Research and Evidence Based Practice critique

Critique quality of the literature reviews conducted for two different types of studies- a quantitative and qualitative research. see attached 

Identify and discuss the research questions, sampling and sampling sizes, research designs (qualitative vs. quantitative), hypothesis, data collection methods, and research findings.

Discuss the credibility of the sources and the research/researchers findings.

400-word minimum/550-word maximum without the references.

Minimum of 3 references  in APA format, must have been published within last 3-5 years

RESEARCH ARTICLE Open Access

“I’ve made this my lifestyle now”: a
prospective qualitative study of motivation
for lifestyle change among people with
newly diagnosed type two diabetes
mellitus
Simon J. Sebire1*, Zoi Toumpakari1, Katrina M. Turner2,3, Ashley R. Cooper1,4, Angie S. Page1,4, Alice Malpass5

and Robert C. Andrews6

Abstract

Background

: Diagnosis with Type 2 Diabetes is an opportunity for individuals to change their physical activity and
dietary behaviours. Diabetes treatment guidelines recommend theory-based, patient-centred care and advocate the
provision of support for patient motivation but the motivational experiences of people newly diagnosed with
diabetes have not been well studied. Framed in self-determination theory, this study aimed to qualitatively explore
how this patient group articulate and experience different types of motivation when attempting lifestyle change.

Methods

: A secondary analysis of semi-structured interview data collected with 30 (n female = 18, n male = 12)
adults who had been newly diagnosed with type two diabetes and were participants in the Early ACTID trial was
undertaken. Deductive directed content analysis was performed using NVivo V10 and researcher triangulation to
identify and describe patient experiences and narratives that reflected the motivation types outlined in self-
determination theory and if/how these changed over time.

Results

: The findings revealed the diversity in motivation quality both between and within individuals over
time and that patients with newly-diagnosed diabetes have multifaceted often competing motivations for
lifestyle behaviour change. Applying self-determination theory, we identified that many participants reported
relatively dominant controlled motivation to comply with lifestyle recommendations, avoid their non-compliance
being “found out” or supress guilt following lapses in behaviour change attempts. Such narratives were accompanied
by experiences of frustrating slow behaviour change progress. More autonomous motivation was expressed as
something often achieved over time and reflected goals to improve health, quality of life or family time.
Motivational internalisation was evident and some participants had integrated their behaviour change to a
new way of life which they found resilient to common barriers.

Conclusions

: Motivation for lifestyle change following diagnosis with type two diabetes is complex and can
be relatively low in self-determination. To achieve the patient empowerment aspirations of current national
health care plans, intervention developers, and clinicians would do well to consider the quality not just quantity of their
patients’ motivation.

Trial registration: ISRCTN ISRCTN92162869. Retrospectively registered

Keywords: Type 2 diabetes, Motivation, Behaviour change, Intervention, Qualitative

* Correspondence: simon.sebire@bristol.ac.uk
1Centre for Exercise, Nutrition & Health Sciences, School for Policy Studies,
University of Bristol, 8 Priory Road, Bristol BS8 1TZ, UK
Full 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 4.0
International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and
reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to
the 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.

Sebire et al. BMC Public Health (2018) 18:204
DOI 10.1186/s12889-018-5114-5

http://crossmark.crossref.org/dialog/?doi=10.1186/s12889-018-5114-5&domain=pdf

http://www.isrctn.com/ISRCTN92162869

mailto:simon.sebire@bristol.ac.uk

http://creativecommons.org/licenses/by/4.0/

http://creativecommons.org/publicdomain/zero/1.0/

Background
Patient empowerment is a cornerstone of contemporary
medicine and is central to national health care plans [1].
Individuals are increasingly encouraged, with support
from professionals, to manage their own health. For this
approach to be effective, a detailed understanding of
how patients experience regulating their behaviour when
they initiate and attempt to sustain health behaviour
change is needed.
Approximately 6% of the adult population in England

have diabetes, 90% are cases of Type 2 Diabetes Mellitus
(T2DM) and the prevalence of T2DM is rising [2]. The
burden of T2DM on individuals’ health (i.e., increased
risk of cardiovascular disease, amputation, kidney dis-
ease, retinopathy and depression) and the economy are
well documented and preventing, managing and treating
diabetes are public health priorities [2].
The point of diagnosis with T2DM is an opportunity

for clinicians to help patients initiate changes in lifestyle
behaviours such as physical activity and diet [3]. Guide-
lines for the care of adults with T2DM build on a foun-
dation of patient-centred care, and advocate the
provision of theory-based patient education at, or soon
after diagnosis to create personalised management plans,
combining advice on diet, increasing physical activity
and losing weight [3]. It is also suggested that patients
try to improve their diet and increase their physical
activity for 3 months before starting medication [4]. Re-
lated guidance (e.g., National Institute for Health and
Care Excellence) on changing lifestyle behaviours such
as physical activity suggests a range of techniques to mo-
tivate and support individual-level change including
helping patients understand the consequences of their
health-related behaviour, goal setting, and devising cop-
ing strategies to prevent relapse [5]. However, despite
this potentially complementary dual focus on patient
self-regulation and motivation, current guidelines do not
consider the degree to which patients’ motivation for
lifestyle change itself is or is not self-regulated.
Recently Fisher et al. [6] have stated that the efficacy

of diabetes-directed interventions “is often dependent
upon on how well a clinician is able to support personal
engagement and motivation of the person with diabetes
to use these new tools and knowledge consistently, and
as directed”. Self-determination theory (SDT) [7] is a
psychological framework of motivational self-regulation
that has been extensively applied to physical activity [8],
diet [9], medication adherence [10] and diabetes control
interventions [11–13]. Rather than considering only the
quantity of people’s motivation (i.e., motivated vs. not
motivated) as in previous work with patients with
T2DM [14], within SDT, the quality of motivation is
considered based on the extent to which it is self-
regulated [15].

Within SDT (Table 1), the most self-determined form
of motivation is intrinsic motivation, where behaviour is
driven by interest, enjoyment or the satisfaction that it
brings. Types of motivation that are not intrinsic but
based on tangible consequences or outcomes can vary in
their level of autonomy/self-regulation. The most
autonomous form (integrated motivation) is where
motivation is derived from an alignment of the out-
comes of a given behaviour (e.g., healthy eating) with a
person’s broader sense of self, values or goals. Less self-
determined, but still considered autonomous, is identi-
fied motivation which is based on personally important
or valued benefits of an activity (e.g., valued health or
social benefits of being active). Introjected motivation is
a form of controlled motivation where self-imposed
sanctions such as avoiding guilt or gaining contingent
self-esteem drive behaviour whereas external motivation
represents motivation based on a desire to comply with
external demands or requests, avoid punishments or to
gain rewards. Finally, amotivation represents an absence
of motivation or intention to act. The dynamic process
(although not necessarily linear) through which individ-
uals’ progress from less to more self-regulation/autono-
mous motivation is called internalisation [15].
Research amongst people with T2DM has shown that

autonomous motivation is positively associated with life-
style behaviours such as physical activity [16], dietary
self-care [17, 18], medication adherence [13], key media-
tors of behaviour change (e.g., action planning) [18] and
sustained improvement in physical health including dia-
betes control [12]. There is also some evidence that con-
trolled motivation (e.g., pressure to comply with advice
or change ones behaviour to please others or to suppress
feelings of guilt) is associated with improved dietary self-
care amongst people with newly-diagnosed T2DM [19].
Together, this evidence highlights the beneficial out-
comes associated with autonomous motivation but also
that controlled motivation may play a role in the lifestyle
changes of people with newly diagnosed diabetes. This is
not surprising given that upon diagnosis, patients com-
monly receive information (e.g., identifying previous life-
style behaviours that may have contributed to diabetes,
possible future health complications, the lifestyle change
needed to manage their symptoms, and weight and
blood glucose targets to meet); interactions that have the
potential to trigger either autonomous (e.g., identifying
personally important reasons for change) or controlled
(e.g., feeling guilty or pressured) motivation for change.
The majority of previous research has studied motiv-

ation amongst people with T2DM using quantitative
questionnaires and existing qualitative research has only
identified motivational factors such as weight manage-
ment and physical and mental well-being as motivating
physical activity among people at risk of diabetes [20].

Sebire et al. BMC Public Health (2018) 18:204 Page 2 of 10

Despite calls from researchers [21], the quality of the
motivation of people who are in early phases of initiating
behaviour change following a diagnosis of diabetes has
not been studied from the patients’ perspective. It is im-
portant to address this gap because understanding peo-
ple’s motivational experiences at critical times of
behaviour change can inform the design of patient-
centred lifestyle interventions or care.
The present study aimed to: (1) qualitatively explore

how people newly diagnosed with T2DM articulate and
experience motivation for lifestyle change as proposed in
SDT, and (2) to examine qualitative evidence for pa-
tients’ motivational internalisation over time (i.e., transi-
tion from controlled to autonomous motivation).

