound in a community inGhana. J Ultrasound Med. 2008;27:1735–43.

29. Robert Tillya SM. Uptake of training on Vscan by midlevel providers workingin rural health facilities in Tanzania: implications for reliability. Journal of Biosafety& Health Education. 2014;02 https://doi.org/10.4172/2332-0893.1000123.

30. WHO Study Group on Training in Diagnostic Ultrasound: Essentials, Principlesand Standards (1996 : Philadelphia, Pennsylvania), Organization WH (1998)Training in diagnostic ultrasound : essentials, principles and standards : reportof a WHO study group. Available: http://apps.who.int/iris/handle/10665/42093.

31. Carrera JM. Obstetric ultrasounds in Africa: is it necessary to promote theirappropriate use? Donald School Journal of Ultrasound in Obstetrics &Gynecology. 2011;5:289–96. https://doi.org/10.5005/jp-journals-10009-1205.

32. Rijken MJ, Lee SJ, Boel ME, Papageorghiou AT, Visser GH, et al. Obstetricultrasound scanning by local health workers in a refugee camp on the Thai-Burmese border. Ultrasound Obstet Gynecol. 2009;34:395–403. https://doi.org/10.1002/uog.7350.

33. Bell G, Wachira B, Denning G. A pilot training program for point-of-careultrasound in Kenya. African Journal of Emergency Medicine. 2016;6:132–7.https://doi.org/10.1016/j.afjem.2016.03.002.

34. Greenwold N, Wallace S, Prost A, Jauniaux E. Implementing an obstetricultrasound training program in rural Africa. Int J Gynaecol Obstet. 2014;124:274–7. https://doi.org/10.1016/j.ijgo.2013.09.018.

35. Swanson JO, Kawooya MG, Swanson DL, Hippe DS, Dungu-Matovu P, et al.The diagnostic impact of limited, screening obstetric ultrasound whenperformed by midwives in rural Uganda. J Perinatol. 2014;34:508–12. https://doi.org/10.1038/jp.2014.54.

36. The Society and College of Radiographers. Standards for the provision of anultrasound service; 2015.

37. Mubuuke AG, Kiguli-Malwadde E, Businge F, Byanyima RK. Utilisation of obstetricsonography at a peri-urban health Centre in Uganda. Pan Afr Med J. 2010;7:24.

38. Bashour H, Hafez R, Abdulsalam A. Syrian women’s perceptions andexperiences of ultrasound screening in pregnancy: implications forantenatal policy. Reprod Health Matters. 2005;13:147–54.

39. Gammeltoft T, Nguyen HT. The commodification of obstetric ultrasoundscanning in Hanoi, Viet Nam. Reprod Health Matters. 2007;15:163–71.https://doi.org/10.1016/S0968-8080(06)29280-2.

40. Ranji A, Dykes AK. Ultrasound screening during pregnancy in Iran: womens’expectations, experiences and number of scans. Midwifery. 2012;28:24–9.https://doi.org/10.1016/j.midw.2010.10.005.

41. Chaturvedi S, Chhabra P, Bharadwaj S, Smanla S, Kannan AT, et al. Fetal sex-determination in Delhi: a population-based investigation. Trop Dr. 2007;37:98–100.

42. Kozuki N, Katz J, Khatry SK, Tielsch JM, LeClerq SC, et al. Community surveyon awareness and use of obstetric ultrasonography in rural Sarlahi District,Nepal. Int J Gynaecol Obstet. 2016;134:126–30. https://doi.org/10.1016/j.ijgo.2016.01.015.

43. Okeke TC, Enwereji JO, Okoro OS, Iferikigwe ES, Ikeako LC, et al. Desire forprenatal gender disclosure among primigravidae in Enugu, Nigeria. PatientPrefer Adherence. 2015;9:429–33. https://doi.org/10.2147/PPA.S77029.

44. Shukar-Ud-Din S, Ubaid F, Shahani E, Saleh F. Reasons for disclosure ofgender to pregnant women during prenatal ultrasonography. Int J WomensHealth. 2013;5:781–5. https://doi.org/10.2147/IJWH.S40685.

45. Mensah YB, Nkyekyer K, Mensah K. The Ghanaian woman’s experience andperception of ultrasound use in antenatal care. Ghana Med J. 2014;48:31.https://doi.org/10.4314/gmj.v48i1.5.

46. Chigbu CO, Odugu B, Okezie O. Implications of incorrect determination offetal sex by ultrasound. Int J Gynaecol Obstet. 2008;100:287–90. https://doi.org/10.1016/j.ijgo.2007.09.021.

