Medication adherence and predictive factors in patients with cardiovascular disease: A comparison study between Australia and Iraq

Abstract

Background: Adherence to cardiac medication regimes is essential for effective treatment of cardiovascular disease but is unsatisfactory in Australia and little studied in Iraq. Aim: This study evaluated and compared adherence to cardiac medications and potentially predictive factors based on the Theory of Planned Behaviour (TPB) in patients with cardiovascular disease admitted to hospital and attending cardiac services in Australia and Iraq. Methods: A cross-sectional multi-centre comparative study involving 246 cardiac patients was conducted in Australia (one hospital in Sydney) and Iraq (three cardiac hospitals in Baghdad) between October 2016 and December 2017. Adherence to medications and related factors were examined using established, validated questionnaires, formally translated and validated into Arabic for Iraqi participants. Binary logistic regression was conducted to determine those factors independently predictive of cardiac medication adherence, in Australia and Iraq. Findings: A significantly higher proportion (64.3%) of Iraqi than Australian (37.5%) cardiac patients reported medium/low levels of adherence to their cardiac medications. After adjusting for confounding factors, the ability to correctly self-administer and refill medications, and beliefs about cardio-protective medication were identified as independent predictors of cardiac medication adherence behaviour in both Australian and Iraqi participants. In Iraq, patients recruited from out-patient cardiac clinics were significantly more likely to report adherent behaviours that patients recruited as in-patients of the cardiac ward. Conclusion: Non-adherence to cardiac medications differed but was sub-optimal in both Australian and Iraqi patient samples, in both countries, adherence was associated with patients’ beliefs about medications, and ability to self-administer and refill medications. Clinical nurses and pharmacists need to investigate these factors at every point in the cardiac trajectory to optimise medication adherence.

Full text

Collegian
26
(2019)
355–365
Contents
lists
available
at
ScienceDirect
Collegian
j
o
ur
nal
ho
mepage:
www.elsevier.com/locate/coll
Medication
adherence
and
predictive
factors
in
patients
with
cardiovascular
disease:
A
comparison
study
between
Australia
and
Iraq
Ali
Hussein
Alek
Al-Ganmia,b,∗,
Sadeq
Al-Fayyadhb,
Mohammed
Baqer
Habeeb
Abd
Alib,
Abdulellah
Modhi
Alotaibia,c,
Leila
Gholizadeha,
Lin
Perrya,d
aUniversity
of
Technology
Sydney,
Faculty
of
Health,
UTS
Building
10,
Level
7,
235-253
Jones
Street,
Ultimo,
Sydney,
NSW,
2007,
Australia
bUniversity
of
Baghdad,
College
of
Nursing,
Bab
Al-Muadham
Square,
Baghdad,
Iraq
cShqra
University,
Faculty
of
Applied
Health
Sciences,
Shqra,
Saudi
Arabia
dSouth
Eastern
Sydney
Local
Health
District,
Prince
of
Wales
Hospital,
G74,
East
Wing,
Edmund
Blacket
Building,
Barker
Street,
Randwick,
NSW,
2031,
Australia
a
r
t
i
c
l
e
i
n
f
o
Article
history:
Received
1
July
2018
Received
in
revised
form
7
September
2018
Accepted
9
October
2018
Keywords:
Cardiovascular
disease
Australia
Iraq
Cardiology
Medication
adherence
Beliefs
Medication
refill
Cardiac
nursing
Theory
of
planned
behaviour
a
b
s
t
r
a
c
t
Background:
Adherence
to
cardiac
medication
regimes
is
essential
for
effective
treatment
of
cardiovas-
cular
disease
but
is
unsatisfactory
in
Australia
and
little
studied
in
Iraq.
Aim:
This
study
evaluated
and
compared
adherence
to
cardiac
medications
and
potentially
predictive
factors
based
on
the
Theory
of
Planned
Behaviour
(TPB)
in
patients
with
cardiovascular
disease
admitted
to
hospital
and
attending
cardiac
services
in
Australia
and
Iraq.
Methods:
A
cross-sectional
multi-centre
comparative
study
involving
246
cardiac
patients
was
con-
ducted
in
Australia
(one
hospital
in
Sydney)
and
Iraq
(three
cardiac
hospitals
in
Baghdad)
between
October
2016
and
December
2017.
Adherence
to
medications
and
related
factors
were
examined
using
established,
validated
questionnaires,
formally
translated
and
validated
into
Arabic
for
Iraqi
participants.
Binary
logistic
regression
was
conducted
to
determine
those
factors
independently
predictive
of
cardiac
medication
adherence,
in
Australia
and
Iraq.
Findings:
A
significantly
higher
proportion
(64.3%)
of
Iraqi
than
Australian
(37.5%)
cardiac
patients
reported
medium/low
levels
of
adherence
to
their
cardiac
medications.
After
adjusting
for
confounding
factors,
the
ability
to
correctly
self-administer
and
refill
medications,
and
beliefs
about
cardio-protective
medication
were
identified
as
independent
predictors
of
cardiac
medication
adherence
behaviour
in
both
Australian
and
Iraqi
participants.