Methods
Study design
A secondary analysis of interview data collected with in-
dividuals who had been newly diagnosed with T2DM
and were participants in the Early ACTID (Early ACTiv-
ity In Diabetes) trial.

The early ACTID trial
Early ACTID was a lifestyle RCT, conducted between
December 2005 and September 2009, at three sites in
South West England, involving 593 adults aged between
30 and 80 years old who had received a diagnosis of
T2DM within the previous 6 months [22]. Patients were
recruited through GP practices and randomised to three
arms; usual care (UC), intensive dietary advice (ID), or
an intensive dietary advice and physical activity interven-
tion (DPAI). Usual care comprised the provision of
standard advice from a dietician at a baseline appoint-
ment, followed by two visits to a doctor blinded to treat-
ment allocation at 6 and 12 months post-randomisation.
The ID arm comprised UC plus 15 20-min appoint-
ments with a nurse or dietician for 12 months following
randomisation where patients were encouraged to
achieve a daily intake reduction of 500 Kcal and 5 to
10% loss of initial weight over 12-months. In addition to
receiving UC and the ID intervention, patients in the

DPAI arm were encouraged to increase their physical ac-
tivity to at least 30 min of brisk walking on 5 days/week.
The ID and DPAI interventions were not based on SDT
or other psychological theory, however behaviour change
techniques included; information provision, setting and
negotiating achievable physical activity and/or eating
behaviour and weight loss goals, weighing at appoint-
ments, barrier identification and assistance to overcome
them, encouragement to self-monitor weight, comple-
tion of food and or/physical activity diaries and wearing
a pedometer.

Interviewee recruitment
At their baseline appointment, each participant was
given information about the qualitative study and asked
to consent to being approached to take part in an inter-
view. Consenting individuals were purposefully sampled
to ensure interviews were held with participants in each
trial arm, with men and women of varying ages, and
from the different trial sites. No relationship was estab-
lished with participants before interview.

Data collection
The aim of the original ACTID qualitative study was to
explore how patients recently diagnosed with T2DM ex-
perience and manage their condition as well as monitor-
ing implementation and identifying improvements to the
intervention. Thirty patients (n female = 18) were inter-
viewed, comprising 6 from the UC arm (n female = 3),
12 from the ID arm (n female = 8) and 12 (n female = 7)
from the DPAI arm. Participants were aged between 40
and 72 years. Semi-structured interviews were con-
ducted (by AM) at 6-months (face-to-face interview)
and 9-months (telephone interview) post-randomisation.
The 6-month interview was timed so as not to influence
participants’ experience of the initial intervention stages.
By 9-months participants had received the majority of
intervention and sufficient time to make lifestyle
changes following diagnosis and inclusion in the trial.
The 6 month interview covered topics including re-

sponse to diagnosis, use of the information, behavioural

Table 1 Types of motivation along the Self-determination Theory continuum and diet/physical activity examples

Amotivation Extrinsic Motivation Intrinsic motivation

Controlled regulations Autonomous regulation

Non-regulation External Regulation Introjected Regulation Identified
Regulation

Integrated
Regulation

Intrinsic
Regulation

Motivation type
description

Lack of motivation
or intention to act

Lifestyle behaviour
change is to avoid
punishment or gain
a reward

Lifestyle change aims
at avoiding guilt or
enhancing self-worth

Lifestyle changes
are personally
important or
valued

Lifestyle behaviours
are in harmony with
other personal values
and goals

Lifestyle behaviours
are enjoyable
or inherently
satisfying to do

Diet / physical
activity example

Not changing one’s
lifestyle behaviours
or passively going
through the motions

Eating less
confectionary to
avoid being told
off by a dietician

Exercising because
one feels they should,
and will feel guilty if
one doesn’t

Maintaining one’s
physical fitness is
a personally
important goal

Eating a healthily is
consistent with one’s
goals to be physically
active

Trying out new
healthy recipes is
satisfying and fun

Sebire et al. BMC Public Health (2018) 18:204 Page 3 of 10

changes and relationship with the clinical team [23]. The
9 month interview covered topics including diet and/or
exercise changes, maintaining change, barriers, coping
with reduce appointment frequency and the trial ending.
Neither the interview guide nor the interviews them-
selves were based on SDT, although the topic of motiv-
ation for change was discussed in depth. Within the
original data collection team, KM and AM discussed
themes and data saturation pertaining to the research
questions. At the time, this was felt to have been
reached. The 6 month (n = 30) and 9 month (n = 29,
one participant did not participate in a 9 month inter-
view) interviews were, on average, 90 and 15 min in dur-
ation respectively. Interview audio recordings were
transcribed verbatim. Ethical approval was given by the
Bath Research Ethics Committee (05/Q2001/5).

Data analysis
The analysis combined both supra (i.e., by answering a
new theoretical question) and supplementary (i.e., a
more in depth investigation of motivation themes which
were not addressed fully in the primary study) secondary
qualitative analysis [24]. Initial analysis involving three
of the authors indicated that there were sufficient ac-
counts to explore the quality of participants’ motivation.
All available data were analysed using directed content
analysis which is appropriate where an existing theory
can guide research questions and initial theory-based
themes and can be supported, challenged or extended
[25]. Analysis was primarily deductive and sought to
identify patient experiences and narratives that reflected
the motivation types outlined in Table 1. Further, a
within-participant analysis probed for evidence of motiv-
ational internalisation between the 6 and 9 month inter-
views. Alongside the deductive analysis, coding was
flexible to allow participants’ narratives and personal
context to guide the complexity of themes and new
themes to emerge.
Transcripts were loaded to NVivo software (QSR

International Pty Ltd. Version 10, 2012) to enable coding
of the data and to facilitate the organisation of codes
into themes and subthemes. Coding was undertaken by
two researchers (ZT & SJS) who had experience of
studying SDT. Both began by coding the same three
transcripts, discussing their initial interpretations/codes
and then refining the initial coding frame (i.e., new codes
being added and existing codes being deleted/refined).
The coding frame was then applied to another five
transcripts, further refined and agreed and then applied
to the remaining 52 transcripts which were divided
equally between ZT and SJS. Researcher triangulation
was undertaken via independent coding of transcripts
and regular meetings to discuss code refinements and

emergent themes. The researchers agreed that the
quotes presented were representative of the themes.

Results
The results are presented as six themes reflecting the
motivation types in SDT (See Table 1) with narratives
reflecting internalisation (or the lack of) integrated
alongside each motivation type. Participant gender,
trial arm allocation (ID: Intensive Dietary Advice,
DPAI: Intensive Dietary Advice & Physical Activity, &
UC: Usual Care) and interview time point are shown
for each quote.

Amotivation
Some participants were reluctant to change and articu-
lated a passivity towards any changes they reported. Ig-
noring one’s diabetes, feeling helpless, not able to
change or resigned to one’s current way of life and not
believing the health benefits of recommended treatments
were also common:

… the other tablets I take is all in the morning, this
one is at night, then one night I forget, then the next
night I forget and I can’t be bothered taking these
because I don’t think – I can’t believe tablets would
do [me any good] – you know? (Mo, male, DPAI, 6
months)

Some participants felt helpless: “At 66 I’m hardly …
going to change very much” (Ronny, male, ID, 6 months)
and others reported acting against advice that was not in
line with a “good life”, deceiving their health practi-
tioners to appear compliant:

Well they don’t know I’m not doing anything, I’m
living a good life and being a good boy. [laughs] I don’t
tell ‘em, I’m back on scones, they just say “how’s your
diet?” And I say “okay”. (James, male, UC, 6 months)

Three months later, James had not changed his motiv-
ation or diet:

Well I never hardly made any [diet changes], only cut
out me scones and Rich Tea biscuits, that’s all I’ve
done, but that’s gone by the board [been abandoned]
the last few months, mind. (James, male, UC, 9 months)

External motivation
The majority of the participants referred to their motiv-
ation being controlled by external sources when first
changing behaviour post-diagnosis. Participants felt re-
stricted in what they were “allowed to eat” (Frank, male,
UC, 6 months) or the exercise that they “have to” do

Sebire et al. BMC Public Health (2018) 18:204 Page 4 of 10

(Nina, female, DPAI, 6 months), found the changes
unenjoyable, and a threat to their quality of life. Ex-
ternal motivation was commonly interpreted from
narratives about patient compliance with the Early
ACTID health practitioners’ recommendations or
goals to avoid negative outcomes (such as increasing
medication).

