This article was downloaded by: [ ] On: 03 May 2012, At: 15:09 Publisher: Taylor & Francis Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK Journal of Health Communication: International Perspectives Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/uhcm20 Effectiveness of mHealth Behavior Change Communication Interventions in Developing Countries: A Systematic Review of the Literature Tilly A. Gurman a , Sara E. Rubin a & Amira A. Roess a a Department of Global Health, George Washington University, Washington, District of Columbia, USA Available online: 01 May 2012 To cite this article: Tilly A. Gurman, Sara E. Rubin & Amira A. Roess (2012): Effectiveness of mHealth Behavior Change Communication Interventions in Developing Countries: A Systematic Review of the Literature, Journal of Health Communication: International Perspectives, 17:sup1, 82-104 To link to this article: http://dx.doi.org/10.1080/10810730.2011.649160 PLEASE SCROLL DOWN FOR ARTICLE Full terms and conditions of use: http://www.tandfonline.com/page/terms-and-conditions This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. The publisher does not give any warranty express or implied or make any representation that the contents will be complete or accurate or up to date. The accuracy of any instructions, formulae, and drug doses should be independently verified with primary sources. The publisher shall not be liable for any loss, actions, claims, proceedings, demand, or costs or damages whatsoever or howsoever caused arising directly or indirectly in connection with or arising out of the use of this material.
24
Embed
Effectiveness of mHealth Behavior Change Communication Interventions in Developing Countries: A Systematic Review of the Literature
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
This article was downloaded by: [ ]On: 03 May 2012, At: 15:09Publisher: Taylor & FrancisInforma Ltd Registered in England and Wales Registered Number: 1072954 Registeredoffice: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK
Journal of Health Communication:International PerspectivesPublication details, including instructions for authors andsubscription information:http://www.tandfonline.com/loi/uhcm20
Effectiveness of mHealth BehaviorChange Communication Interventionsin Developing Countries: A SystematicReview of the LiteratureTilly A. Gurman a , Sara E. Rubin a & Amira A. Roess aa Department of Global Health, George Washington University,Washington, District of Columbia, USA
Available online: 01 May 2012
To cite this article: Tilly A. Gurman, Sara E. Rubin & Amira A. Roess (2012): Effectiveness of mHealthBehavior Change Communication Interventions in Developing Countries: A Systematic Review of theLiterature, Journal of Health Communication: International Perspectives, 17:sup1, 82-104
To link to this article: http://dx.doi.org/10.1080/10810730.2011.649160
PLEASE SCROLL DOWN FOR ARTICLE
Full terms and conditions of use: http://www.tandfonline.com/page/terms-and-conditions
This article may be used for research, teaching, and private study purposes. Anysubstantial or systematic reproduction, redistribution, reselling, loan, sub-licensing,systematic supply, or distribution in any form to anyone is expressly forbidden.
The publisher does not give any warranty express or implied or make any representationthat the contents will be complete or accurate or up to date. The accuracy of anyinstructions, formulae, and drug doses should be independently verified with primarysources. The publisher shall not be liable for any loss, actions, claims, proceedings,demand, or costs or damages whatsoever or howsoever caused arising directly orindirectly in connection with or arising out of the use of this material.
Effectiveness of mHealth Behavior Change Communication Interventions in Developing
Countries: A Systematic Review of the Literature
TILLY A. GURMAN, SARA E. RUBIN, AND AMIRA A. ROESS
Department of Global Health, George Washington University, Washington, District of Columbia, USA
Mobile health (mHealth) technologies and telecommunication have rapidly been integrated into the health care delivery system, particularly in developing countries. Resources have been allocated to developing mHealth interventions, including those that use mobile technology for behavior change communication (BCC). Although the majority of mobile phone users worldwide live in the developing world, most research evaluating BCC mHealth interventions has taken place in developed countries. The purpose of this study was to conduct a systematic review of the literature to determine how much evidence currently exists for mHealth BCC interventions. In addition to analyzing available research for methodological rigor and strength of evidence, the authors assessed interventions for quality, applying a set of 9 standards recommended by mHealth experts. The authors reviewed 44 articles; 16 (36%) reported evaluation data from BCC mHealth interventions in a developing country. The majority of BCC mHealth interventions were implemented in Africa (n = 10) and Asia (n = 4). HIV/AIDS (n = 10) and family planning/pregnancy (n = 4) were the health topics most frequently addressed by interventions. Studies did not consistently demonstrate significant effects of exposure to BCC mHealth interventions on the intended audience. The majority of publications (n = 12) described interventions that used two-way communication in their message delivery design. Although most publications described interventions that conducted formative research about the intended audience (n = 10), less than half (n = 6) described targeting or tailoring the content. Although mHealth is viewed as a promising tool with the ability to foster behavior change, more evaluations of current interventions need to be conducted to establish stronger evidence.
Estimates indicate that half of all people living in remote areas of the world will have access to a mobile phone by 2012 (United Nations Foundation/Vodafone Foundation, 2009), with the greatest growth for mobile phones being primarily in low- and middle-income countries (Mechael, 2009). As a result, mobile phone technology is increasingly viewed as a promising communication channel that offers the potential to improve health care delivery and promote behavior change among vulnerable populations. Some of the more attractive features of mobile phones include the pay-as-you-go
Address correspondence to Amira A. Roess, Department of Global Health, George Washington University, Suite 200, 2175 K Street NW, Washington, DC 20037, USA. E-mail: [email protected]
Dow
nloa
ded
by [
] a
t 15:
09 0
3 M
ay 2
012
mHealth Behavior Change Communication 83
and the short message service (SMS) capacities. SMS is the most widely used form of communication globally and is a simple transfer of data usually from person-to-person in the form of 160 characters, but the message can also be sent in bulk and from computer to person or vice versa (Atun et al., 2006). Proponents suggest that advantages of using SMS to communicate with others include its ability to disseminate information immediately, assure a certain level of confidentiality, confirmation of delivery, and cost little (Atun et al., 2006).
This integration of mobile telecommunication technologies into the health arena is also known as mobile health (mHealth; Mechael, 2009). No systematic literature review to date has focused exclusively on the effectiveness of behavior change communication (BCC) mHealth interventions in developing countries. The current systematic review fills this gap and identifies recommendations for future BCC interventions and research using mHealth.
Experts in mHealth have outlined recommendations about how to use mHealth technologies strategically and effectively for BCC (GSMA Development Fund, 2010; K4Health, 2011; McNamara, 2007; Mechael, 2009; Mechael & Sloninsky, 2007). These recommendations address the technology specifications, the intended audience, the design of BCC messages, and the evaluation. The current systematic review analyzes available literature with these recommendations in mind in order to offer insight about the overall quality of the interventions themselves.
Method
Search Strategy
This systematic review consisted of several data collection steps. Publications were identified through four separate mechanisms. First, an Internet keyword search via Scopus, PubMed, MEDLINE, LexisNexis, and GoogleScholar identified potential peer-review publications. Keywords used in these searches included the following: text, text message, short message service, SMS, cell phone, phone, mobile phone, mobile health, mHealth, eHealth, health communication, health education, behavior, behavior change, prevention, and intervention. Second, we manually searched specific journals that had recently published articles on mHealth (The Lancet, the Journal of Health Communication, and Health Affairs). Third, gray literature was identified by searching the websites of organizations and agencies that are currently engaged in international mHealth BCC efforts (e.g., Population Services International, Academy for Educational Development, The Johns Hopkins University Center for Communication Programs, United States Agency for International Development, The World Bank, CORE Group, and the World Health Organization) and by querying the CORE Group mHealth listserv. Last, we reviewed reference lists within individual publications to ensure an exhaustive search.