In
Iraq,
patients
recruited
from
out-patient
cardiac
clinics
were
significantly
more
likely
to
report
adherent
behaviours
that
patients
recruited
as
in-patients
of
the
cardiac
ward.
Conclusion:
Non-adherence
to
cardiac
medications
differed
but
was
sub-optimal
in
both
Australian
and
Iraqi
patient
samples,
in
both
countries,
adherence
was
associated
with
patients’
beliefs
about
med-
ications,
and
ability
to
self-administer
and
refill
medications.
Clinical
nurses
and
pharmacists
need
to
investigate
these
factors
at
every
point
in
the
cardiac
trajectory
to
optimise
medication
adherence.
©
2018
Australian
College
of
Nursing
Ltd.
Published
by
Elsevier
Ltd.
∗Corresponding
author
at:
University
of
Technology
Sydney,
Faculty
of
Health,
UTS
Building
10,
Level
7,
235-253
Jones
Street,
Ultimo,
Sydney,
NSW,
2007,
Australia;
University
of
Baghdad,
College
of
Nursing,
Bab
Al-Muadham
Square,
Baghdad,
Iraq.
E-mail
address:
ali.h.al-ganmi@student.uts.edu.au
(A.H.A.
Al-Ganmi).
https://doi.org/10.1016/j.colegn.2018.10.002
1322-7696/©
2018
Australian
College
of
Nursing
Ltd.
Published
by
Elsevier
Ltd.

356
A.H.A.
Al-Ganmi
et
al.
/
Collegian
26
(2019)
355–365
Summary
of
relevance
Problem
Cardiovascular
disease
remains
the
major
cause
of
mortality
in
developed
and
developing
countries.
What
is
already
known
Cardiovascular
medication
adherence
is
suboptimal,
increas-
ing
the
risk
of
morbidity.
Few
medication
adherence
investigations
have
been
conducted
in
Australia
and
none
have
been
published
from
Iraq.
What
this
paper
adds
This
is
the
first
study
to
compare
adherence
to
cardiac
medications
and
factors
potentially
predictive
of
medica-
tion
non-adherence
in
patients
with
cardiovascular
disease
in
Australia
and
Iraq.
Outcomes
can
support
the
development
of
nurse-led
behavioural
approaches
to
improve
medication
adherence
in
patients
with
cardiovascular
disease.
Findings
flag
the
importance
for
nurses
should
focus
on
patients’
beliefs
about
their
medication
and
their
ability
to
self-administer
and
refill
medications,
taking
into
account
the
socio-eco-cultural
factors
on
medication
adherence.
1.
Introduction
Cardiovascular
disease
(CVD)
is
the
major
cause
of
death
in
both
developed
and
developing
countries
(Wirtz,
Kaplan,
Kwan,
&
Laing,
2016).
CVD
encompasses
a
variety
of
disease
entities
and
specific
symptom
complexes,
treated
by
a
variety
of
approaches,
pharmaceutical,
surgical
and
interventional
radiology.
In
devel-
oped
countries,
the
annual
mortality
associated
with
CVD
in
2013
was
4
million
people,
with
more
than
1.4
million
dying
prematurely
before
the
age
of
75
years
(Townsend
et
al.,
2016).
In
Australia,
CVD
accounts
for
18%
of
the
total
burden
of
disease
and
was
respon-
sible
for
43,963
deaths
in
2016
and
more
than
490,000
hospital
admissions
in
2014–2015
(Heart
Foundation,
2016).
According
to
the
Australian
Pharmaceutical
Benefits
Scheme
(PBS),
the
total
cost
of
CVD
medications
in
2015-16
was
AU$1.448
billion
for
a
total
estimated
population
of
23.781.200
people
(Australian
Bureau
of
Statistics,
2015),
comprising
20%
of
the
total
health
expenditure
(Australian
Institute
of
Health
&
Welfare,
2017).
Prevalence
rates
of
CVD
are
increasing,
and
particularly
in
Middle
Eastern
countries
such
as
Iraq,
where
CVD
accounted
for
27,500
deaths
in
2012,
16.5%
of
all-cause
mortality
(World
Health
Organization,
2013).
The
Iraqi
government
subsidises
the
cost
of
a
wide
range
of
prescription
medicines,
including
those
for
CVD.
Government
health
expenditure
was
US$82.2
billion
in
2010
with
an
estimated
39%
spent
on
cardiovascular
medications,
totalling
around
US$32
billion
for
a
population
of
30.868.156
people
(Al
Hilfi,
Lafta,
&
Burnham,
2013).
The
standard
of
treatment
and
care
remains
suboptimal
in
Iraq
and
the
health
care
system
is
poorly
effective
in
managing
long-term
disease
such
as
CVD
(Alwan,
2004).
Long-term
management
of
CVD
is
largely
based
on
life-style
modifications
such
as
diet,
physical
exercises,
smoking
cessation
and
medications
are
usually
required
over
sustained
periods,
often
for
life
(Australian
Institute
of
Health
&
Welfare,
2017).
Hence,
adherence
to
medication
regimes
and
the
ability
to
maintain
treat-
ment
as
prescribed
is
essential
to
managing
symptoms,
delaying
or
preventing
disease
progression,
premature
disability
and
death
(World
Health
Organisation,
2003).