I can use those [books] for reference for restricting
carbohydrate intake, which I’m trying to do at the
moment. It’s the latest aim of my – or goal of my
dietician. And that way I can hopefully reduce these
glucose levels a bit more … I don’t want to get a third
tablet a day, that’s what I’m trying to avoid. (Stuart,
Male, ID, 6 months)

Some participants’ behaviours were motivated
by wanting to demonstrate that they had been
“behaving” between appointments, completing be-
haviour change activities such as diet diaries, and
a fear that non-compliance would be identified by
measurements:

Knowing that I’m going to see [practitioner] every so
often … part of me says “well you’ve got to behave
yourself girl because you’re going to be going and
seeing them, and they’ll know if you haven’t been
behaving yourself, your weight and your levels and
everything else” (Wendy, Female, ID, 9 months)

Narratives rich in external motivation coincided with
reports of challenges to behaviour change and disap-
pointment when effort was not rewarded with the de-
sired outcomes such as weight loss:

So we’re there cutting, cutting, cutting, and doing
everything they say and the weight doesn’t come off, ‘oh
you’ve got to get your output up’, so my output is now
up, the weight is not coming off, you know? (Hugh,
Male, DPAI, 6 months)

The majority of participants who articulated exter-
nal motivation at 6 months, appeared to remain con-
trolled by external sources 3 months later. During
their second interview their low enjoyment of lifestyle
changes persisted, and they reported challenges, lapses
and slow progress towards change that they did not
feel in control of:

Things have just gone on like a slow train without any
stations. (Rose, Female, DPAI, 9 months)

Participants (in all trial arms) were commonly con-
cerned about losing the professional support when the

intervention ended that provided them with motivation
they needed to sustain change.

Through not seeing [practitioner] so much,
no one is on top of me, keeping … I am like
a little boy really, I suppose, I need somebody
to give me a kick up the backside … and I
know it’s coming to an end now as well, so I’m
getting a little bit lax I think. (Mo, male, DPAI,
9 months)

Introjected motivation
Motivation based on personal pressures such as
avoiding guilt were commonly rooted in partially-
internalised lifestyle advice. Participants’ understood
the reasons for lifestyle change but appeared relatively
controlled by these reasons rather than describing
them as personally valued. Participants commonly re-
ported introjected exercise motivation to burn calories
consumed during dietary lapses:

Admittedly if I’ve eaten something that I know I
shouldn’t have and I haven’t yet exercised, I’m more
determined to go and exercise then, because … I’ve got
to get rid of that, I’ve got to burn it off. (Wendy,
Female, ID, 9 months)

Similarly, introjected motivation was common where
participants had hung their self-worth on successful be-
haviour change but not achieved it. They judged their
behaviour as “wrong” versus right and themselves as
“naughty”, “stupid” or at “fault”.

I’m going to be good and go back on my diet, it’s
entirely my fault because I know the rights and the
wrongs and the dos and the don’t dos … it was also
summer and barbeques and I just think I’d been very
naughty and stupid and allowed this to happen, but
I do see the dietician on Thursday, so it is entirely my
own fault. (Diana, female, DPAI, 9 months)

Showing the interplay between external and intro-
jected motivation over time, at 6 months Clive (Male,
DPAI) was motivated by the reward of preventing a
worsening of his diabetes but viewed certain dietary
changes as punishment:

I’m very motivated by rewards if you like, and I
consider the reward for not doing that is my
diabetes not getting any worse, and if I can hold it
at that, then I’ll punish myself with other things,
such as giving up sugars and sweets (Clive, Male,
DPAI, 6 months)

Sebire et al. BMC Public Health (2018) 18:204 Page 5 of 10

Three months later, Clive’s diabetes had worsened, he
had been prescribed medication to improve his blood
sugar control and felt guilty for not having made
changes sooner. He now felt internal pressure to adopt
the lifestyle advice “more seriously”.

I was cursing myself for not giving it [unhealthy
food] all up and trying to keep on the diet rather
than have to go onto tablets … I’ve still left myself
subject to eat some sweet things a week, and had
I known that it would take me out of the diet
side onto tablets I would have made even more
effort I think … So yeah, I really must settle down
with these tablets – and take it a little bit more
seriously than I have done. (Clive, Male, DPAI, 9
months)

Identified motivation
At diagnosis, all participants had been informed of the
potential health implications of diabetes and the benefits
of lifestyle changes for glucose control. It was therefore
unsurprising that many referred to valuing the experi-
enced or anticipated health benefits of diet or physical
activity changes including feeling fitter, healthier, im-
proved quality of life and well-being. Others identified
with maintaining their health in order to spend time
with family and/or fulfil caring responsibilities:

(I) don’t want to go blind, I don’t want my legs chopped
off and I’d like to live a bit longer. [laughs] And I’ve got
a family to look after. It’s all that sort of stuff. So I want
to stay as healthy as I can for as long as I can,
preferably with as little medication as I can. (Pete,
Male, DPAI, 6 months)

Like Pete, many participants reported identified motiv-
ation based on the importance they placed on avoiding
deleterious health consequences of diabetes. While this
form of motivation was initially an external motivation,
and accompanied for some by fear of the consequences
of not changing, many participants had internalised the
health threat to become a personally valued outcome of
change. Commonly, these motivations were based on
participants wanting to avoid the ill health experienced
by their relatives or friends:

Knowing how my mother went with her diabetes and
the fact that before she died she could only just about
see, I thought “well I don’t want that happening to
me”. (Monica, female, UC, 6 months)

Further, the personal importance placed on avoiding ill
health was accompanied by participants taking responsibility

for change, following advice and wanting to avoid future
feelings of shame associated with not changing:

I don’t want it [heart problems] to happen to me
because I’m not following recommended lines. If it
happened to me for any other reason then that’s
something that I probably can’t control, but if there
is something that I can contribute to stop it happening
then I will do so. (Robert, male, UC, 6 months)

For some participants, their diagnosis with diabetes
provided “a stronger motivation that is bigger than the
other motivations for not doing it” (Alice, Female, ID,
6 months). This was in contrast to their previous, less
effective appearance-based motivations for dieting:

That’s what I was like with my diets, I [would] not-
stick to them, because – well you don’t think there’s
any health reason why you should stick to this thing
apart from vanity with slimming. But as soon as
you’re told you’re diabetic, it’s in your best interests to
get that weight off and stick to a healthy diet. So I
suppose in a roundabout way that complication helps
you another way, and then I feel much better now for
that. (Mary, female, DPAI, 6 months)

Participants carried their identified motivation from 6
to 9 months but also expressed their need for continued
collaborative professional or social support to help them
monitor their diabetes and/or support their motivation:

I am frightened of it [Early ACTID] ending. I
would like to be monitored by experts, but I
don’t think they are experts at my surgery … if
I was monitored regularly like I am now, I would
know that it was the diabetes that I allowed to
get out of control and I’ll do something about it.
Without knowing I’m very worried. (Frank, Male,
UC, 9 months)

Integrated motivation
Some participants described how over time their lifestyle
behaviours had become motivated by forming a new pat-
tern, routine, or a way of life. Many of these statements
were articulated at the 6 month interviews suggesting
that some participants had internalised the motivation
for lifestyle change relatively quickly after diagnosis:

I think I’ve entered now into a pattern and a way of
life (Jill, female, DPAI, 6 months).

Many narratives like this reflected an internalisation
process whereby participants’ motivation was initially

Sebire et al. BMC Public Health (2018) 18:204 Page 6 of 10

controlled but over time aligned lifestyle recommenda-
tions with their perception of a good quality of life:

Obviously it’s becoming a way of life now, what I’m
doing, it’s becoming a way of life, and I don’t think
about it so much. The first few months I did, I thought
“I mustn’t do this, I mustn’t do that”, and [practitioner]
said “I think you’re being a bit hard on yourself here,
you’re being a bit too strict and you won’t keep it up”.
But now I’m settled into a very comfortable way of life
with it. (Nora, female, ID, 6 months)

References

to behavioural habits, viewing exercise or
healthy eating as a personal characteristic, and flexible
resilience to challenges such as dietary lapses or bad
weather were common alongside integrated motivation:

I’ve made this my lifestyle now, my exercise, it’s just
such a routine now that I get up, I have my breakfast
and I go out, and I just do it really. If it’s raining I go,
if it’s absolutely lashing I might wait and say “I’ll go
this afternoon”, but very rarely do I miss. (Penny,
Female, ID, 9 months)

Personal characteristics that supported internalisation
of lifestyle changes included listening to advice from
practitioners, taking responsibility for change, being
positive, and persistent. Participants also referred to
making gradual changes, perceiving that the time was
right for change and balancing their new and more
established lifestyle components (e.g., social life)

Systematically [it] becomes part of your life … I try in
life not to make a hard life any harder, just try and
make it easy by getting to grips with this, listening to
what you’re told and being positive about this. (Nora,
female, ID, 6 months)

Intrinsic motivation
Several participants reported enjoying their new way of
eating or the relaxation and satisfaction brought about
by exercise:

The exercise now has become so much a part of my life
that I don’t think about it as the fact that I must do it as a
chore, I do it because I enjoy it, I go out, the hour walking
sorts the brain out for the day, its relaxing, sometimes I’m
out and I’m back before I’ve gone, before I realise [laughs].
Yes, I enjoy it. (Penny, Female, ID, 6 months)

Intrinsic motivation was sometimes supported by
identified and introjected motivations at the moment of
deciding to exercise:

Some days I have to make myself do it, but I always
feel better when I come back. I always feel I’ve got
more energy, and more get up and go when I’ve been.
It sounds daft don’t it- it’s almost like somebody strikes
match and you do it and then you come back. (Sylvia,
female, DPAI, 6 months)

Robert demonstrated internalisation where his motiv-
ation transitioned from eating to follow recommenda-
tions to a combination of integrated (his eating “style
has changed”) and intrinsic motivation (he now enjoys
different foods). He thought that it was important to be
motivated by enjoyment rather than pressure.