Once a possible publication was identified, its title and abstract were reviewed to assess whether it might meet the inclusion criteria for this systematic review. Publications that were not excluded at the title/abstract stage underwent a full-text review. After the various levels of review, the result was the sample of publications that met all the eligibility criteria. (See Figure 1 for a decision tree describing the inclusion/exclusion process.) Eligibility criteria for inclusion were as follows: study used mHealth technology in its interventions for BCC in low- and middle-income countries; study included formative, process, or summative/outcome evaluation
Dow
nloa
ded
by [
] a
t 15:
09 0
3 M
ay 2
012
84 T. A. Gurman et al.
that assessed the mHealth intervention; and study was a peer-reviewed article, gray literature, internal organization report, or conference paper/PowerPoint presentation. Given the lack of research available in the field, limiting solely to peer-reviewed literature was not feasible.
The initial search for abstracts resulted in 44 articles that were reviewed for relevance to the research question. The main factors for ultimately excluding many articles included the following: study was conducted in a high-income country; study provided descriptive summaries of mHealth programs but failed to provide an evaluation of the program; study provided a short description of multiple mHealth programs without providing specific details on the BCC mHealth intervention; and study focused on mHealth informatics, health worker training, or other subsets of mHealth outside the realm of BCC.
Quality Assessment
In addition to evaluating publications for their methodological rigor and quality of evidence, the interventions were assessed for quality based on mHealth BCC intervention components recommended by experts (GSMA Development Fund, 2010; K4Health, 2011; McNamara, 2007; Mechael, 2009; Mechael & Sloninsky, 2007).
Technology-related components considered critical to the success of BCC mHealth interventions include selecting the appropriate technology for the intended audience, location, and context. Because BCC mHealth interventions may deal with sensitive health issues and phone sharing commonly practiced in developing countries, the ability to ensure privacy is also key (Atun et al., 2006; McNamara, 2007). In addition, a goal of BCC mHealth interventions in low- and middle-income countries should be to minimize costs while maximizing the benefits (Mechael et al., 2010).
Figure 1. Decision tree of included and excluded publications for mHealth behavior change communication literature review.
Dow
nloa
ded
by [
] a
t 15:
09 0
3 M
ay 2
012
mHealth Behavior Change Communication 85
Creating interventions with the intended audience’s level of comprehension—including not only language and literacy competency but also the ability to understand and use technology—will help to foster greater behavior change. Message design considerations include creating targeted and tailored content which engages the user in two-way communication, allowing users to interact and ask questions instead of simply receiving information (Atun et al., 2006; Mechael et al., 2010). In addition, despite the anytime/anywhere nature of mobile phones, the timing of communication messages (i.e., time of day, frequency, and sequencing) should be considered in order to communicate at a convenient time in a way in which the end user is receptive to the information (Atun et al., 2006).
Last, quality components related to the evaluation include conducting formative research in order to design culturally sensitive interventions which truly understand the audience, in terms of health needs and telephone usage. In addition, to obtain the strongest evidence and entice donors to continue funding, experts have suggested the need for long-term evaluation of mHealth BCC interventions (Mechael et al., 2010).
Results
Although a total of 44 articles received full-text review, only 16 (36% of the 44 articles) met the inclusion criteria for the systematic review. (See Table 1 for a summary of the 16 articles.) The majority of publications reported findings from summative evaluations (n = 10), followed by formative (n = 5) and then process (n = 1) evaluation. For three of the publications, the same organization, Text To Change, was involved in the intervention (“Using an interactive,” 2010; Danis et al., 2010; Hoefman & Apunyo, 2010). Quantitative research methodologies were most often represented, with nine publications that were solely quantitative and four that were mixed-methods.
The majority of studies were located in Africa (n = 10), Asia (n = 4), and multicountries (n = 2) consisting of Sub-Saharan Africa, South America, and South Asia. Studies most commonly occurred in Uganda (n = 3), South Africa (n = 3), and India (n = 3). The topical focus of the articles included HIV/AIDS (n = 10), family planning/pregnancy (n = 4), self-breast exam (n = 1), general health appointments at a clinic (n = 1), and tuberculosis (n = 1). No existing literature offers an explanation as to why more than two thirds of the existing literature focuses on HIV/AIDS.
Five publications focused on interventions for reminder to do a certain behavior. However, as discussed later only some required a response indicating that the behavior had been completed, and most of them that do require responses are entirely based on self-reported data. This ranged from several different kinds of behaviors such as taking a drug treatment, conducting a self-breast examination, and keeping track of the menstrual cycle for a traditional family planning method. Four publications specifically focused on quiz-based SMS in which a server would send out knowledge questions to participants, they would be encouraged to reply and receive the correct answer if wrong or congratulated if correct.
Quality Assessment
For the application of mHealth recommendations, the following areas were used the most in the reviewed studies: two-way communication (n = 12), technology platform selection (n = 10), and understanding the audience (n = 10). Lesser used recommendations include the following: comparison studies (n = 2), long-term evaluation (n = 2), and timing of communication (n = 3).
Dow
nloa
ded
by [
] a
t 15:
09 0
3 M
ay 2
012
Tab
le 1
. Su
mm
ary
of li
tera
ture
rev
iew
Reg
ion
and
coun
try
Aut
hor
(o
r or
gani
zati
on)
Sum
mar
y R
elev
ant
find
ings
App
licat
ion
of m
Hea
lth
reco
mm
enda
tion
sE
valu
atio
n
Sub
-Sah
aran
Afr
ica
Dem
ocra
tic
Rep
ublic
of
Con
go
Lig
ne V
erte
(C
orke
r,
2010
)
•T
oll-
free
fa
mily
-pl
anni
ng
hotl
ine
•L
esso
ns
lear
ned
from
3
year
s of
the
ho
tlin
e
•M
ore
than
80,
000
calls
mad
eto
ho
tlin
e ov
er 3
yea
rs; m
en a
re t
he
prim
ary
user
s of
the
hot
line—
80%
an
nual
ly
•C
allt
ime
of2
min
isin
suff
icie
ntto
an
swer
all
ques
tion
s (p
art o
f the
dea
l w
ith
VO
DA
CO
M li
mit
ed to
2 m
in)
•C
once
pto
fah
otlin
eno
twel
lun
ders
tood
in th
e D
emoc
rati
c R
epub
lic o
f Con
go—
20%
of c
alls
on
a to
pic
unre
late
d to
fam
ily p
lann
ing
•C
ost
ofp
rogr
am:a
nnua
lope
rati
ng
cost
s ab
out
$8,0
00 o
n th
e ba
sis
of
per-
call
rate
of
$0.3
6 an
d sa
lary
for
th
e ed
ucat
ors
•Se
lect
app
ropr
iate
te
chno
logy
•E
nsur
epr
ivac
y•
Min
imiz
eco
sts
•C
onsi
der
com
preh
ensi
on•
Use
tw
o-w
ay
com
mun
icat
ion
•T
arge
tan
dta
ilor
cont
ent
•U
nder
stan
dau
dien
ce•
Con
duct
long
-te
rm e
valu
atio
n
•T
ype
of
eval
uati
on:
outc
ome
•M
etho
dolo
gy:
mix
ed m
etho
ds•
Des
ign:
le
sson
s le
arne
d
docu
men
t,
uncl
ear
on t
he
desi
gn
Gha
naM
obile
Mid
wif
e (G
ram
een
Fou
ndat
ion,
20
11)
•“M
obile
m
idw
ife”
al
low
s pr
egna
nt
wom
en a
nd
thei
r fa
mili
es
to r
ecei
ve
wee
kly
•42
%w
holi
sten
toth
epr
imar
ym
essa
ge a
lso
opt t
o lis
ten
to a
sec
ond
mes
sage
and
36%
opt
to li
sten
to a
th
ird
mes
sage
•
Con
tent
tra
nsla
ted
into
sev
eral
G
hana
ian
lang
uage
s an
d ta
rget
ed
tow
ard
mot
hers
, fat
hers
, and
ex
tend
ed f
amily
•Se
lect
app
ropr
iate
te
chno
logy
•M
inim
ize
cost
s•
Con
side
rco
mpr
ehen
sion
•T
arge
tan
dta
ilor
cont
ent
•T
ype
of
eval
uati
on:
form
ativ
e•
Met
hodo
logy
:m
ixed
met
hods
•D
esig
n:
less
ons
lear
ned
docu
men
t
86
Dow
nloa
ded
by [
] a
t 15:
09 0
3 M
ay 2
012
(Con
tinu
ed)
m
essa
ges
(99%
cho
ose
voic
e ov
er
text
mes
sage
)
•P
regn
ancy
que
stio
nbo
xse
tup
befo
re v
oice
pro
gram
to h
ear
wha
t qu
esti
ons
wer
e on
the
min
ds o
f m
othe
rs a
nd fa
mily
mem
bers
•C
ost
ofr
ecor
ding
and
tra
nsla
ting
17
7 m
essa
ges
was
$22
,000
per
la
ngua
ge•
Peo
ple
pref
erre
dto
hea
ra
soft
er
“aun
tie”
voi
ce to
del
iver
the
mes
sage
•U
nder
stan
dau
dien
ce
Gha
naT
ext
Me!