Unless
patients
living
with
CVD
are
able
to
maintain
treatments
as
prescribed,
the
resources
com-
mitted
by
the
pharmaceutical
industry
and
healthcare
providers
to
drug
development,
diagnosis
and
prescription
are
wasted.
2.
Background
Medication
adherence
is
defined
as
the
extent
to
which
patients
take
medications
as
prescribed
by
their
health
care
providers
(World
Health
Organisation,
2003).
Poor
medication
adherence
is
a
key
factor
impeding
disease
control
among
those
with
CVD
(World
Health
Organisation,
2003).
Despite
evidence
of
the
effectiveness
of
cardiac
medications
such
as
anti-platelet
agents,
beta-blockers,
angiotensin-converting
enzyme
inhibitors
(ACEIs),
angiotensin
receptor
blockers
(ARBs),
and
aldosterone
receptor
antagonists
(ARA)
(Levy,
Huang,
Huang,
&
Michael
Ho,
2018)
to
manage
CVD
symptoms,
non-adherence
rates
range
from
44.4%
to
61%
across
settings
and
countries
(Paradkar
&
Sinha,
2017).
Studies
report
that
adherence
to
cardio-protective
medications
is
sub-optimal
in
both
resource-limited
and
resource-rich
countries
(Chowdhury,
Khan,
&
Heydon,
2013).
This
is
a
major
problem
internationally;
as
such,
there
may
be
lessons
to
learn
from
multi-national
comparisons,
studying
the
problem
in
different
settings
and
contexts.
Long-term
cardiac
medication
adherence
by
Australian
patients
with
CVD
is
unsatisfactory,
with
non-adherence
rates
ranging
between
14%
and
43%
from
2010
to
2014
(McKenzie,
McLaughlin,
Clark,
&
Doi,
2015).
To
date,
few
investigations
of
medication
adher-
ence
have
been
conducted
with
Australian
patients
with
chronic
conditions,
especially
patients
with
CVD,
and
no
published
stud-
ies
report
adherence
to
cardiovascular
medications
in
Iraq.
Many
medication
adherence
studies
have
been
conducted
in
developed
countries,
but
it
is
difficult
to
extrapolate
the
results
of
these
studies
to
developing
countries
such
as
Iraq
due
to
disparities
in
health-
care
systems,
social
and
cultural
factors
that
affect
health
beliefs
and
practice
(Sharrad,
Hassali,
&
Shafie,
2009).
Medication
non-
adherence
is
a
complex
and
multi-faceted
problem,
influenced
by
a
wide
range
of
sociodemographic
and
economic
characteris-
tics,
behavioural
and
cognitive
problems,
medication
complexity
and
social
support
systems
(Marshall,
Wolfe,
&
McKevitt,
2012).
Cultural
differences
among
patients
from
diverse
national
back-
grounds,
racial
and
ethnic
groups
may
contribute
to
disparities
of
cardiovascular
medication
adherence
(Traylor,
Schmittdiel,
Uratsu,
Mangione,
&
Subramanian,
2010).
Cultural
differences
can
neg-
atively
influence
patients’
beliefs
about
medication
and
their
perceptions
of
treatment,
and
difference
in
generation,
household
composition
and
religion
may
also
create
variability
in
beliefs
and
behaviours
(Horne
et
al.,
2004).
Multiple
different
factors
may,
therefore,
contribute
to
medication
non-adherence
in
developed
and
developing
countries
(Cooney
et
al.,
2009).
In
patients
with
CVD,
negative
perceptions
of
medications
(Choudhry
et
al.,
2011),
beliefs
about
benefits
and
harms
of
medications
(Horne
et
al.,
2013),
low
levels
of
social
support
(Park,
Howie-Esquivel,
Whooley,
&
Dracup,
2015),
low
self-efficacy
(Greer,
Milner,
Marcello,
&
Mazin,
2015),
poor
medication
self-
management,
attitudes
and
sociocultural
norms
(Martin,
Williams,
Haskard,
&
DiMatteo,
2005)
and
problems
refilling
medications
have
all
been
found
to
influence
adherence
to
cardiac
medica-
tions
(Kripalani,
Risser,
Gatti,
&
Jacobson,
2009).
Other
contributing
factors
include
the
complexity
of
treatment,
patients’
forgetful-
ness
(Karakurt
&
Kas¸
ikc¸
i,
2012),
the
presence
of
comorbidities
(Al
Qasem,
Smith,
&
Clifford,
2011)
and
medication
side
effects
(Molloy
&
O’Carroll,
2017).
At
an
individual
level,
the
engagement
of
individuals
in
health-related
behaviours,
such
as
medication
adherence,
may
be
explained
by
applying
behavioural
theories.
This
study
used
Ajzen
(1991)
Theory
of
Planned
Behaviour
(TBP)
to
identify
factors
pre-
dictive
of
medication
adherence
in
patients
with
CVD
(Fig.
1).