When you see somebody alongside you with a
meal that you like but you can’t have, I don’t find
it envious anymore that they are eating something
I would like to have eaten, my style has changed,
I now enjoy a different type of food, but the
important thing is I enjoy it, it’s not simply because
I know that that’s what I’ve got to eat. (Robert,
male, UC, 6 months)

Discussion

This study is the first qualitative examination of the
types of motivation for lifestyle change articulated in
SDT amongst people newly diagnosed with T2DM. As
the participants were involved in a lifestyle intervention,
it could be argued that almost all had some motivation
to change, but despite this, the findings highlight the di-
versity in motivation quality both between and within
participants. The prospective data facilitated an analysis
of motivation transitions.
Diagnosis with T2DM provokes a range of emotional

responses [26], close scrutiny of patients’ lifestyle, threats
to people’s social and personal identity and the need to
construct a new identity representations [27]. It is not
surprising therefore that many participants’ motivation
for change was controlled, or not self-regulated. External
motivation for diet or physical activity change was expe-
rienced as participants complying with what they per-
ceived to be restrictive dietary advice and through fear
of non-compliance (i.e., “lapses”) being identified in ap-
pointments or assessments. On the SDT continuum
(Table 1) external and introjected motivation are located
adjacently and in the theory, motivation is viewed as dy-
namic rather than static [15]. This was supported in our
findings as participants often experienced these motiva-
tions concurrently, by complying with recommendations
and labelling themselves as “good” or “naughty” and
their behaviour as “right” or “wrong” based on the extent
to which their behaviour change was successful. Consist-
ent with previous research with exercisers pursuing

Sebire et al. BMC Public Health (2018) 18:204 Page 7 of 10

extrinsic (relatively controlling) goals [28], participants
whose motivation was relatively controlled experienced
frustrating slow progress towards rigidly defined end
point goals (e.g., weight loss). These experiences of con-
trolled motivation amongst people with newly diagnosed
T2DM are a source of internal conflict and potential
barriers to their development of self-regulation.
Previous work has shown that autonomous motivation

is associated with physical activity and healthy eating [8,
9]. Identified motivation (i.e., personally important valu-
ing of a behaviour) mainly stemmed from the value par-
ticipants placed on health, quality of life and family
responsibilities which they understood to be compro-
mised by uncontrolled diabetes. Improved health is an
intrinsic goal [29] which, relative to extrinsic goals, such
as improved appearance, is associated with autonomous
motivation and physical activity behaviour [30, 31]. The
results add experiential support to this finding as for
some participants health-based reasons for change,
prompted by their T2DM diagnosis, were more motivat-
ing than their previous extrinsic appearance-based
weight loss goals. Diabetes diagnosis may offer an oppor-
tunity to help individuals identify meaningful intrinsic
goals (e.g., health or family time) which will likely under-
pin autonomous motivation.
Despite the participants being relatively newly diagnosed

with T2DM some reported integrated motivation (i.e., phys-
ical activity or healthy eating being part of their identity),
which plays an important role in motivating diet [9] and
physical activity [32]. Having internalised early controlled
motivations (i.e., moved from controlled to autonomous
motivation), participants’ new lifestyle had become a pat-
tern or a way of life which was robust to challenges. Inte-
grated motivation developed over time and internalisation
was supported by personal factors such as a positive atti-
tude, resilience to barriers (e.g., bad weather), persistence
and practitioners who encouraged gradual change.
Intrinsic motivation (i.e., being motivated by enjoy-

ment, interest and satisfaction) was articulated least fre-
quently, although some participants enjoyed their
exercise and diet changes and this was commonly sup-
ported by integrated motivation. It is to be expected that
new physical activity and eating behaviours may not yet
be intrinsically motivated in a sample such as ours, and
it is possible that for some patients, or for some behav-
iours (e.g., cutting down on high sugar foods) which pa-
tients find enjoyable, identified motivation for change
(i.e., identifying a health-based value) may be a more
realistic and adequate motivational target. Indeed it is
suggested that maintenance of lifestyle behaviours, such
as exercise, is most likely when a person has a combin-
ation of intrinsic, identified and integrated motivation
types [33]. Collectively, the findings suggest that if
T2DM patients can be supported to internalise their

motivation to the point of identifying a personal benefit,
or integrate changes as part of an enjoyable way of life,
such changes may be more sustainable and resilient to
common challenges to behaviour change (e.g., lack of
time, periods of holiday, & changes in routine).
Recent quantitative research using SDT has sought to

identify how different types of motivation for physical
activity commonly cluster within individuals [21, 32, 34].
Amongst adults with T2DM, Gourlan et al. [21] identi-
fied a “self-determined” profile (high scores on autono-
mous motivation types), a “moderate” profile (all
motivation types moderately endorsed), and a “high
combined” profile (all motivation types strongly en-
dorsed plus moderate amotivation). Our findings add
further experiential evidence to support the existence of
these multifaceted motivation profiles which commonly
include both autonomous and controlled motivation.
For example, amongst patients who reported largely
identified and intrinsic motivation, low-level controlled
motivation (often introjected) supported their mainten-
ance of behaviour change at times. Together, the find-
ings support calls for future research to take a theory-
driven person- rather than variable-centred approach to
understanding motivation [35] and indicate that mixed-
methods approaches may be particularly illuminating.
Largely regardless of their dominant motivation, par-

ticipants articulated a need for structure in their care,
commonly through provision of expert guidance and
support. However, the nature of the structure sought dif-
fered depending on participants’ motivation. Specifically,
participants mainly motivated by autonomous reasons
sought support for their ongoing self-regulation (with
particular interest in ongoing assessment of health out-
comes), whereas participants mainly motivated by con-
trolled reasons sought more continuous provision of
motivation (i.e., being pushed or prompted) by a practi-
tioner, family member or friend with references to pater-
nalistic perspectives (e.g., “like a little boy”). The
provision of structure is a cornerstone of autonomy-
supportive clinical/interpersonal interactions, which aim
to facilitate patients’ autonomous motivation and com-
petence [36]. This study highlights the importance of
considering the long term provision of support/structure
for people newly diagnosed with T2DM for two reasons;
first, the transition from controlled to autonomous mo-
tivation can take time, and ongoing continuity in expert
support can create a space to facilitate patients’ internal-
isation, and second, it is clear that even participants who
were relatively self-regulated did not want to be left on
their own, rather they wanted professional support to
“keep on track”. Our findings therefore support the dis-
tinction drawn in SDT between the provision of
autonomy-support (i.e., support for self-regulation) and
independence (i.e., being left to fend for oneself).

Sebire et al. BMC Public Health (2018) 18:204 Page 8 of 10

The findings of this study suggest that to achieve the
patient empowerment aspirations of current national
health care plans [1], clinicians would do well to con-
sider the quality not just quantity of their patients’ mo-
tivation. Research suggests that physicians may not
know whether their T2DM patients are motivated to
change or not and recommend the regular measurement
of patient motivation [14]. While our findings support a
greater focus on patient motivation, we would argue that
considering the quality of motivation is of primary im-
portance. Our findings complement a recent framework
for supporting engagement and motivation for behaviour
change in people with diabetes which draws on multiple
patient-centred approaches including SDT [6]. This
framework provides clinicians with a pragmatic, three-
step approach to building a supportive clinician-patient
relationship. Our findings support this work by document-
ing patient motivational experiences in line with the
framework’s underpinning theory that clinicians will likely
experience in conversations with patients about behaviour
change. Previous work has identified how concepts from
SDT could be integrated into medical training [37] which
would help clinicians become attuned to patients’ motiv-
ation quality and support patients’ motivational needs.