F
lash
Me!
(C
lem
mon
s,
2009
)
•H
elpl
ine
for
MSM
to
ans
wer
qu
esti
ons
and
enco
urag
e th
em t
o se
ek
care
•M
SMc
all
and
hang
up
or t
ext
thei
r nu
mbe
r to
“f
lash
” it
, an
d th
en
coun
selo
r ca
lls t
hem
ba
ck
•In
the
fir
stm
onth
,fiv
ehe
lplin
eco
unse
lors
spo
ke w
ith
439
MSM
, fo
r an
ave
rage
of
20 m
in; 1
,000
te
xts
wer
e m
isse
d in
the
fir
st m
onth
be
caus
e th
ere
wer
e no
t en
ough
co
unse
lors
to
mee
t de
man
d•
Ave
rage
of
88M
SMc
ouns
eled
per
m
onth
ver
sus
50 M
SM c
ompa
red
to in
-per
son
coun
selin
g.•
Bef
ore
the
surv
ey,9
8%o
fM
SMn
oup
taki
ng H
IV/A
IDS
serv
ices
, but
af
ter
the
prog
ram
the
re w
as a
54%
up
take
of
serv
ices
•
Sixf
old
incr
ease
of
upta
kea
tC
entr
efo
r P
opul
ar E
duca
tion
and
Hum
an
Rig
hts,
Gha
na D
rop-
In C
ente
r on
e m
onth
aft
er la
unch
of
hotl
ine
•Se
lect
app
ropr
iate
te
chno
logy
•E
nsur
epr
ivac
y•
Use
tw
o-w
ay
com
mun
icat
ion
•T
arge
tan
dta
ilor
cont
ent
•K
now
aud
ienc
e
•T
ype
of
eval
uati
on:
outc
ome
•M
etho
dolo
gy:
quan
tita
tive
•D
esig
n:q
uasi
-ex
peri
men
tal
87
Dow
nloa
ded
by [
] a
t 15:
09 0
3 M
ay 2
012
Reg
ion
and
coun
try
Aut
hor
(o
r or
gani
zati
on)
Sum
mar
y R
elev
ant
find
ings
App
licat
ion
of m
Hea
lth
reco
mm
enda
tion
sE
valu
atio
n
Ken
yaW
elT
el K
enya
1
(Les
ter,
201
0)•
Rem
inde
rpr
ogra
m
com
pari
ng
anti
retr
ovir
al
adhe
renc
e af
ter
rece
ivin
g te
xt m
essa
ges
to t
hose
who
re
ceiv
ed n
o re
min
ders
•Se
lf-r
epor
ted
adhe
renc
eof
62%
(vir
al
supp
ress
ion
57%
) in
inte
rven
tion
gr
oup;
50%
adh
eren
ce (
vira
l su
ppre
ssio
n 48
%)
in c
ontr
ol g
roup
•Se
lf-r
epor
ted
adhe
renc
esi
gnif
ican
tly
bett
er in
con
trol
gro
up; o
dds
rati
o =
.57,
p =
.002
8 •
No
seco
ndar
you
tcom
ess
how
ed
sign
ific
ance
suc
h as
mal
e se
x, u
rban
re
side
nce,
mob
ile p
hone
ow
ners
hip
•C
ost:
Int
erve
ntio
nco
sts
abou
t$0
.05
per
text
mes
sage
and
a t
otal
of
$20
per
100
pati
ents
per
mon
th, a
nd
follo
w-u
p vo
ice
calls
ave
rage
d $3
.75
per
nurs
e pe
r m
onth
•98
%o
fin
terv
enti
ong
roup
wou
ld
reco
mm
end
it t
o a
frie
nd; 1
91 o
ut o
f 19
4 w
ante
d to
con
tinu
e th
e pr
ogra
m
•M
inim
ize
cost
s•
Use
tw
o-w
ay
com
mun
icat
ion
•C
ondu
ctlo
ng-t
erm
ev
alua
tion
•T
ype
of
eval
uati
on:
outc
ome
•M
etho
dolo
gy:
quan
tita
tive
•D
esig
n:
rand
omiz
ed
clin
ical
tri
al;
n =
538
; SM
S in
terv
enti
on
grou
p (n
= 2
73),
co
ntro
l gr
oup
wit
h no
rem
inde
r (n
= 2
65);
the
in
terv
enti
on
grou
p re
ceiv
ed
wee
kly
text
m
essa
ge
rem
inde
rs
to t
ake
anti
retr
ovir
als
from
nur
ses
and
wer
e as
ked
to
resp
ond
88
Tab
le 1
. C
onti
nued
Dow
nloa
ded
by [
] a
t 15:
09 0
3 M
ay 2
012
w
ithi
n 48
hr;
the
co
ntro
l gro
up
cons
iste
d of
265
w
ho r
ecei
ved
no r
emin
ders
; se
lf-r
epor
ted
anti
retr
ovir
al
adhe
renc
e w
as
mon
itor
ed f
or 3
0 da
ys a
nd t
hen
at
6- a
nd 1
2 m
onth
fo
llow
-up
Nig
eria
Lea
rnin
g ab
out L
ivin
g (M
obile
4Goo
d,
no d
ate)
•C
lient
sas
kqu
esti
ons
abou
t se
xual
and
re
prod
ucti
ve
heal
th v
ia
call,
tex
t m
essa
ge, o
r w
eb t
hrou
gh
the
“My
Que
stio
n”
serv
ice;
qu
esti
ons
ar
e st
ored
on
a
mob
ile
•B
yth
een
dof
pilo
tph
ase,
alm
ost
9,00
0 yo
ung
peop
le h
ad b
een
reac
hed
•14
mon
ths
afte
rla
unch
of
the
init
iati
ve, “
My
Que
stio
n” h
ad
rece
ived
mor
e th
an 6
0,00
0 qu
esti
ons
via
text
mes
sage
; “M
y A
nsw
er”
part
icip
atio
n in
crea
sed
mon
th t
o m
onth
thr
ough
out
the
pilo
t•
76%
of
user
sfe
ltq
uest
ions
wer
ean
swer
ed p
rope
rly;
24%
sai
d us
e se
rvic
e be
caus
e it
’s f
ree.