The
TBP
encompasses
attitudes
(e.g.
positive
or
negative
beliefs
about
the
behaviour),
subjective
norms
(e.g.
the
degree
to
which
social
consensus
values
and
role
models
reinforce
behaviours;
per-
ceived
social
support
to
perform
the
behaviour),
and
perceived
behavioural
control
(e.g.
perceptions
of
self-efficacy
to
perform
the
behaviour)
(Table
1).
The
theory
explains
how
these
factors
might
impact
patients’
intentions
or
directly
influence
their
behaviours

A.H.A.
Al-Ganmi
et
al.
/
Collegian
26
(2019)
355–365
357
Table
1
Study
instruments,
adherence
characteristics
measured
and
their
relation
to
the
Theory
of
Planned
Behaviour.
Instruments
Behaviours
Measured
The
Theory
of
Planned
Behaviour
Elements
Medication
Adherence
Questionnaire
(4-item
Morisky
scale)
(MAQ)
•
Forgetfulness,
•
Carelessness,
•
Adverse
effects
•
Efficacy
•
Behaviour
(adherence
to
medication)
The
Adherence
to
Refills
and
Medications
Scale
(ARMS)
•
Correct
medications
self-administration
•
Ability
to
refill
medication
on
schedule
•
Behaviour
(ability
to
self-administer
and
refill
medications)
The
Belief
about
Medicine
Questionnaire
(BaMQ)
•
Belief
about
the
necessity
of
medications
•
Belief
and
concerns
about
the
medication
•
Belief
about
the
overuse
of
medication
•
Belief
about
harms
of
medication
•
Attitudes
toward
the
behaviour
(predisposition
toward
adherence)
The
Medication
Adherence
Self-Efficacy
Scale-Revised
(MASES-R)
•
Confidence
in
taking
medications
•
Ability
to
take
medications
as
part
of
everyday
routine
•
Perceived
behavioural
control
(adherence
under
the
person’s
control)
Medication
Specific
Social
Support
(MSSS) •
Social
support •
Subjective
norms
(perceived
social
support
for
adherence)
such
as
adherence
to
cardiac
medications
(Armitage
&
Conner,
2001).
The
social
and
cultural
contexts
of
Australia
and
Iraq
differ,
as
developed
and
developing
nations
of
the
West
and
the
Mid-
dle
East.
However,
both
have
high
prevalence
rates
of
CVD,
and
for
both
it
is
crucial
to
understand
adherence
to
CVD
medica-
tions
and
factors
which
predict
this.
Cross-national
comparisons
might
offer
information,
for
example
cultural
insights,
not
avail-
able
with
a
single
country
study.
The
aim
of
this
study,
therefore,
was
to
explore
medication
adherence
behaviours
of
patients
with
CVD,
admitted
to
hospital
and
attending
out-patient
services,
in
Australia
and
Iraq
and
to
determine
factors
predictive
of
adherence
to
cardiac
medications
regimes.
Findings
could
inform
develop-
ment
of
preventive
strategies
to
improve
medication
adherence
internationally
through
tailored
nurse-led
adherence
interventions
utilising
reliable
adherence
assessment
tools
and
the
most
effective
and
appropriate
adherence
approaches.
3.
Methods
3.1.
Aims
and
objectives
The
aim
of
this
study
was
to
evaluate
and
compare
adherence
to
cardiac
medications
and
factors
potentially
predictive
of
this
in
patients
with
CVD
in
Australia
and
Iraq.
Objectives
were
to:
1)
Identify
and
compare
the
levels
of
cardiac
medication
adherence
in
patients
with
CVD
in
Australia
and
Iraq.
2)
Examine
and
compare
socio
demographic,
health
related,
attitudinal
and
behavioural
factors
potentially
predictive
of
medication
adherence
in
these
countries
using
the
lens
of
the
Theory
of
Planned
Behaviour
(TPB)
(Ajzen,
1991).
3.2.
Design
This
was
a
cross-sectional
multi-centre
comparative
study
con-
ducted
in
Australia
(one
hospital
in
Sydney)
and
Iraq
(three
cardiac
hospitals
in
Baghdad).
Survey
design
was
used
to
explore
cardiac
patients’
medication
adherence.
3.3.
Study
settings
In
Australia,
participants
were
recruited
from
the
in-patient
cardiology
ward
and
out-patient
cardiac
rehabilitation
centre
of
Fig.
1.
Components
of
the
Theory
of
Planned
Behaviour
model
(TPB)
(Ajzen,
1991)
used
to
explain
medication
adherence
in
this
study.
an
acute
tertiary
hospital
in
Sydney.
This
hospital
has
inpatient
diagnostic
and
interventional
cardiac
services
including
a
cardio-
thoracic
intensive
care
and
sub-acute
surgical
ward,
a
coronary
care
unit
and
sub-acute
cardiology
ward.
The
out-patient
cardiac
reha-
bilitation
service
screened
in-patients
and
external
referrals
for
delivery
of
cardiac
rehabilitation
programs
including
structured,
supervised
group
exercise
and
information
sessions,
and
referral
to
relevant
services.
In
Iraq
patients
were
recruited
at
three
cardiac
hospitals
in
Baghdad,
Iraq.