Strengths and limitations
The qualitative data provided a rich person-centred re-
source with which we were able to extend previous
variable-centred quantitative literature. The sample size
was relatively large and the initial interviews were exten-
sive. The repeated interviews helped us hear personal
experiences of the dynamic nature of behaviour change
and motivation. Despite these strengths, the follow-up
interviews were shorter (although the transcripts sug-
gested that discussions were detailed) and although there
can be strength in not basing interviews on theory, more
theory-driven follow up interviews would have allowed a
more in-depth analysis of motivation change. Finally,
while we have reported our secondary analysis methods
transparently and used researcher triangulation to agree
our interpretations, due to the lapse between data collec-
tion and the analysis, it was not possible to use other
strategies, such as member checking.

Conclusions
Understanding T2DM patients’ motivation for lifestyle be-
haviour change is considered central to successful patient-
centred care [6]. Combining qualitative methods with
SDT, we have identified the diverse motivational experi-
ences of people newly diagnosed with T2DM. Participants
who reported relatively dominant controlled motivation
experienced initial behaviour change but this was often
accompanied by internal conflict, frustration and a need
for continual external prompting. Participants’ reporting

more autonomous motivation approached behaviour
change with more flexibility had integrated it in to a new
way of life and wanted ongoing support for their self-
regulation. The findings highlight the importance of un-
derstanding the quality of motivation in this group and
carefully considering the types of motivation that are tar-
geted in lifestyle interventions for people with T2DM.

Abbreviations

DPAI: Dietary advice and physical activity; HbA1c: Glycated haemoglobin
(A1c); ID: Intensive dietary advice; SDT: Self-determination theory;
T2DM: Type 2 diabetes mellitus; UC: Usual care

Acknowledgements
Not applicable.

Ethical approval and consent to participate

Ethical approval for the Early-ACTID study was given by the Bath Research
Ethics Committee (05/Q2001/5) and all participants provided consent to the
Early-ACTID study and additionally to be interviewed.

Funding

This study was supported by the NIHR Biomedical Research Centre at the
University Hospitals Bristol NHS Foundation Trust and the University of
Bristol. The views expressed in this publication are those of the author (s)
and not necessarily those of the NHS, the National Institute for Health
Research or the Department of Health. The Early ACTID study was funded
by Diabetes UK and the UK Department of Health. The funders were not
involved in the design of the study, the collection, analysis or interpretation
of the data nor the writing of the manuscript.

Availability of data and materials

The datasets generated and/or analysed during the current study are not
publicly available due to the level of personal information that is contained
in the qualitative transcripts.

Authors’ contributions

SJS, ASP and ARC conceived of the study and sought funding to undertake
the secondary data analysis. KMT and AM designed and conducted the
interviews. SJS, ZT & KMT designed the analytical approach, SS & ZT undertook
the secondary data analysis and all authors provided critical input into the
interpretation of the findings. SS wrote the first draft of the paper and
coordinated contributions from the co-authors. All authors made critical
comments on the drafts of the paper and approved the final submission.

Consent for publication

Not applicable.

Competing interests

The authors declare that they have no competing interests.

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published
maps and institutional affiliations.

Author details

1Centre for Exercise, Nutrition & Health Sciences, School for Policy Studies,
University of Bristol, 8 Priory Road, Bristol BS8 1TZ, UK. 2Department of
Population Health Sciences, Bristol Medical School, University of Bristol,
Bristol, UK. 3The National Institute for Health Research Collaboration for
Leadership in Applied Health Research and Care West (NIHR CLAHRC West)
at University Hospitals Bristol NHS Foundation Trust, Bristol, UK. 4National
Institute for Health Research Bristol Biomedical Research Centre, University
Hospitals Bristol NHS Foundation Trust and University of Bristol, Bristol, UK.
5Centre for Academic Primary Care, Bristol Medical School, University of
Bristol, Bristol, UK. 6Institute of Biomedical and Clinical Sciences, University of
Exeter Medical School, Medical Research, RILD Level 3, Barrack Road, Exeter,
Devon EX2 5DW, UK.

Sebire et al. BMC Public Health (2018) 18:204 Page 9 of 10

Received: 11 May 2017 Accepted: 23 January 2018

References
1. NHS England. NHS five year forward view. London; 2014.
2. Gatineau M, et al. Adult obesity and type 2 diabetes. Public Health England;

2014.
3. NICE. Type 2 diabetes in adults: management. 2015.
4. Inzucchi SE, et al. Management of hyperglycemia in type 2 diabetes: a

patient-centered approach: position statement of the American Diabetes
Association (ADA) and the European Association for the Study of Diabetes
(EASD). Diabetes Care. 2012;35(6):1364–79.

5. NICE, Behaviour change: general approaches. 2007.
6. Fisher L, et al. A practical framework for encouraging and supporting

positive behaviour change in diabetes. Diabet Med. 2017;34(12):1658–66.
7. Deci EL, Ryan RM. The “what” and “why” of goal pursuits: human needs and

the self-determination of behavior. Psychol Inq. 2000;11:227–68.
8. Teixeira PJ, et al. Exercise, physical activity, and self-determination theory: a

systematic review. Int J Behav Nutr Phys Act. 2012;9:78.
9. Teixeira PJ, Patrick H, Mata J. Why we eat what we eat: the role of autonomous

motivation in eating behaviour regulation. Nutr Bull. 2011;36(1):102–7.
10. Williams GC, et al. Autonomous regulation and long-term medication

adherence in adult outpatients. Health Psychol. 1998;17(3):269–76.
11. Moreau M, Gagnon MP, Boudreau F. Development of a fully automated,

web-based, tailored intervention promoting regular physical activity among
insufficiently active adults with type 2 diabetes: integrating the I-change
model, self-determination theory, and motivational interviewing
components. JMIR Res Protoc. 2015;4(1):e25.

12. Williams GC, et al. Testing a self-determination theory process model for
promoting glycemic control through diabetes self-management. Health
Psychol. 2004;23(1):58–66.

13. Williams GC, et al. Reducing the health risks of diabetes: how self-
determination theory may help improve medication adherence and quality
of life. Diabetes Educ. 2009;35(3):484–92.

14. Linmans JJ, Knottnerus JA, Spigt M. How motivated are patients with type 2
diabetes to change their lifestyle? A survey among patients and healthcare
professionals. Prim Care Diabetes. 2015;9(6):439–45.

15. Ryan RM, Deci EL. Self-determination theory and the facilitation of intrinsic
motivation, social development, and well-being. Am Psychol. 2000;55:68–78.

16. Sweet SN, et al. Understanding physical activity in adults with type 2 diabetes
after completing an exercise intervention trial: a mediation model of self-
efficacy and autonomous motivation. Psychol Health Med. 2009;14(4):419–29.

17. Senecal C, Nouwen A, White D. Motivation and dietary self-care in adults
with diabetes: are self-efficacy and autonomous self-regulation
complementary or competing constructs? Health Psychol. 2000;19(5):452–7.

18. Julien E, Senecal C, Guay F. Longitudinal relations among perceived
autonomy support from health care practitioners, motivation, coping
strategies and dietary compliance in a sample of adults with type 2
diabetes. J Health Psychol. 2009;14(3):457–70.

19. Nouwen A, et al. Longitudinal motivational predictors of dietary self-care
and diabetes control in adults with newly diagnosed type 2 diabetes
mellitus. Health Psychol. 2011;30(6):771–9.

20. Korkiakangas EE, et al. Motivators and barriers to exercise among adults
with a high risk of type 2 diabetes–a qualitative study. Scand J Caring Sci.
2011;25(1):62–9.

21. Gourlan M, Trouilloud D, Boiche J. Motivational profiles for physical activity
practice in adults with type 2 diabetes: a self-determination theory
perspective. Behav Med. 2016;42(4):227–37. https://doi.org/10.1080/
08964289.2014.1001810.

22. Andrews RC, et al. Diet or diet plus physical activity versus usual care in
patients with newly diagnosed type 2 diabetes: the early ACTID randomised
controlled trial. Lancet. 2011;378(9786):129–39.

23. Malpass A, Andrews R, Turner KM. Patients with type 2 diabetes experiences
of making multiple lifestyle changes: a qualitative study. Patient Educ
Couns. 2009;74(2):258–63.

24. Heaton, J., Secondary analysis of qualitative data: a review of the literature.
2000, Social Policy Research Unit, University of York.

25. Hsieh HF, Shannon SE. Three approaches to qualitative content analysis.
Qual Health Res. 2005;15(9):1277–88.

26. Peel E, et al. Diagnosis of type 2 diabetes: a qualitative analysis of patients’
emotional reactions and views about information provision. Patient Educ
Couns. 2004;53(3):269–75.

27. Amorim MM, et al. Identity representations of people with diabetes. Qual
Health Res. 2014;24(7):913–22.

28. Sebire SJ, et al. “coveting thy neighbour’s legs”: a qualitative study of
exercisers’ experiences of intrinsic and extrinsic goal pursuit. J Sport Exerc
Psychol. 2013;35(3):308–21.