Of
24%
who
w
ere
not
sati
sfie
d, m
ore
than
50%
di
d no
t re
ceiv
e re
ply
to t
ext
mes
sage
s be
caus
e of
“ne
twor
k fl
uctu
atio
ns”
•Se
lect
app
ropr
iate
te
chno
logy
•E
nsur
epr
ivac
y•
Use
tw
o-w
ay
com
mun
icat
ion
•T
arge
tan
dta
ilor
cont
ent
•U
nder
stan
dth
eau
dien
ce
•T
ype
of
eval
uati
on:
outc
ome
•M
etho
dolo
gy:
mix
ed m
etho
ds•
Des
ign:
qua
si-
expe
rim
enta
l (n
= 9
,000
),
uncl
ear
whe
ther
th
at n
umbe
r si
gned
up
volu
ntar
ily
and
uncl
ear
how
man
y pa
rtic
ipat
ed in
su
rvey
89
(Con
tinu
ed)
Dow
nloa
ded
by [
] a
t 15:
09 0
3 M
ay 2
012
Reg
ion
and
coun
try
Aut
hor
(o
r or
gani
zati
on)
Sum
mar
y R
elev
ant
find
ings
App
licat
ion
of m
Hea
lth
reco
mm
enda
tion
sE
valu
atio
n
pl
atfo
rm
and
then
co
unse
lors
re
spon
d to
th
e qu
esti
ons
wit
hin
24 h
r•
The
“M
yA
nsw
er”
serv
ice
offe
rs
teen
s th
e ch
ance
to
answ
er q
uiz
ques
tion
s an
d w
in a
pri
zeSo
uth
Afr
ica
Cel
l-L
ife
(Ski
nner
, n.d
.)•
Pro
cess
re
sear
ch t
o de
term
ine
the
best
way
to
bet
ter
dist
ribu
te
info
rmat
ion
thro
ugh
mob
ile p
hone
s
•Im
port
ance
of
two-
way
tex
tm
essa
ges;
tai
lor
text
mes
sage
s so
pe
ople
can
sel
ect
to r
ecei
ve a
sm
all
num
ber
or b
ulk;
pri
vacy
on
HIV
st
atus
was
a c
once
rn•
Whe
nas
ked
wha
the
alth
in
form
atio
n th
ey w
ante
d to
rec
eive
on
cel
l pho
ne: 6
1.4%
, HIV
-rel
ated
in
form
atio
n; 5
8.6%
, TB
tre
atm
ent;
51
.9%
, avo
id
•Se
lect
app
ropr
iate
te
chno
logy
•E
nsur
epr
ivac
y•
Con
side
rco
mpr
ehen
sion
•T
arge
tan
dta
ilor
cont
ent
•T
ype
of
eval
uati
on:
proc
ess
•M
etho
dolo
gy:
qual
itat
ive
•D
esig
n:1
0in
-dep
th
inte
rvie
ws
and
5 fo
cus
grou
ps
90
Tab
le 1
. C
onti
nued
Dow
nloa
ded
by [
] a
t 15:
09 0
3 M
ay 2
012
to
mem
bers
of
a h
ealt
h or
gani
zati
on
H
IV in
fect
ion
info
rmat
ion;
51.
4%,
the
proc
ess
of H
IV t
esti
ng•
Pre
ferr
edla
ngua
ge:X
hosa
,61.
9%;
Eng
lish,
16.
7%
•T
ime
the
com
mun
icat
ion
appr
opri
atel
y•
Und
erst
and
the
audi
ence
Sout
h A
fric
aP
roje
ct
Mas
ilule
ke•
Bul
kte
xt
mes
sagi
ng
serv
ice
sent
1
mill
ion
text
s pe
r da
y fo
r 36
5 da
ys t
o So
uth
Afr
ican
s•
Mes
sage
is
call
to a
ctio
n to
vis
it H
IV/
TB
cal
l cen
ters
to
obt
ain
info
rmat
ion
or
get
test
ed
•1.
2m
illio
nca
llst
oho
tlin
eat
trib
uted
to
SM
S pr
ogra
m (
300%
incr
ease
in
over
all c
alls
to
hotl
ine)
•E
nglis
han
dZ
ulu
offe
red,
but
Zul
um
essa
ges
outp
erfo
rmed
•C
onsi
der
com
preh
ensi
on•
Use
tw
o-w
ay
com
mun
icat
ion
•T
arge
tan
dta
ilor
cont
ent
•T
ype
of
eval
uati
on:
outc
ome
•M
etho
dolo
gy:
quan
tita
tive
•D
esig
n:n
one
liste
d
Uga
nda
Hea
lth
Chi
ld
and
Tex
t to
C
hang
e
•B
ulk
text
m
essa
ges
deliv
ered
to
3,00
0 pe
ople
in
are
as o
f U
gand
a
•58
.4%
ans
wer
edc
orre
ctly
(m
ulti
ple
choi
ce)
how
bab
ies
can
acqu
ire
HIV
.•
Con
clud
ead
ding
voi
ces
egm
ent
wou
ld h
elp
reac
h th
e ill
iter
ate
(48%
of
tar
get)
•C
onsi
der
com
preh
ensi
on•
Use
tw
o-w
ay
com
mun
icat
ion
•T
ype
of
eval
uati
on:
outc
ome
•M
etho
dolo
gy:
quan
tita
tive
91
(Con
tinu
ed)
Dow
nloa
ded
by [
] a
t 15:
09 0
3 M
ay 2
012
Reg
ion
and
coun
try
Aut
hor
(o
r or
gani
zati
on)
Sum
mar
y R
elev
ant
find
ings
App
licat
ion
of m
Hea
lth
reco
mm
enda
tion
sE
valu
atio
n
w
here
Hea
lth
Chi
ld h
as
prog
ram
min
g•
Goa
lof
quiz
to
bui
ld
know
ledg
e on
an
tena
tal c
are/
preg
nanc
y,
HIV
/AID
S,
and
mal
aria
pr
even
tion
•A
vera
ge
resp
onse
rat
e to
the
sur
vey
was
33%
(70
0 pa
rtic
ipan
ts)
•26
4pe
ople
sho
wed
up
ath
ealt
hcl
inic
s m
enti
onin
g th
is S
MS
prog
ram
•O
fth
e70
0pa
rtic
ipan
ts,t
hey
attr
ibut
e kn
owin
g ab
out
heal
th
serv
ices
thr
ough
the
fol
low
ing:
38%
co
mm
unit
y he
alth
wor
kers
, 28%
ra
dio,
25%
mob
ile p
hone
s
•D
esig
n:q
uasi
-ex
peri
men
tal
Uga
nda
Tex
t to
Cha
nge
(Dan
is, 2
010)
•
Qui
z-ba
sed
bulk
SM
S pr
ogra
m
to in
crea
se
know
ledg
e of
H
IV/A
IDS
•SM
Squ
izq
uest
ions
ans
wer
ed
corr
ectl
y ra
nged
fro
m t
hree
gro
ups
rang
ed b
etw
een
from
84.