47. Buscarini E, Lutz H, Mirk P. Manual of diagnostic ultrasound, volume 2.Geneva: World Health Organization; 2013.

48. Rijken MJ, Gilder ME, Thwin MM, Ladda Kajeechewa HM, Wiladphaingern J,et al. Refugee and migrant women’s views of antenatal ultrasound on theThai Burmese border: a mixed methods study. PLoS One. 2012;7:e34018.https://doi.org/10.1371/journal.pone.0034018.

49. Firth ER, Mlay P, Walker R, Sill PR. Pregnant women’s beliefs, expectationsand experiences of antenatal ultrasound in northern Tanzania. Afr J ReprodHealth. 2011;15:91–107.

50. Oluoch DA, Mwangome N, Kemp B, Seale AC, Koech A, et al. You cannot knowif it’s a baby or not a baby’: uptake, provision and perceptions of antenatal careand routine antenatal ultrasound scanning in rural Kenya. BMC PregnancyChildbirth. 2015;15:127. https://doi.org/10.1186/s12884-015-0565-5.

51. Ugwu AC, Udoh BE, Eze JC, Eze PC. Awareness of information, expectationsand experiences among women for obstetric sonography in a south EastNigeria population. East Afr J Public Health. 2011;8:132–4.

52. Kozuki N, Mullany LC, Khatry SK, Ghimire RK, Paudel S, et al. Accuracy ofhome-based Ultrasonographic diagnosis of obstetric risk factors by primary-level health Care Workers in Rural Nepal. Obstet Gynecol. 2016;128:604–12.https://doi.org/10.1097/AOG.0000000000001558.

53. Blaivas M, Kuhn W, Reynolds B, Brannam L. Change in differential diagnosisand patient management with the use of portable ultrasound in a remotesetting. Wilderness Environ Med. 2005;16:38–41.

Kim et al. Reproductive Health (2018) 15:129 Page 25 of 26

54. Roberts T, Henderson J, Mugford M, Bricker L, Neilson J, et al. Antenatalultrasound screening for fetal abnormalities: a systematic review of studiesof cost and cost effectiveness. BJOG. 2002;109:44–56.

55. Swanson D, Lokangaka A, Bauserman M, Swanson J, Nathan RO, et al.Challenges of implementing antenatal ultrasound screening in a rural studysite: a case study from the Democratic Republic of the Congo. Global health,science and practice. 2017;5:315–24. https://doi.org/10.9745/GHSP-D-16-00191.

56. Papp Z, Fekete T. The evolving role of ultrasound in obstetrics/gynecologypractice. Int J Gynecol Obstet. 2003;82:339–46.

57. Hofmeyr GJ. Routine ultrasound examination in early pregnancy: is itworthwhile in low-income countries? Ultrasound Obstet Gynecol. 2009;34:367–70. https://doi.org/10.1002/uog.7352.

58. W H O. WHO recommendations on antenatal Care for a Positive PregnancyExperience: ultrasound examination. Geneva: World Health Organization;2018.

59. Henwood PC, Mackenzie DC, Rempell JS, Murray AF, Leo MM, et al. Apractical guide to self-sustaining point-of-care ultrasound educationprograms in resource-limited settings. Ann Emerg Med. 2014;64:277–285.e2.https://doi.org/10.1016/j.annemergmed.2014.04.013.

60. Shah S, Bellows BA, Adedipe AA, Totten JE, Backlund BH, et al. Perceivedbarriers in the use of ultrasound in developing countries. Crit Ultrasound J.2015;7:28. https://doi.org/10.1186/s13089-015-0028-2.

61. Kawooya MG, Goldberg BB, De Groot W, Matovu PD, Malwadde EK, et al.Evaluation of US training for the past 6 years at ECUREI, the worldFederation for Ultrasound in medicine and biology (WFUMB) Centre ofExcellence, Kampala, Uganda. Acad Radiol. 2010;17:392–8. https://doi.org/10.1016/j.acra.2009.10.009.

62. Bagayoko CO, Traoré D, Thevoz L, Diabaté S, Pecoul D, et al. Medical andeconomic benefits of telehealth in low- and middle-income countries:results of a study in four district hospitals in Mali. BMC Health Serv Res.2014;14(Suppl 1):S9. https://doi.org/10.1186/1472-6963-14-S1-S9.

63. Mindel S. Role of imager in developing world. Lancet. 1997;350:426–9.https://doi.org/10.1016/S0140-6736(97)03340-0.

64. Maru DS, Schwarz R, Jason A, Basu S, Sharma A, et al. Turning a blind eye:the mobilization of radiology services in resource-poor regions. Glob Health.2010;6:18. https://doi.org/10.1186/1744-8603-6-18.

Kim et al. Reproductive Health (2018) 15:129 Page 26 of 26