29. Sebire SJ, Standage M, Vansteenkiste M. Development and validation of the goal
content for exercise questionnaire. J Sport Exerc Psychol. 2008;30(4):353–77.

30. Sebire SJ, Standage M, Vansteenkiste M. Examining intrinsic versus extrinsic
exercise goals: cognitive, affective, and behavioral outcomes. J Sport Exerc
Psychol. 2009;31(2):189–210.

31. Gunnell KE, et al. Goal contents, motivation, psychological need satisfaction,
well-being and physical activity: a test of self-determination theory over 6
months. Psychol Sport Exerc. 2014;15(1):19–29.

32. Miquelon P, Chamberland P, Castonguay A. The contribution of integrated
regulation to adults’ motivational profiles for physical activity: a self-
determination theory perspective. Int J Sport Exerc Psychol. 2017;15(5):488–
507. https://doi.org/10.1080/1612197X.2016.1155637.

33. Ryan RM, Deci EL. Active human nature: self-determination theory and the
promotion and maintenance of sport, exercise and health. In: Hagger MS,
Chatzisarantis NLD, editors. Intrinsic motivation and self-determination in
exercise and sport. Champaign: Human Kinetics; 2007. p. 1–19.

34. Lindwall M, et al. Stirring the motivational soup: within-person latent
profiles of motivation in exercise. Int J Behav Nutr Phys Act. 2017;14(1):4.

35. Vansteenkiste M, Mouratidis A. Emerging trends and future directions for
the field of motivation psychology: a special issue in honor of Prof. Dr. Willy
lens. Psychologica Belgica. 2016;56(3):317–41.

36. Williams GC, et al. Variation in perceived competence, glycemic control, and
patient satisfaction: relationship to autonomy support from physicians.
Patient Educ Couns. 2005;57(1):39–45.

37. Ten Cate TJ, Kusurkar RA, Williams GC. How self-determination theory can
assist our understanding of the teaching and learning processes in medical
education. AMEE guide no. 59. Med Teach. 2011;33(12):961–73.

• We accept pre-submission inquiries
• Our selector tool helps you to find the most relevant journal
• We provide round the clock customer support
• Convenient online submission
• Thorough peer review
• Inclusion in PubMed and all major indexing services
• Maximum visibility for your research

Submit your manuscript at
www.biomedcentral.com/submit

Submit your next manuscript to BioMed Central
and we will help you at every step:

Sebire et al. BMC Public Health (2018) 18:204 Page 10 of 10

http://dx.doi.org/10.1080/08964289.2014.1001810

http://dx.doi.org/10.1080/08964289.2014.1001810

http://dx.doi.org/10.1080/1612197X.2016.1155637

Abstract
Background
Methods
Results
Conclusions
Trial registration
Background
Methods
Study design
The early ACTID trial
Interviewee recruitment
Data collection
Data analysis
Results
Amotivation
External motivation
Introjected motivation
Identified motivation
Integrated motivation
Intrinsic motivation
Discussion
Strengths and limitations
Conclusions
Abbreviations
Ethical approval and consent to participate
Funding
Availability of data and materials
Authors’ contributions
Consent for publication
Competing interests
Publisher’s Note
Author details
References

Oncotarget14516www.impactjournals.com/oncotarget

www.impactjournals.com/oncotarget/ Oncotarget, 2017, Vol. 8, (No. 9), pp: 14516-14524

Impact of diabetes on the risk of bedsore in patients undergoing
surgery: an updated quantitative analysis of cohort studies

Mining Liang1,*, Qiongni Chen2,*, Yang Zhang3, Li He1, Jianjian Wang1, Yiwen Cai1,
Lezhi Li2
1Department of Psychiatry, The Second Xiangya Hospital, Central South University, Changsha, Hunan Province, China
2Department of Nursing, The Second Xiangya Hospital, Central South University, Hunan Province, Changsha, China
3Nursing Teaching and Research Institute, Medical College of Guangxi University of Science and Technology, Liuzhou, Guangxi
Province, China

*These authors contributed equally to this work

Correspondence to: Lezhi Li, email: llz6511@126.com

Keywords: diabetes, bedsore, surgery, meta-analysis

Received: November 26, 2016 Accepted: December 21, 2016 Published: December 27, 2016

ABSTRACT
Diabetes is a major cause of morbidity for patients undergoing surgery and can

increase the incidence of some postoperative complications such as bedsores. We
conducted a meta-analysis of observational studies to examine whether patients with
diabetes undergoing surgery had high risk of bedsore. We performed a systematic
literature search in Pubmed, Embase and the Cochrane Library Central Register of
Controlled Trials database from inception to November 2016. Studies were selected
if they reported estimates of the relative risk (RR) for bedsore risk in postoperative
diabetic patients compared with that of in non-diabetic patients. Random-effects
meta-analysis was conducted to pool the estimates. A total of 16 studies with 24,112
individuals were included in our meta-analysis. The pooled RR of bedsore development
for patients with diatetes was 1.77 (95% CI 1.45 to 2.16). The results of subgroup
analyses were consistent when stratified by surgery type, study design, research
region, sample size, inclusion period, analysis method and study quality. There was
evidence of publication bias among studies and a sensitivity analysis using the Duval
and Tweedie “trim-and-fill” method did not significantly alter the pooled results
(adjusted RR 1.17, 95% CI 1.02 to 1.36).This meta-analysis provides indications
that diabetic patients undergoing surgery could have a higher risk of developing
bedsores. Further large-scale prospective trials should be implemented to comfirm
the association.

INTRODUCTION

Bedsore, also known as pressure ulcer, is a common
cause of prolonging length of hospital stay for patients
with surgery. It has been reported that the length of
hospital stay of surgical patients could increase by 3.5 to
5 days on average when a bedsore occurs [1, 2]. For some
severe cases, the length of stay for bedsores could even be
longer than 15 days [3], which adds tremendous financial
burden on the patient and healthcare facility.

Several risk factors and aetiologies have been
reported to contribute to the development of bedsores

during perioperative period. Traditionally, it is considered
that patients with advanced age, malnutrition (lower levels
of hematocrit or albumin), poor circulation or smoking
may have a higher risk of bedsores [4–7]. Moreover,
for patients with surgery, some other factors such as
anesthesia and surgery type, length of surgery, patient
position during the surgery, warming or moisture devices
used,and padding type the patients used [8–11] could also
affect the development of bedsores.

Numerous studies have explored the role of patients
with preexisting diabetes on the development of bedsore.
Despite the fact that some studies have reported significant

Research Paper

Oncotarget14517www.impactjournals.com/oncotarget

association between diabetes and risk of surgery-related
bedsore, some others have reported varying results on this
association. It was noted in several studies that surgical
patients with diabetes had higher risk of bedsore than
those without diabetes [12–18], while still others showed
null association [19–23]. Although two previous meta-
analyses have explored this topic and found significant
association between diabetes and surgery-related bedsore
[24, 25], limited sample size and significant heterogeneity
which was not sufficiently examined made the results less
reliable. Therefore, there is an urgent need to update the
evidence of association between preexisting diabetes and
surgery-related bedsore.

RESULTS

Search and selection of studies

The initial literature search yielded 1046 abstracts of
which 31 were considered potentially relevant for full-text
review. Totally, 16 studies including 24,112 participants
met our eligibility criteria and were involved in the meta-

analysis [12–23, 26–29]. Figure 1 gives the detailed
process for study selection of this meta-analysis.

Study characteristics

Table 1 presents the baseline characteristics of the
16 included studies. In summary, the included studies
were published between 1994 and 2013 with a sample
size ranging from 102 to 9400. Nine of the studies
were conducted in the USA and six in Europe. For case
ascertainment, 11 studies had a prospective study design,
and 5 had a retrospective study design. Four types of
surgical procedures including general surgery, hip surgery,
cardiac surgery and lower extremity amputations were
involved. Seven studies investigated patients more than
70 years in age, and 9 less than 70 years in age. Ten studies
applied univariate analysis and 6 studies used multivariate
analysis as statistical method. The NOS scores for the
assessment of methodological quality for cohort studies
ranged from 5 to 8, with scores ≥ 6 in 14 studies and
scores < 6 in 2 studies. The NOS score for the included studies were summarized in Table 3.

Figure 1: Flow diagram of the study selection.

Oncotarget14518www.impactjournals.com/oncotarget

Relationship between diabetes and risk of bedsore

Sixteeen cohort studies investigating the relationship
between diabetes and risk of bedsore in surgical patients
were included in our meta-analysis. The pooled RR was
1.77 (95% CI, 1.45 to 2.16) and there was statistical inter-
study heterogeneity (I2 = 62.7%; P < 0.001) (Figure 2).