6% t
o 91
.9%
•T
echn
ical
err
orin
res
pond
ing
to
text
mes
sage
s ra
nged
fro
m 3
.16
to 9
.70%
bec
ause
tex
t m
essa
ge in
co
nver
sati
onal
for
mat
tha
t co
mpu
ter
cann
ot p
roce
ss
•Se
lect
app
ropr
iate
te
chno
logy
•C
onsi
der
com
preh
ensi
on
•U
set
wo-
way
co
mm
unic
atio
n
•T
ype
of
eval
uati
on:
outc
ome
•M
etho
dolo
gy:
quan
tita
tive
•D
esig
n:q
uasi
-ex
peri
men
tal;
thre
e sa
mpl
es
92
Tab
le 1
. C
onti
nued
Dow
nloa
ded
by [
] a
t 15:
09 0
3 M
ay 2
012
•In
terv
enti
on
cond
ucte
d in
th
ree
stud
ies:
O
ne g
roup
ta
rget
ed a
re
gion
al a
reas
of
10,
000
cell
phon
e us
ers
in
Uga
nda,
the
ot
her
targ
eted
tw
o se
para
te
grou
ps o
f 5,
000
fact
ory
wor
kers
in
Sout
heas
tern
U
gand
a.
•H
ealt
hce
nter
sat
fac
tory
qui
zto
wns
re
ceiv
ed t
hree
fold
incr
ease
in H
IV
test
req
uest
s fr
om w
orke
rs (
no
stat
isti
cs o
n th
is p
rovi
ded)
for
thre
e in
terv
enti
ons;
n
= 1
0,00
0;
n =
5,0
00,
n =
5,0
00;
part
icip
atio
n ra
tes
vari
ed f
rom
5%
to
10%
in
the
Dis
tric
t Q
uiz
to a
bout
50%
in
Fac
tory
1 Q
uiz;
a
tota
l of
1,84
6 pa
rtic
ipan
ts
opte
d in
to t
he
prog
ram
by
text
ing
that
th
ey w
ante
d to
pa
rtic
ipat
eU
gand
aT
ext
to C
hang
e (H
oefm
an, n
o da
te)
•B
ulk
text
m
essa
ges
deliv
ered
to
8,00
0 us
ers
and
seve
ral
opte
d in
;
•T
hew
eek
afte
rth
epr
ogra
m,3
98
HIV
tes
ts w
ere
carr
ied
out,
dou
ble
the
185
in t
he p
revi
ous
wee
k•
96%
sta
ted
that
the
sur
vey
help
ed
them
gai
n H
IV k
now
ledg
e•
On
aver
age,
74%
cor
rect
ans
wer
s•
19%
of
part
icip
ants
wer
efe
mal
e
•C
onsi
der
com
preh
ensi
on•
Use
tw
o-w
ay
com
mun
icat
ion
•T
ype
of
eval
uati
on:
outc
ome
•M
etho
dolo
gy:
quan
tita
tive
93
(Con
tinu
ed)
Dow
nloa
ded
by [
] a
t 15:
09 0
3 M
ay 2
012
Reg
ion
and
coun
try
Aut
hor
(o
r or
gani
zati
on)
Sum
mar
y R
elev
ant
find
ings
App
licat
ion
of m
Hea
lth
reco
mm
enda
tion
sE
valu
atio
n
ov
eral
l, 8,
272
subs
crib
ed f
or
the
surv
ey, b
ut
1,22
2 di
d no
t re
spon
d to
any
qu
esti
ons
•G
oalo
fqu
iz
to b
uild
kn
owle
dge
on H
IV
and
fam
ily
plan
ning
in
add
itio
n en
cour
agin
g fr
ee H
IV/
AID
S te
stin
g th
e fo
llow
ing
wee
k
•U
nder
stan
dth
eau
dien
ce•
Des
ign:
qua
si-
expe
rim
enta
l
Asi
a
Chi
na[N
o pr
ogra
m
nam
e] (
Che
n,
2007
)
•R
emin
der
prog
ram
for
ge
nera
l hea
lth
upco
min
g
•T
her
ates
of
atte
ndan
cew
ere
80.5
%
in c
ontr
ol g
roup
, 87.
5% in
the
SM
S gr
oup
and
88.3
% in
the
voi
ce g
roup
. T
he in
terv
enti
on g
roup
s si
gnif
ican
tly
high
er t
han
•Se
lect
app
ropr
iate
te
chno
logy
•M
inim
ize
cost
s•
Und
erst
and
the
audi
ence
•T
ype
of
eval
uati
on:
outc
ome
•M
etho
dolo
gy:
quan
tita
tive
94
Tab
le 1
. C
onti
nued
Dow
nloa
ded
by [
] a
t 15:
09 0
3 M
ay 2
012
ap
poin
tmen
ts;
one
grou
p re
ceiv
ed t
ext
mes
sage
re
min
ders
, on
e gr
oup
rece
ived
voi
ce
rem
inde
rs,
anot
her
rece
ived
no
rem
inde
rs
co
ntro
l (p
= .0
01).
•
No
stat
isti
cald
iffe
renc
ebe
twee
nth
eSM
S an
d vo
ice
grou
p (p
= .6
70)
•C
ost
effe
ctiv
enes
s:S
MS
grou
pco
st
.31
Yua
n an
d vo
ice
grou
p w
as .4
8 Y
uan
per
part
icip
ant;
tex
t m
essa
ge
rem
inde
rs w
ere
mor
e co
st-e
ffec
tive
•D
esig
n:
rand
omiz
ed
cont
rol t
rial
; 1,
859
had
appo
intm
ent
at a
he
alth
clin
ic a
nd
wer
e ra
ndom
ly
sele
cted
to
part
icip
ate,
of
whi
ch 6
19 w
ere
plac
ed in
con
trol
gr
oup,
620
in
the
SMS
grou
p,
and
620
in t
he
voic
e gr
oup;
th
e in
terv
enti
on
grou
ps r
ecei
ved
rem
inde
rs a
bout
th
eir
upco
min
g m
eeti
ng a
t th
e cl
inic
Indi
aC
ycle
Tel
(Lav
oie,
20
09)
•F
ocus
gro
ups
to d
iscu
ss
pref
eren
ce f
or
a te
xt m
essa
ge
•M
ena
ndw
omen
inte
rest
edin
ser
vice
bu
t la
ck k
now
ledg
e of
“fe
rtile
day
”•
Mes
sage
sho
uld
say
“saf
e/un
safe
da
y,”
part
icip
ants
vie
wed
•C
onsi
der
com
preh
ensi
on•
Und
erst
and
the
audi
ence
•T
ype
of
eval
uati
on:
form
ativ
e
95
(Con
tinu
ed)
Dow
nloa
ded
by [
] a
t 15:
09 0
3 M
ay 2
012
Reg
ion
and
coun
try
Aut
hor
(o
r or
gani
zati
on)
Sum
mar
y R
elev
ant
find
ings
App
licat
ion
of m
Hea
lth
reco
mm
enda
tion
sE
valu
atio
n
pr
ogra
m
to h
elp
part
icip
ants
ke
ep t
rack
of
men
stru
al
cycl
e in
line
w
ith
the
stan
dard
day
m
etho
d
“fer
tile
day
” as
deg
radi
ng t
o
wom
en•
Mes
sage
sho
uld
beH
indi
wor
ds
spel
led
out
in E
nglis
h le
tter
s:
Hin
glis
h•
Kee
pfr
eque
ncy
ofm
essa
ges
low
and
co
nten
t sh
ort
•M
etho
dolo
gy:
qual
itat
ive
•D
esig
n:F
our
grou
ps c
onsi
sted
of
wom
en, t
wo
grou
ps c
onsi
sted
of
men
, and
one
gr
oup
of c
oupl
es,
for
a to
tal o
f 54
pa
rtic
ipan
ts w
ho
are
aged
18
to
28 y
ears
wit
h a
fam
ily-p
lann
ing
need
Indi
a[N
o pr
ogra
m
nam
e]
(Kho
khar
, 20
09)
•O
ffic
ew
omen
re
ceiv
e m
onth
ly
rem
inde
rs v
ia
text
mes
sage
to
con
duct
m
onth
ly s
elf-
brea
st e
xam
; th
ey m
ust
repl
y if
the
y co
mpl
eted
it,
and
if n
ot, w
hy
•T
hem
ain
barr
iers