These
were
major
tertiary
teaching
hospitals
offering
a
range
of
cardiac
services
including
cardiac
out-patient
clinics
and
in-patient
cardiac
services
for
approximately
570
beds
with
intensive
and
coronary
care
units,
cardiology
and
cardiac
surgery
departments.
These
three
public
hospitals
are
operated
by
the
Iraqi
Ministry
of
Health,
which
provides
cardiac
services
for
people
referred
nation-wide,
including
cardiac
emergency
and
critical
care
cases.
At
the
time
of
this
study,
most
Iraqi
hospitals
were
struggling
to
recover
from
years
of
war,
shortages
in
health
facilities
and
the
health
workforce,
with
inadequately
trained
healthcare
professionals
and
poor
health-care
finance
(The
Iraqi
Ministry
of
Health,
2012).

358
A.H.A.
Al-Ganmi
et
al.
/
Collegian
26
(2019)
355–365
3.4.
Participants,
sampling
and
sample
size
Participants
were
patients
who
had
been
admitted
to
hospital
for
an
acute
cardiac
event
or
referred
to
and
attended
a
cardiac
reha-
bilitation
program
or
out-patient
cardiac
clinic
between
October
2016
and
December
2017.
Similar
inclusion
criteria
were
applied
in
both
countries:
18
years
of
age
or
older;
diagnosed
with
cardiac
disease;
currently
taking
at
least
one
cardio-protective
medication
and
having
primary
responsibility
for
taking
their
own
medication
before
admission
to
hospital
or
when
attending
cardiac
rehabil-
itation.
Participants
were
required
to
be
able
to
read,
speak
and
understand
English
(in
Australia)
or
Arabic
(in
Iraq).
Patients
who
were
blind,
deaf
or
medically
deemed
unable
to
provide
consent
were
excluded.
Patients
in
the
cardiac
ward
were
excluded
if
they
were
newly
diagnosed
with
no
previous
history
of
cardiovascu-
lar
disease
and
hence
no
prior
history
of
taking
prescribed
cardiac
medication.
Recruitment
occurred
under
the
supervision
of
the
clinical
nurse
consultant
for
cardiac
rehabilitation
and
the
clinical
pharmacist
for
the
cardiac
ward
in
Australia
and
the
clinical
nurse
specialist
for
all
sites
in
Iraq,
with
the
agreement
of
the
directors
of
nursing
and
the
cardiology
consultants
of
both
sites.
Patients
were
screened
using
the
study
inclusion/exclusion
criteria
by
the
clinical
nurse
consultant/specialist
and
the
clinical
pharmacist.
Eligible
members
of
consecutive
cohorts
of
patients
who
expressed
interest
in
the
project
were
referred
to
the
researcher.
In
Australia,
approvals
to
conduct
this
study
were
granted
by
the
appropriate
health
district
and
university
Human
Research
Ethics
Committees
in
June
2016
(references:
16/085
(16/085
-HREC/16/POWH/218;
ETH16-0635),
and
the
health
district
Human
Research
Ethics
Committees
from
the
three
hospitals
in
Iraq
approved
the
study
in
June
2017.
Partici-
pants’
privacy
and
confidentiality
were
maintained
at
all
times
and
informed
consent
obtained
from
all
participants.
3.5.
Sample
size
determination
A
sample
size
of
120
participants
in
each
country
was
calcu-
lated
to
demonstrate
a
moderate
sized
effect
(␣
=
0.05,
5%
level
of
significant)
and
power
=
0.80
(Ma,
Zhou,
Zhou,
&
Huang,
2014).
We
anticipated
approximately
50%
of
eligible
patients
might
partici-
pate.
3.6.
Data
collection
procedures
In
order
to
assess
medication
adherence
and
associated
factors,
a
valid,
reliable
and
culturally
acceptable
instrument
is
required.
Data
collection
procedures
at
both
sites
were
performed
under
the
same
methods
and
conditions
using
the
same
instruments
for
the
respective
language
after
formal
translation
procedures
had
been
performed
in
Iraq.
In
both
Australia
and
Iraq,
data
collection
took
place
during
patients’
stay
in
the
inpatient
cardiac
ward,
during
attendance
at
cardiac
rehabilitation
sessions,
or
when
visiting
an
outpatient
clinic.
The
study
used
a
paper-based
self-report
questionnaire.
At
enrolment,
participants
received
a
survey
package
including
an
information
sheet,
a
consent
form
and
the
study
instruments.
The
researcher
also
verbally
explained
the
purpose
and
methods
of
the
study
face-to-face.
3.7.
The
survey
The
survey
comprised
a
number
of
validated
questionnaires
designed
to
gather
data
about
medication
adherence,
patient
beliefs,
behavioural,
attitudinal
and
other
factors
associated
with
adherence/non-adherence.
For
Iraqi
participants,
all
question-
naires
were
formally
translated
and
validated
in
the
Arabic
language
using
best
practice
translation
guidelines
(Sousa
&
Rojjanasrirat,
2011).
Study
instruments
were
translated
into
Arabic
and
then
back-
translated
to
English
by
two
researchers
independently
(Epstein,
Santo,
&
Guillemin,
2015).
The
researcher
used
the
back-translation
method
for
achieving
linguistic
translations
equivalent
to
the
items
in
the
source
language
(Duffy,
2006).