Methodological quality of the studies

Table 2 presented the summary RRs for bedsore
risk and diabetes from 14 high quality studies (≥ 6) and
two low quality studies (< 6). In terms of methodological quality of studies, the summary RRs of bedsore risk were 1.72 (95% CI 1.40 to 2.10) in high quality studies and 2.07 (95% CI 1.04 to 4.14) in low quality studies, respectively,

in comparison between surgical patients with diabetes
and without diabetes. There was statistically significant
difference for inter-study heterogeneity (P = 0.009).

Type of surgery

Four types of sugery were involved in the
studies, with 6 of general surgery, 4 of hip surgery, 4 of
cardiac surgery and 2 of lower extremity amputations,
respectively. The summary RRs estimated for bedsore
incidence were 1.71 (95% CI 1.40 to 2.09) for general
surgery, 1.78 (95% CI 1.14 to 2.78 ) for hip surgery,
1.98 (95% CI 1.41 to 2.79) for cardiac surgery and 1.44
(95% CI 0.93 to 2.24) for lower extremity amputations,
respectively. No statistically significant difference for
inter-study heterogeneity (P = 0.238) was noted.

Table 1: Characteristics of the included studies on the risk of bedsore in diabetic patients undergoing
surgery

Oncotarget14519www.impactjournals.com/oncotarget

Study design

As demonstrated in Table 2, the pooled RRs
evaluated for bedsore risk were 1.96 (95% CI 1.52 to 2.52)
for prospective studies and 1.31 (95% CI 1.07 to 1.59)
for retrospective studies, respectively, with no significant
difference for inter-study heterogeneity (P = 0.017).

Sample size

The summarised RRs for bedsore risk stratified by
sample size were 1.66 (95% CI 1.21 to 2.29) for studies
with large sample size (≥ 1000) and 1.93 (95% CI 1.57
to 2.38) for studies with small sample size (< 1000). We found statistically significant difference for inter-study heterogeneity (P = 0.019).

Research region

Six and 9 studies were conducted in Europe and
USA, respectively. The summary RRs for bedsore risk
were 1.94 (95% CI 1.26 to 2.99) for studies conducted
in Europe and 1.62 (95% CI 1.33 to 1.97) for studies
conducted in USA. No statistically significant difference
for inter-study heterogeneity was found (P = 0.523).

Inclusion period

Four studies included participants before year 2000
and the pooled RR was 1.38 (95% CI 1.09 to 1.76); and
4 included participants after year 2000 with the pooled RR
of 1.61 (95% CI 1.30 to 2.00). We did not find statistically
significant difference for inter-study heterogeneity
(P = 0.089).

Age

The summarised RRs for bedsore risk stratified
by patient age were 1.66 (95% CI 1.19 to 2.32) for
studies with patients having older age (≥ 70 years) and
1.77 (95% CI 1.48 to 2.11) for studies with patients having
less older age (< 70). There was no statistically significant difference for inter-study heterogeneity (P = 0.078).

Statistical analysis method

Ten studies applied univariate analysis to analyze the
risk estimates and 6 studies applied multivariate analysis.
The results showed that the pooled RRs for bedsore risk
were 2.08 (95% CI 1.73 to 2.50) for studies with univariate
analysis and 1.44 (95% CI 1.11 to 1.88) for studies with

Figure 2: Association between diabetes and the risk of bedsore in patients undergoing surgery.

Oncotarget14520www.impactjournals.com/oncotarget

multivariate analysis. Statistically significant difference for
inter-study heterogeneity (P < 0.001) was found.

Publication bias and sensitivity analyses

The shape of the funnel plot for the studies on the
diabetes and bedsore risk seemed asymmetrical. In addition,
Egger’s adjusted rank correlation test showed potential
evidence of publication bias (P = 0.002). We further test
whether publication bias significantly influenced the
pooled risk estimates by using trim and filled method and
the adjusted RR indicated the same trend with the result
of the primary analysis (RR 1.17 95% CI 1.02 to 1.36).

A sensitivity analysis was carried out by excluding one
study at each time and then recalculating the pooled RRs
for the remaining ones to test the effect of each study on
the overall estimates. We did not find the alteration of in
the direction of the estimate when any one of the included
study was excluded. This analysis confirmed the robustness
of the positive association between diabetes and bedsore
risk in surgical patients.

DISCUSSION

Our systematic review and meta-analysis
summarizing the results of 16 observational studies,

Table 2: Subgroup analyses of the associations between diabetes and the risk of bedsore in patients
undergoing surgery

Variables RR 95% CI
Degree of

heterogeneity
(I2 statistics; %)

P
No. of

included
Studies

Pa

Total 1.77 1.45 to 2.16 62.7 < 0.001 16 Study quality 0.009

Score ≥ 6 1.72 1.40 to 2.10 58.1 0.002 14
< 6 2.07 1.04 to 4.14 65.4 0.089 2

Surgery type 0.238
General surgery 1.71 1.40 to 2.09 0 0.496 6
Hip surgery 1.78 1.14 to 2.78 88.4 < 0.001 4 Cardiac surgery 1.98 1.41 to 2.79 0 0.859 4 LEAs 1.44 0.93 to 2.24 0 0.414 2

Study design 0.017
Prospective 1.96 1.52 to 2.52 68.3 < 0.001 11 Retrospective 1.31 1.07 to 1.59 2.9 0.398 5

Sample size 0.019
≥ 1000 1.66 1.21 to 2.29 82.6 < 0.001 6 < 1000 1.93 1.57 to 2.38 0 0.856 10

Research region 0.523
Europe 1.94 1.26 to 2.99 80.8 < 0.001 6 USA 1.62 1.33 to 1.97 30.8 0.162 9

Inclusion period 0.089
Before year 2000 1.38 1.09 to 1.76 16.8 0.308 4
After year 2000 1.61 1.30 to 2.00 0 0.575 4

Age 0.078
≥ 70 1.66 1.19 to 2.32 78.3 < 0.001 7 < 70 1.77 1.48 to 2.11 0 0.695 9

Analysis method < 0.001 Univariate 2.08 1.73 to 2.50 10.6 0.342 10 Multivariate 1.44 1.11 to 1.88 64.8 0.014 6

Abbreviations: CI, confidence interval; LEA, Lower Extremity Amputations; RR, relative risk.
Pa: P values from the test of homogeneity between strata.

Oncotarget14521www.impactjournals.com/oncotarget

which comprised a total of 24,112 participants on the
association between diabetes and risk of bedsore support
the evidence that the risk of developing bedsore among
surgical patients exposed to diabetes was 1.77 times that
of the non-exposed patients. Analyses stratified by surgical
site suggest a greater risk increase for cardiac surgery than
for other three investigated surgeries (general surgery,
hip surgery or lower extremity amputations), though no
statistical significance is found among different surgery
types in this meta-analysis.

This updated meta-analysis further confirms and
extends the preliminary findings of the two previous
published meta-analyses [24, 25]. The first one performed
by Liu et al. [25], reported a 115% (OR, 2.15; 95% CI
1.62 to 2.84) higher risk of surgery-related bedsore in
diabetic patients compared with that of in non-diabetic
patients. The other one conducted by Kang et al. found
similar result (surgery related bedsore risk: diabates
vs. non-diabetes OR, 1.74; 95% CI 1.40 to 2.15) [24].
Our findings are consistent with the results of previous
systematic reviews. We also explored the effect of
different surgery types and other potential variables
more thoroughly on the combined estimates than the
previous ones. Compared with the study by Kang et al.,
our meta-analysis has added more statistical power to
test the surgery type subgroup and examined some other
variables which could explain the potential heterogeneity.
This study found that diabetic patients having general
surgery, hip surgery and cardiac surgery all had significant

higher bedsore risk than non-diabetic patients. However,
we did not find that association for patients having lower
extremity amputations.

Multiple mechanisms can contribute to the
deveopment and severity of bedsores, which may result
from capillary occlusion by external pressure, leading to
the shut off of blood supply, cell death, necrosis removal
and ulceration. The severity of the bedsore is determined
by the length of time pressure is applied to the local
region. Moreover, for a patient receiving surgery, the
incidence rate of bedsores is mainly determined by the
duration and intensity of the shearing force given upon the
tissue during surgery. For the impact of different surgery
types on the risk of bedsore development, we noted that
patients with cardiact surgery had the higher risk (RR 1.98,
95% CI 1.41 to 2.79) than patients with general surgery
or hip surgery, while patients with lower extremity
amputations had the lowest risk (RR 1.44, 95% CI 0.93
to 2.24). We propose that the trauma severity of surgery
to the body could be a major influential factor determining
the risk of developing bedsore.

We noted moderate inter-study heterogeneity in our
meta-analysis (I2 = 62.7%, Pheterogeneity < 0.001). Sensitivity analyses indicated that exclusion of any one of the study did not significantly alter the summary estimate. The trim- and-fill model and multiple subgroup analyses stratified by some main clinical variables were in agreement with the initial findings, indicating that the result of this meta- analysis was robust and not affected by publication bias.