for
tho
sew
hod
id
not
do t
he b
reas
t se
lf-e
xam
incl
uded
th
e fo
llow
ing:
for
got
to d
o it
/will
do
it n
ow, 5
4%; b
usy,
47%
; anx
iety
, 12
%; p
ain
in b
reas
ts, 4
%; s
ome
ques
tion
reg
ardi
ng e
xam
, 4%
•In
the
fir
stm
onth
,onl
y42
.4%
had
co
nduc
ted
the
exam
bef
ore
rece
ivin
g th
e m
essa
ge a
nd b
y th
e si
xth
mon
th
72.6
% h
ad c
ondu
cted
the
exa
m
•Se
lect
app
ropr
iate
te
chno
logy
•U
set
wo-
way
co
mm
unic
atio
n•
Tim
eth
eco
mm
unic
atio
n ap
prop
riat
ely
•U
nder
stan
dth
eau
dien
ce
•T
ype
of
eval
uati
on:
outc
ome
•M
etho
dolo
gy:
quan
tita
tive
•D
esig
n:q
uasi
-ex
peri
men
tal;
106
fem
ale
empl
oyee
s to
vo
ice
call
96
Tab
le 1
. C
onti
nued
Dow
nloa
ded
by [
] a
t 15:
09 0
3 M
ay 2
012
betw
een
the
ages
of
22 t
o 54
yea
rs in
a
priv
ate
sect
or
offi
ce in
Del
hi
volu
ntee
red
to
part
icip
ate
in
the
stud
y; a
ll w
omen
rec
eive
d te
xt m
essa
ge
rem
inde
rs t
o co
nduc
t br
east
se
lf-e
xam
and
se
lf-r
epor
ted
if
they
con
duct
ed
exam
Indi
aP
roje
ct P
raga
ti
(Sam
basi
van,
20
11)
•P
rere
cord
ed
voic
e m
essa
ging
sy
stem
to
reac
h ur
ban
sex
wor
kers
in
Ban
galo
re a
nd
offe
r
•In
augu
rati
onin
vita
tion
:Of
35
calle
d, 9
3.10
% li
sten
ed t
o en
tire
m
essa
ge, a
nd 1
0 of
the
29
peop
le
who
ans
wer
ed t
he c
alls
att
ende
d th
e ev
ent
•M
icro
fina
nce
call:
Of
23c
alle
d,2
2w
ere
reac
habl
e an
d 20
list
ened
to
the
full
mes
sage
; no
diff
eren
ce w
as
•Se
lect
app
ropr
iate
te
chno
logy
•E
nsur
epr
ivac
y•
Con
side
rco
mpr
ehen
sion
•T
ime
the
com
mun
icat
ion
appr
opri
atel
y
•T
ype
of
eval
uati
on:
outc
ome
•M
etho
dolo
gy:
mix
ed m
etho
ds•
Des
ign:
Fou
rin
terv
enti
on
even
ts o
ccur
red
urba
n se
x
97
(Con
tinu
ed)
Dow
nloa
ded
by [
] a
t 15:
09 0
3 M
ay 2
012
Reg
ion
and
coun
try
Aut
hor
(o
r or
gani
zati
on)
Sum
mar
y R
elev
ant
find
ings
App
licat
ion
of m
Hea
lth
reco
mm
enda
tion
sE
valu
atio
n
in
vita
tion
s to
pr
ogra
ms
or
rem
inde
rs t
o vi
sit
heal
th
clin
ic/p
ay b
ills
fo
und
in t
hose
who
got
the
cal
l an
d pa
id o
n ti
me
and
thos
e w
ho
did
not
•M
edic
alt
esti
ngr
emin
ders
:90%
of
thos
e co
ntac
ted
conn
ecte
d an
d 59
%
liste
ned
to t
he w
hole
mes
sage
•C
ompu
ter
trai
ning
:Of
tota
lnum
ber
dial
ed, 6
3 pe
ople
(19
%)
atte
nded
the
tr
aini
ng
•U
nder
stan
dth
eau
dien
ce
wor
kers
: 35
calle
d fo
r th
e Sw
ati M
anne
In
augu
rati
on
even
t, 3
8 ca
lled
for
mic
rofi
nanc
e re
min
ders
, 230
ca
lls fo
r m
edic
al
test
ing
rem
inde
rs,
627
invi
ted
to
a co
mpu
ter
trai
ning
cla
ss;
com
plet
ed
21 s
truc
ture
d in
terv
iew
s w
ith
urba
n se
x w
orke
rs
Mul
tico
untr
y
Sout
h
Afr
ica,
N
icar
agua
, P
akis
tan
Sim
Pill
, Sim
Med
, X
outT
B
(Bar
clay
, 200
9)
•A
nov
ervi
ew
arti
cle
that
pr
ovid
es
find
ings
on
thre
e m
ajor
T
B r
egim
en
rem
inde
r
•Si
mP
ill:A
fter
pat
ient
sre
ceiv
ed
text
s fo
r 10
mon
ths,
dru
g ad
here
nce
stab
ilize
d be
twee
n 86
% a
nd 9
2%
and
trea
tmen
t su
cces
s ra
te o
f 94
%;
Sim
Pill
nur
se c
ould
kee
p ta
bs o
n 50
–60
pati
ents
inst
ead
of ju
st 1
0
•M
inim
ize
cost
s•
Use
tw
o-w
ay
com
mun
icat
ion
•T
ype
of
eval
uati
on:
outc
ome
98
Tab
le 1
. C
onti
nued
Dow
nloa
ded
by [
] a
t 15:
09 0
3 M
ay 2
012
pr
ogra
ms;
Si
mP
ill h
ad a
pi
lot s
tudy
with
15
5 T
B p
atie
nts
at th
ree
clin
ics
in C
ape
Tow
n;
the
othe
r tw
o pr
ogra
ms
offe
r re
sults
, but
not
in
form
atio
n on
pop
ulat
ion
stud
ied
•Si
mM
ed:C
heap
erc
ompe
tito
rto
Si
mP
ill t
hat
asks
res
pond
ents
to
pres
s sp
eed
dial
but
ton
afte
r ta
king
m
edic
atio
n, w
hich
rec
ords
the
ir
adhe
renc
e in
a d
atab
ase
•X
outT
B:P
atie
nts
urin
ate
onfi
lter
pa
per
afte
r ta
king
med
icat
ion
whi
ch
reve
als
a co
de, a
nd p
atie
nt m
ust s
end
code
via
text
mes
sage
; eco
nom
ic
ince
ntiv
e fo
r th
ose
who
par
tici
pate
; pr
ogra
m h
as b
een
test
ed in
Nic
arag
ua
and
now
Pak
ista
n (n
o re
sult
s lis
ted)
•M
etho
dolo
gy:
quan
tita
tive
•D
esig
n:la
rgel
yun
clea
r
Tan
zani
a,
Ken
yam
4RH
(L
’Eng
le,
2009
)•
Inte
rvie
ws
to p
rovi
de
insi
ght o
n ho
w
to d
esig
n a
mob
ile p
hone
pr
ogra
m fo
r re
prod
ucti
ve
heal
th in
Dar
es
Sal
aam
, T
anza
nia
and
Nai
robi
, K
enya
wer
e in
terv
iew
ed
•A
lmos
tal
lres
pond
ents
sai
dth
ey
wou
ld s
hare
info
rmat
ion
wit
h fa
mily
an
d fr
iend
s•
Res
pond
ents
wou
ldt
rust
fam
ily
plan
ning
info
rmat
ion
rece
ived
via
ph
one
•R
espo
nden
tss
aid
they
wan
ted
to
lear
n ab
out
cont
race
ptiv
e m
etho
ds
thro
ugh
text
mes
sage
•R
espo
nden
tss
aid
they
rea
dal
ltex
tm
essa
ges
and
wou
ld n
ot d
elet
e as
s u-
min
g a
mes
sage
is s
pam
bef
ore
read
ing
•Se
lect
app
ropr
iate
te
chno
logy
•E
nsur
epr
ivac
y
•T
ype
of
eval
uati
on:
form
ativ
e•
Met
hodo
logy
:m
ixed
met
hods
•D
esig
n:
inte
rvie
wed
40
clie
nts
at t
wo
fam
ily-p
lann
ing
clin
ics
99
Dow
nloa
ded
by [
] a
t 15:
09 0
3 M
ay 2
012
100 T. A. Gurman et al.
Selecting the appropriate technology (n = 10) includes studies that weighed the pros and cons of various technologies or explained the reasoning for choosing one technology over another. The two articles from Ghana use the “flashing” method of encouraging users to call or text a hotline, hang up, and then they were either called back for counseling (Clemmons, 2009) or added to a weekly voice health message program (Grameen Foundation, 2011). The pricing structure for phone use in Ghana makes this technique a practical choice, and one article cited SMS was an option, but 99% of users chose to receive voice calls (Grameen Foundation, 2011).
Ensuring privacy (n = 5) is a particular concern for vulnerable populations such as men who have sex with men. An intervention from Ghana discussed men who have sex with men flashing their number to a hotline and a counselor calling them back within 24 hours (Clemmons, 2009). While phone sharing can be an issue in terms of privacy, voice calls provide a greater buffer of privacy for those sharing phones than SMS.
Although five publications mention the ability of mobile technology to minimize cost, in comparison with other health care option, few articles document this effectively. The randomized clinical trial in China effectively demonstrated the statistical significant improvement in attendance of appointments when given SMS/voice reminders over no reminder, and the study showed no statistical difference between SMS or voice (p = .670; Chen, Fang, Chen, & Dai, 2008). This led the study to the conclusion that because SMS is cheaper than voice, 0.31 Yuan versus 0.48 Yuan, respectively, SMS is the more cost-effective option (Chen et al., 2008).
Considering comprehension (n = 10) refers to the linguistic and literacy competency of the audience as well as their ability to understand and use the technology. In one study in the Democratic Republic of the Congo, the concept of a hotline was not well understood (Corker, 2010). The study found that 20% of calls to the hotline were to ask for money, free phone credit or ask questions and discuss issues entirely unrelated to family planning (Corker, 2010). An article from Uganda that sent SMS quizzes to participants noted a 3.16 to 9.70% errors caused by people responding to SMS in conversational format instead of a quiz answer the computer database could read and convert.
The majority of publications (n = 12) used two-way communication in their communication message design. For example, a randomized clinical trial of 538 HIV-infected adults in Kenya taking antiretroviral treatments had an intervention group who received SMS reminders and a control group with no reminder (Lester et al., 2010). Every Monday, nurses would send a message to the patients asking how they were doing and request the reply of sawa (“doing well”) or shida (indicating a problem). Patients who did not respond in a 48-hour time frame would receive a call from a nurse (Lester et al., 2010). This example of two-way communication allowed users to indicate a problem or ask questions to a nurse instead of simply receiving a reminder (one-way).
Less than half (n = 6) of publications described targeting and tailoring the content for the intended audience. An article from Ghana discusses “mobile midwife,” which allows women or family members to sign up to receive voice messages once a week (Grameen Foundation, 2011). The messages are tailored specifically to her week of pregnancy, offered in several Ghanaian languages, and each message received she can choose to listen to a second or third message. The study found those who listen to the first message, 42% opt to listen to a second, and 36% opt to listen to a third message in a row (Grameen Foundation, 2011).
Dow
nloa
ded
by [
] a
t 15:
09 0
3 M
ay 2
012
mHealth Behavior Change Communication 101
Timing of communication (n = 3) requires studies to narrow down technology use from anytime/anywhere to an allotted time of day, frequency, or sequence of messages that is agreeable to the user. One article that dealt with urban sex workers in India considered time of day and length of message that would be least invasive to these individuals. The study found that 93.1% of urban sex workers listened to the entirety of a 19 second message, only 59% listened to a 30 second message (Sambasivan, Weber, & Cutrell, 2011).
Understanding the audience (n = 10) requires studies to discuss the population’s health needs and phone usage needs to best design an intervention. Articles varied in description of ethnographic research into cultural background and preferences regarding health interventions and technological communication. One article in particular described research to capture technology used, time of date contact, length of contact, frequency of contact, two- or one-way communication, and the types of messages offered to them (Sambasivan et al., 2011). This article used 21 in-depth interviews that described the phone use and demographic of urban sex workers. This group of individuals had high uptake of cell phones, 97% compared with women in India (Sambasivan et al., 2011). The formative research informed the decision to use voice messages instead of text for this population.
Long-term Evaluation (n = 2) requires long-term follow-up on a specific study or monitoring and reporting results of a program over a period of at least 1 year. The aforementioned SMS reminder program for antiretroviral treatment required self-reporting antiretroviral adherence to be monitored for 30 days with a follow-up at 6 and 12 months (Lester et al., 2010). Since many of the studies involve interventions of only a couple of weeks and no follow-up, this study offers a good example of long-term evaluation.
Discussion
This review of literature provided a full survey of evaluations for mHealth behavior change programs in low- and middle-income countries. The quality assessment in this review used several characteristics that should be incorporated into mHealth BCC interventions in such countries. The current literature offers a broad spectrum of quality regarding methodology and content, but most articles provided comprehensive information on the effectiveness of mHealth interventions.
The review began with 44 articles and was ultimately narrowed down to 16 articles of which 5 were peer-reviewed and 11 were gray literature. The majority of publications failed to meet the criteria of this study primarily because so much mHealth research is ongoing within the United States and other developed countries. Studies conducted in low- and middle-income countries that were excluded either were purely descriptive in nature or entirely lacked evaluation components. This finding provides a major gap in the literature and more attention should be directed toward programs with evaluation components.
The field of mHealth research is still in the infancy stage and there is a need for more thorough evaluation, follow-up from programs, and greater availability of research results publicly accessible on the Internet. Some articles had methodological flaws or lacked adequate sample sizes to draw statistically significant results, and three of the 16 articles did not provide methodological procedures of any kind; however, several articles provided adequate description of methods. As information and research become available, a more comprehensive
Dow
nloa
ded
by [
] a
t 15:
09 0
3 M
ay 2
012
102 T. A. Gurman et al.
review will be able to draw conclusions on the effectiveness of mHealth interventions on BCC.
The quality assessment for this review points to successes and limitations within the existing literature. Specific areas in the quality assessment that were used extensively among a broad range of articles included two-way communication (n = 12), technology platform selection (n = 10), and understanding the audience (n = 10), while those elements demonstrated to a lesser extent in the literature included long-term evaluation (n = 2) and timing of communication (n = 3).
Two-way communication, technology platform selection, and understanding the audience are all factors that deal with the target audience’s usage and preferences for technology. As noted above, these three areas were particularly well represented in the reviewed articles. However, timing of communication was one of the worst represented elements, but topically it should fit into understanding the technological needs of a population. This gap might exist because timing of communication requires pretesting of messages and more formative research than was undergone in the existing studies (Lavoie, Puleio, & Jha, 2009). Pretesting messages should be a priority to understand and learn about the specific audience’s preferences in technology use, language, and health needs.