Following
the
instrument
translation
recommendations,
an
expert
translator
translated
the
English
versions
of
the
questionnaires
into
Arabic.
The
Arabic
translated
version
was
then
back-translated
to
English
by
an
inde-
pendent
expert
with
rich
clinical
experience
in
cardiology
nursing,
bilingual
and
fluent
in
both
English
and
Arabic
languages
but
unfa-
miliar
with
the
study.
The
researcher
then
discussed
and
resolved
minor
rewording
conflicts.
All
translations
were
conducted
using
clear,
direct
and
simple
phrases
and
questions
suitable
for
older
adults
with
low
literacy
levels.
Content
and
face
validity
of
the
study
questionnaires
were
assessed
and
confirmed
by
seven
content
experts
with
extensive
experience
in
cardiac
disease
of
various
spe-
cialty
backgrounds
(cardiologist,
senior
pharmacists,
cardiac
nurse
specialist
and
nursing
educators).
They
reviewed
the
question-
naires
using
the
Content
Validity
index,
resulting
in
the
wording
of
some
items
being
slightly
revised.
The
translated
questionnaires
were
piloted
with
five
cardiac
patients
from
the
three
Iraqi
hospi-
tals;
further
minor
modifications
were
made
so
the
questionnaires
could
be
easily
understood.
Sociodemographic
and
health
data
were
collected,
and
completion
took
10–15
minutes
(Table
2).
3.7.1.
Medication
adherence
questionnaire
The
Medication
Adherence
Questionnaire
(MAQ)
scale
(Morisky,
Green,
&
Levine,
1986),
is
a
short,
self-reported
measure
designed
to
assess
medication
adherence
behaviour
and
barriers
such
as
for-
getfulness,
carelessness,
adverse
effects
and
efficacy.
It
contains
four
simple
dichotomous
questions
(yes/no),
scoring
one
point
for
each
“yes”
response
and
zero
point
for
each
“no”
response.
Scores
are
summed
to
derive
a
total
score
with
higher
numbers
demon-
strating
greater
medication
non-adherence
i.e.
0
=high
to
3–4=
low
medication
adherence
behaviours.
The
validity
and
reliability
of
the
MAQ
were
originally
established
in
patients
with
hypertension,
with
internal
consistency
of
␣
=
0.61,
sensitivity
of
0.81
and
speci-
ficity
of
0.44
(Lavsa,
Holzworth,
&
Ansani,
2011).
The
MAQ
score
has
been
demonstrated
to
be
a
significant
independent
predictor
of
cardiovascular
medication
nonadherence
by
multivariate
logistic
regression
(Shalansky,
Levy,
&
Ignaszewski,
2004).
3.7.2.
The
adherence
to
refills
and
medications
scale
The
12-item
scale
Adherence
to
Refills
and
Medications
Scale
(ARMS)
(Kripalani
et
al.,
2009)
was
used
to
determine
medication
adherence
self-regulation.
The
ARMS
consists
of
12-items,
the
first
8
items
in
the
scale
assessing
the
ability
to
self-administration
for
the
prescribed
medications
and
the
last
4
items
evaluating
the
patient’s
ability
to
take
medications
on
schedule.
Each
item
is
scored
on
a
four-point
scale
ranging
from
l
=
none
of
the
time
to
4
=
all
the
time.
Scores
are
summed
to
derive
a
total
score
with
higher
numbers
demonstrating
better
refill
ability
for
medications
on
schedule.
The
ARMS
and
its
subscales
have
been
shown
to
cor-
relate
highly
with
other
measures
of
medication
adherence
such
as
the
four-item
scale
by
Morisky
and
colleagues,
and
medica-
tion
refill
adherence
(Mayberry,
Gonzalez,
Wallston,
Kripalani,
&
Osborn,
2013).
The
ARMS
has
demonstrated
high
internal
consis-
tency
among
patients
with
low
literacy
skills
(Kripalani
et
al.,
2009).
3.7.3.
The
belief
about
medicine
questionnaire
The
short
(eight-item)
version
of
the
Belief
about
Medicine
Questionnaire
(BaMQ)
(Horne,
Weinman,
&
Hankins,
1999)
was
used
to
evaluate
patients’
beliefs
about
medications.
The
BaMQ
identifies
patients’
beliefs
in
the
necessity
of
and
concerns
about

A.H.A.
Al-Ganmi
et
al.
/
Collegian
26
(2019)
355–365
359
Table
2
Sociodemographic
and
health
characteristics
of
cardiac
patients
recruited
in
Australia
and
Iraq.