While understanding the audience was one of the better addressed elements, rigorous methods should continue to be used that take into account cultural factors and population specific issues and needs. Although comprehension was discussed in terms of people understanding how to use technology, programs that use wide-scale mass text messaging services should better troubleshoot the systems to avoid system fluctuations and glitches in replying to participants. In addition, where hotlines are unknown concepts, mass media campaign or other techniques should be used to better educate the population about the service (Clemmons, 2009).
The lack of long-term evaluation could be a result of an emerging field that has yet to conduct this type of research, or a lack of resources being directed toward an important area. Future interventions should focus on incorporating long-term evaluation to show the lasting effects of mHealth interventions. Similarly, although some articles mentioned cost (n = 5), more should conduct cost-benefit analyses similar to (Chen et al., 2008) to demonstrate which technologies are most effective and cost efficient. Greater follow-up with study participants in the long-term should become a priority. Because the studies are so new, it is unknown if further follow-up or development of formative research will lead to more conclusive results. The evaluation process should continue even after the immediate project ends to understand the long-term effects of the program.
A possible limitation of this literature review is that three of the 16 articles were sponsored by the same organization, Text To Change; however, the articles studied different populations in different locations (Danis et al., 2010; Hoefman & Apunyo, 2010; “Using an interactive,” 2010). Because the field is still emerging, recent research could be under review for peer-reviewed journals, but the data are yet to be publicly available for review. In addition, the articles focused on Sub-Saharan Africa and Asia (India and China) but left many other regions of the world unaccounted for in the field of mHealth BCC.
Another limitation of this review is that the studies targeted several different populations. For instance, in some articles at-risk populations such as urban sex workers/female sex workers (Sambasivan et al., 2011) and men who have sex with men (Clemmons, 2009) were the targeted population, while other studies targeted higher
Dow
nloa
ded
by [
] a
t 15:
09 0
3 M
ay 2
012
mHealth Behavior Change Communication 103
income populations like educated women doing breast self-exams in India (Khokhar, 2009) and higher income populations for general health appointment reminders in China (Chen et al., 2008). As more research becomes available higher and lower income populations should be considered separately regarding their needs in mHealth studies. Likewise, issues such as privacy and gender are much greater concerns in vulnerable populations compared with high-income populations and should be considered separately in future reviews.
mHealth is a promising field of study that may improve the effect of BCC programs, but more studies need to be conducted with a greater emphasis on formative research and long-term evaluation. This review offers 10 main recommendations to incorporate into mHealth interventions in low- and middle-income countries and provides a status update on the areas of success and limitations. As the field continues to develop, mHealth reviews of BCC should further segment studies by income level or topic area, which was not possible in this review due to lack of available information.
References
Atun, R. A., Greene, S. A., Morris, S., Thomas, L., McKee, M., Franklin, V., … Soalen, R. (2006). The role of mobile phones in increasing accessibility and efficiency in healthcare. Retrieved from http://www.mobileactive.org/files/file_uploads/vodafone_policy_paper_4_march06.pdf
Chen,Z.,Fang,L.,Chen,L.,&Dai,H.(2008).ComparisonofanSMStextmessagingandphone reminder to improve attendance at a health promotion center: A randomized controlled trial. Journal of Zhejiang University Science, 9(1), 34–38.
Clemmons, L. (2009). Reaching MSM in Ghana with HIV and AIDS interventions. Most-at-risk-populations Technical Working Group Meeting, Chennai, India, February 18–24.
Corker, J. (2010). ‘Ligne verte’ toll-free hotline: Using cell phones to increase access to family planning information in the Democratic Republic of Congo. Cases in Public Health Communication & Marketing, 4, 23–37. Retrieved from http://www.gwumc.edu/sphhs/departments/pch/phcm/casesjournal/volume4/peer-reviewed/cases_4_03.pdf
Danis, C., Ellis, J., Kellogg, W., van Beijima, H., Hoefman, B., Daniels, S., & Loggers, J.-W. (2010). Mobile phones for health education in the developing world: SMS as a user interface. Paper presented at the First Annual Symposium on Computing for Development, London, United Kingdom, December 17–18. Retrieved from http://dev2010.news.cs.nyu.edu/papers/13-danis.pdf
Grameen Foundation. (2011). Mobile technology for community health in Ghana. Washington, DC: Author. Retrieved from http://www.mobileactive.org/research/mobile-technology-community-health-ghana
GSMA Development Fund. (2010). Women & mobile: A global opportunity, A study on the mobile phone gender gap in low and middle-income countries. London, United Kingdom: Author. Retrieved from http://www.mobileactive.org/research/women-mobile-global-opportunity-study-mobile-phone-gender-gap-low-and-middle-income-countri
Hoefman, B., & Apunyo, B. (2010). Using SMS for HIV/AIDS education and to expand the use of HIV testing and counseling services at the AIDS information Centre, Uganda. Mobile Communication Technology for Development (M4D), Kampala, Uganda.
K4Health. (2011). Planning for mHealth BCC: Questions to ask. Retrieved from http://www.k4health.org/toolkits/mhealth/planning-mhealth-bcc-questions-ask
Khokhar, A. (2009). Short text messages (SMS) as a reminder system for making working women from Delhi breast aware. Asian Pacific Journal of Cancer Prevention, 10, 319–322.
Lavoie, K., Puleio, M., & Jha, P. (2009). Family planning via mobile phones: Proof-of-concept testing in India. Washington, DC: Institute for Reproductive Health, Georgetown
Dow
nloa
ded
by [
] a
t 15:
09 0
3 M
ay 2
012
104 T. A. Gurman et al.
University. Retrieved from http://www.mobileactive.org/research/family-planning-mobile-phones
Lester, R. T., Ritvo, P., Mills, E. J., Kariri, A., Karanja, S., Chung, M. H., … Plummer, F. A. (2010). Effects of a mobile phone short message service on antiretroviral treatment adherence in Kenya (WelTel Kenya 1): A randomised trial. The Lancet, 376, 1838–1845.
McNamara, K. (2007). Improving health, connecting people: The role of ICTs in the health sector of developing countries (Framework Paper No. 7). Healthlink Worldwide, InfoDev, AfriAfya, ISHED. Retrieved from http://www.docstoc.com/docs/44938925/Improving-health-connecting-people-the-role-of-ICTs-in-the
Mechael, P. (2009). The case for mHealth in developing countries. Innovations. 4(1), 103–118. Mechael, P., Batavia, H., Kaonga, N., Searle, S., Kwan, A., Goldberger, A., … Ossman, J. (2010).
Barriers and gaps affecting mHealth in low and middle income countries: Policy white paper. New York, NY: Center for Global Health and Economic Development, The Earth Institute, Columbia University.
Mechael, P., & Sloninsky, D. (2007). Toward the development of an mHealth strategy. Geneva, Switzerland: World Health Organization.
Sambasivan, N., Weber, J. S., & Cutrell, E. (2011). Designing a phone broadcasting system for urban sex workers in India. Paper presented at the ACM SIGCHI Conference on Human Factors in Computing Systems, Vancouver, British Columbia, Canada, May. Retrieved from http://www.ics.uci.edu/~nsambasi/CHI2011_SexWork.pdf
United Nations Foundation/Vodafone Foundation. (2009). mHealth for development: The opportunity of mobile technology for healthcare in the developing world. Washington, DC: Authors. Retrieved from http://www.globalproblems-globalsolutions-files.org/unf_website/assets/publications/technology/mhealth/mHealth_for_Development_full.pdf
Using an interactive health service, SMS messaging, to improve maternal and child health service delivery. (2010). Nakawa, Kampala, Uganda: Text to Change. Retrieved from http:www.texttochange.org/news/using-text-and-voice-messages-improve-maternal-and-child-health-uganda