Variables
Total
(N
=
246)
Australian
Patients
(N
=
120)
Iraqi
patients
(N
=
126)
Chi2test
df
P-values
Gender
n
(%)
Male
Female
161
(65.4)
85
(34.6)
78
(65.0)
42
(35.0)
83
(65.9)
43
(34.1)
0.21
1
0.89
Patient
services
n
(%)
Cardiac
Rehabilitation/
Outpatient
clinics
Cardiac
ward
67
(27.2)
179
(72.8)
31
(25.8)
89
(74.2)
36
(28.6)
90
(71.4)
0.23
1
0.63
Employment
Status
n
(%)
Employed
Unemployed
Retired
76
(30.9)
72
(29.3)
98
(39.8)
31
(25.8)
11
(9.2)
78
(65.0)
45
(35.7)
61
(48.4)
20
(15.9)
71.52
2
0.001*
Living
Arrangement
n
(%)
Lives
alone
Lives
with
spouse/partner/others
41
(16.7)
205
(83.3)
27
(22.5)
93
(77.5)
14
(11.1)
112
(88.9)
5.74
1
0.017*
Marital
Status
n
(%)
Married/
co-habiting
Single
194
(78.9)
52
(21.1)
77
(64.2)
43
(35.8)
117
(92.9)
9
(7.1)
30.35
1
0.001*
Education
n
(%)
<Year
11
Year
12
Certificate
or
Diploma
Bachelor
Degree
Graduate
Qualification
95
(38.6)
71
(28.9)
41
(16.7)
28
(11.4)
11
(4.5)
49
(40.8)
20
(16.7)
28
(23.3)
13
(10.8)
10
(8.3)
46
(36.5)
51
(40.5)
13
(10.3)
15
(11.9)
1
(0.8)
26.49
4
0.001*
Comorbidity
n
(%)
None
Any
168
(68.3)
78
(31.7)
94
(78.3)
26
(21.7)
74
(58.7)
52
(41.3)
10.90
1
0.001*
Diabetes
Mellitus
n
(%)
No
Yes
157
(63.8)
89
(36.2)
77
(64.2)
43
(35.8)
80
(63.5)
46
(36.5)
0.12
1
0.912
Medications
recall
n
(%)
Recalled
all
medications
Can’t
recall
all
medications
155
(63.0)
91
(37.0)
83
(69.2)
37
(30.8)
72
(57.1)
54
(42.9)
3.81
1
0.51
t-test
df
P-values
Age,
mean
(SD)*/
years
69.5
(13.9)
69.1
(11.6)
54.3
(12.0)
t
=
9.82
244
0.001**
Number
of
cardiac
medications
taken
per
day;
mean
(SD)
2.9
(1.2)
2.5
(0.98)
3.3
(1.2)
t
=
−5.71
244
0.001**
Note:
(SD)
Standard
Deviation,
*Significant
at
0.05,
**Significant
at
0.01.
their
medicines.
The
BaMQ
is
composed
of
subscales
(8
items)
to
assess
specific
necessity,
specific
concerns,
general
overuse
and
general
harm.
Each
item
is
measured
using
a
5-point
Lik-
ert
scale
range
from
l
=
strongly
disagree
to
5
=
strongly
agree.
Scores
are
summed
to
derive
a
total
score
with
higher
numbers
demonstrating
more
positive
beliefs.
The
internal
consistency
of
the
BMQ
scale
has
been
evaluated
using
Cronbach’s
alpha,
show-
ing
inter-item
correlation
of
General
Harm
=
0.44,
p
<
0.01);
General
Overuse
=
0.64,
p
<
0.01);
Specific
Necessity
=
0.79,
p
<
0.01);
and
Specific
Concerns
=
0.56,
p
<
0.01)
(Schüz
et
al.,
2011).
The
BaMQ
has
been
shown
to
correlate
significantly
with
other
adherence-
related
scales
such
as
the
MAQ,
the
Morisky
Medication
Adherence
Scale
(MMAQ),
and
medication
adherence
rating
scale
(MARS-5)
(Mårdby,
Åkerlind,
&
Jörgensen,
2007).
3.7.4.
The
medication
adherence
self-efficacy
scale-revised
The
13-item
Medication
Adherence
Self-Efficacy
Scale-Revised
(MASES-R)
(Fernandez,
Chaplin,
Schoenthaler,
&
Ogedegbe,
2008)
was
used
to
evaluate
an
individual’s
ability
to
adhere
to
their
medication
schedule
under
various
challenging
circumstances.
The
MASES-R
consists
of
items
that
specifically
examine
patients’
con-
fidence
in
taking
medications
in
specific
circumstances
and
their
ability
to
take
medications
as
part
of
everyday
routine.
Each
item
is
measured
using
a
4-point
Likert
scale,
ranging
from
0
=
not
at
all
sure
to
3
=
extremely
sure,
with
greater
self-efficacy
indicated
by
higher
scores
and
a
single
score
derived
as
the
mean
of
all
items.
The
MASES-R
has
been
found
correlate
to
significantly
with
electronic
medication
adherence
records
(MEMS)
at
3-months,
indicating
support
for
the
predictive
validity
of
the
MASES-R
(Fernandez
et
al.,
2008).
3.7.5.
Medication
specific
social
support
An
eight-item
Medication
Specific
Social
Support
(MSSS)
scale
(Lehavot
et
al.,
2011)
was
used
to
identify
how
often
patients
receive
assistance
from
others
with
their
medication.
Each
item
is
measured
using
a
5-point
Likert
score
for
each
item
ranging
from
0
=
never
to
4
=
very
often.,
with
a
high
total
mean
of
all
items
representing
a
high
level
of
medication-specific
support.
3.8.
Data
analysis
Data
were
checked
and
cleaned
prior
to
entry
into
SPSS
for
Windows
version
23.
Descriptive
statistics
were
used
to
analyse
data
related
to
the
patients’
baseline
characteristics
in
Australia
and
Iraq,
level
of
medication
adherence,
medication
adherence
self-efficacy,
beliefs
about
medication
and
social
support.
Opti-
mal
adherence
was
defined
as
having
a
score
of
greater
than
two
on
the
four-item
Morisky
medication
adherence
scale.
Bivariate
analyses
were
conducted
using
the
Spearman
Correlation
Coeffi-
cient
to
examine
factors
potentially
associated
with
medication
adherence.
The
levels
of
medication
adherence
were
categorised
as
high
and
low/medium
and
compared
between
the
two
patient
groups
by
Chi-square
(␹2)
test.
Variables
significantly
associated
with
medication
adherence
in
bivariate
analyses
were
examined
by
logistic
regression,
with
significance
set
at
P
<
0.25
in
the
pre-
liminary
bivariate
analysis
for
entry
into
regression
models
and

360
A.H.A.
Al-Ganmi
et
al.
/
Collegian
26
(2019)
355–365
P
<
0.05
for
the
regression
analysis
(Polit,
1996).
Two
sided
tests
were
conducted
and
for
all
analyses,
p
values
of
<0.05
were
consid-
ered
statistically
significant.
4.
Results
4.1.
Characteristics
of
participants
Recruitment
progressed
at
approximately
13
participants
per
month
in
a
single
general
hospital
in
Australia
and
20
participants
per
month
at
three
specialised
cardiac
hospitals
in
Iraq.
Between
October
2016
and
December
2017,
in
total
246
patients
with
CVD
were
recruited,
n
=
120
in
Australia,
n
=
126
in
Iraq,
as:
a)
in-patients
of
a
cardiac
ward
(n
=
179)
and
b)
out-patients
attending
cardiac
rehabilitation
in
Australia
and
out-patient
clinics
in
Iraq
(n
=
67).
The
characteristics
of
participants
from
Australia
and
Iraq
are
presented
in
Table
2.
Sociodemographic,
health
and
medication-
related
continuous
data
were
provided
by
means
and
standard
deviations,
and
categorical
data
by
frequencies
and
percentages.
Differences
between
patients
from
Australia
and
Iraq
were
tested
using
independent
samples
t-tests
for
continuous
variables
and
Chi-square
(␹2)
test
for
categorical
variables.
Overall,
Iraqi
car-
diac
participants
were
significantly
younger
(t
=
9.82,
p
=
0.001)
and
more
likely
to
be
unemployed
(chi2=
71.52,
p
=
0.017),
married
or
in
a
co-habiting
relationship
(chi2=
30.35,
p
=
0.001).
They
were
also
significantly
more
likely
to
have
a
lower
level
of
education
(chi2=
26.49,
p
=
0.001),
comorbid
disease
(chi2=
10.90,
p
=
0.001),
and
to
take
significantly
more
and
different
classes
of
cardiac
med-
ications
per
day
than
Australian
cardiac
participants
(t=
−5.71,
p
=
0.001).
4.2.
Medication
adherence
Significantly
more
participants
from
Iraq
reported
medium/low
levels
of
adherence
to
their
cardiac
medications
compared
to
par-
ticipants
from
Australia
(64.3%
versus
37.5%,
respectively)
and
fewer
reported
high
levels
of
adherence
to
their
cardiac
med-
ications
than
participants
recruited
in
Australia
(35.4%
versus
62.5%);
both
p
=
0.001
(Table
3).
Significant
associations
were
sought
between
socio-demographic
and
medication-related
vari-
ables
and
medication
adherence
(Table
4).
In
neither
country
were
the
sociodemographic
or
health
variables
significantly
associated
with
medication
adherence.
The
ARMS
included
12-items
measuring
the
ability
to
self-
administer
prescribed
medications
and
evaluating
the
patient’s
ability
to
take
medications
on
schedule.
Mean
scores
for
the
abil-
ity
to
self-administer
medications
for
the
Australian
participants
(ranging
from
23
to
40)
and
the
Iraqi
participants
(ranging
from
20
to
40)
were
38.05
(SD
=
2.69)
and
33.72
(SD
=
4.98),
respec-
tively
(Table
5).
Mean
scores
for
the
ability
to
refill
medications
on
schedule
(the
last
4
items)
for
the
Australian
and
Iraqi
par-
ticipants
(ranging
from
8
to
16)
were
14.95
(SD
=
1.26)
and
12.64
(SD
=
2.21),
respectively.
Mean
scores
for
the
ability
to
refill
med-
ication
were
highest
in
the
Australian
participants
for
the
item
“How
often
do
you
put
off
refilling
your
medicines
because
they
cost
too
much
money?”
(mean
=
3.94,
SD
=
0.23)
and
in
Iraqi
par-
ticipants
for
the
item
“How
often
do
you
decide
not
to
take
your
medicine?”
(mean
=
3.59,
SD
=
0.68).
Lowest
reported
mean
scores
for
both
Australian
and
Iraqi
participants
were
for
the
item
“How
often
do