Table 3: Quality assessment of the included studies
Selection Comparability Outcome

Study ID

Represent
ativeness

of the

exposed
cohort

Selection
of the
non

exposed
cohort

Ascertain
ment of

exposure

Demonstration
that outcome

of interest
was not

present at
start of study

Comparability
of cohorts on

the basis of the
design or
analysis

Assessment
of outcome

Was follow-up
long enough for

outcomes to
occur

Adequacy
of follow

up of
cohorts

Quality
score

1 Zambonato 2013      5
2 Ekström 2013        8
3 Tschannen 2012        8
4 Bulfone 2012       6
5 Norris-DOI 2011       6
6 Slowikowski 2010        8
7 Aragón-Sánchez 2010       6
8 Haleem 2008      5
9 Frankel 2007       7
10 Pokorny 2003       6
11 Baumgarten 2003        8
12 Spittle 2001       6
13 Schultz 1999        8
14 Stordeur 1998       6
15 Lewicki-

preoperative
1997      

6
16 Papantonio 1994        7

Newcastle-Ottawa Scale for assessing the quality of studies in meta-analysis.
Note: A study can be awarded a maximum of one star for each numbered item within the Selection and Outcome categories. A maximum of two stars can be given for
Comparability.

Oncotarget14522www.impactjournals.com/oncotarget

Nevertheless, we should interpret the results with caution
due to the common occurance of publication bias [30] and
statistical tests to detect publication bias are incomplete.

Despite the previous published studies investigating
the association between diabetes and risk of surgery-related
bedsore, the statistical power was quite limited for the small
sample sizes of these studies (ranging from 67 to 616).
To the best of our knowledge, our study is the most
comprehensive one with the largest sample size to evaluate
this association. Furthermore, exhaustive search strategies
were developed to garantee the inclusion of almost all of
the eligible studies, generating 16 studies and data from
24,112 individuals. Such a large sample size could provide
us a precise and important risk estimates. Moreover, based
on the subgroup analyses, our study also showed that

bedsore risk increased among different types of
surgery although statistical significance was not noted
for lower extremity amputations probably due to limited
sample size. Lastly, consistent and stable sensitivity
analyses and result of trim and filled method made the
results more strengthened.

Several limitations in our study should be
acknowledged. First, variations of treatment or nursing
procedures for different types of surgery, may result
in variations in risk estimates. Secondly, in order to
assess the effect of different blood glucose levels or
patient body mass index on the different risk of bedsore,
related subgroup analyses should ideally be performed.
However, due to the nature of study-level data instead
of patient-level data, the available data did not allow us
to conduct such assessment. Thirdly, 10 of 16 studies
used univariate analysis instead of multivariate analysis
to obtain the risk estimates as they did not adjust for
some potential influential confounders, such as gender,
patient age, diabetes duration and type, which could
lead to inaccurately generating the pooled estimates. In
addition, for the studies using multivariate analysis, the
adjustment variables varied considerably. Moreover,
the data sources from observational studies restricted
the power to fully explore the influence of unmeasured
confounding variables and observational studies could
not establish a causal relationship between exposure
factor of diabetes and risk of bedsore. Finally, some of the
study authors could not be contacted for retrieving some
necessary data. Despite the limitations of the current study,
the major clinical implication lies in that for some types of
surgery, clinicians should take more care of patients with
diabetes to mininize the development of surgery-related
bedsore and improve the quality of patient life during
hospitalization.

In conclusion, our systematic review and meta-
analysis provide evidence that diabetic patients having
surgery could have a higher risk of developing bedsore.
This association is almost independent of surgery type and
other study characteristics. However, further large-scale
prospective studies should be implemented to further test
the association.

MATERIALS AND METHODS

Search strategy

We systematically searched Pubmed, EMBASE,
and the Cochrane Library without language restriction
through November 2016 for related peer-reviewed studies
that examined an association between diabetes and risk
of bedsore in patients undergoing surgery. We performed
this systematic review and meta-analysis based on the
Preferred Reporting Items for Systematic Reviews and
Meta-Analyses (PRISMA) (Supplementary Table S4) [31].
Two authors (M.L. and Q.C.) independently conducted
the literature search using the terms: (surgery OR surgical
OR operation OR operative) AND (diabetes mellitus OR
diabetes) AND (pressure sore* OR pressure ulcer* OR
bedsore* OR decubitus). Manual searches of reference
lists of relevant studies obtained from the initial searches
were also conducted for some missing citations. Detailed
search strategies of each database are provided in
Supplementary Appendix.

Study selection

Two reviewers (M.L. and Q.C.) independently
assessed all records through reading the titles and/or
abstracts for potentially eligible studies. In case there were
different opinions, a senior reviewer (L.L ) would join to
discuss and resolve the disagreement. We included studies
in this meta-analysis if they satisfied the following criteria:
(i) observational studies including cohort or case–control
studies; (ii) investigating diabetes and risk of bedsore in
patients undergoing surgery; (iii) providing odds ratios
(ORs)/relative risks (RRs) along with 95% confidence
intervals (95% CIs) or sufficient information to calculate
them, for bedsore risk stratified by diabetes in patients
having operation. We included patients with history or
diagnosis of diabetes, irrespective of diabetes type (1 or 2),
disease severity, duration or anti-diabetic drug use due to
unavailability of those data.

Data extraction and quality assessment

Data were extracted independently according to a
predesigned form by two reviewers (Y.Z. and L.H.) and the
results were crosschecked. A third reviewer(L.L.) would
reevaluated the extracted data if any disagreements occurred.
The following data were extracted from each study:
first author, publication year, study region, study design,
inclusion period, opertion site, number of participants, sex,
mean/median age,body mass index, treatment regimen,
analysis method, follow up period, adjustment variables,
and risk estimates for association between bedsore risk and
diabetes in patients having operation.

Two reviewers (M.L. and L.H.) independently
assessed the methodological quality of each included
study using Newcastle-Ottawa quality assessment

Oncotarget14523www.impactjournals.com/oncotarget

scale (NOS) for cohort studies, which included
3 domains (4 points for selection, 2 points for comparability
and 3 points for exposure/outcome) totaling 9 points
(Table 3). We categoried score less than 6 as low quality
and score of 6 or more than 6 as high quality. Discrepancies
were resolved by consensus with a senior reviewer (L.L.).

Statistical analyses

We quantified the relationship between diabetes and
risk of bedsore using an inverse variance method using
DerSimonian and Laird random-effects models [32]. All
statistical analyses were carried out with Stata Statistical
Software (version 12.0; StataCorp LP, College Station, TX,
USA) by two reviewers (M.L. and L.H.). Between-study
heterogeneity was assessed using the chi-square statistic and
quantified by I², with an I2 statistic more than 50% defining
significant heterogeneity [33, 34]. We further investigated
potential sources of between-study heterogeneity by
subgroup analyses based on some baseline variables
(study quality, surgery type, study design, sample size,
research region, inclusion period, patient age and analysis
method). Egger’s regression model was quantified to assess
publication bias [35]. If publication bias existed, we used
the trim-and-fill method to adjust the pooled estimates
of the potential unpublished studies in the meta-analysis,
which were compared with the original pooled RRs [36].
Sensitivity analysis was also conducted to investigate the
influence of each study on the separate analyses of cohort
studies [26]. All statistical analyses were two-sided with a
P value less than 0.05 indicating significant difference.

CONFLICTS OF INTEREST

The authors declare no potential conflicts of interest.

FINANCIAL SUPPORT

None.

Authors ̕contributions

Study concept and design (ML LL); acquisition of
data (ML QC); analysis and interpretation of data (ML
QC YZ LH LL); drafting of the manuscript (ML QC
LL); critical revision of the manuscript for important
intellectual content (ML QC YZ LH JW YC LL); study
supervision.

REFERENCES

1. Price MC, Whitney JD, King CA. Development of a risk
assessment tool for intraoperative pressure ulcers. Journal
of Wound Ostomy & Continence Nursing. 2005; 32:19–30.

2. Aronovitch SA. Intraoperatively acquired pressure ulcer
prevalence: a national study. Journal of Wound Ostomy &
Continence Nursing. 1999; 26:130–136.

3. Allman RM, Goode PS, Burst N, Bartolucci AA,
Thomas DR. Pressure ulcers, hospital complications, and
disease severity: impact on hospital costs and length of stay.
Advances in skin & wound care. 1999; 12:22–30.

4. Fu Shaw L, Chang PC, Lee JF, Kung HY, Tung TH.
Incidence and predicted risk factors of pressure ulcers in
surgical patients: experience at a medical center in Taipei,
Taiwan. BioMed research international. 2014; 2014.

5. Lannering C, Ernsth Bravell M, Midlöv P, Östgren CJ,
Mölstad S. Factors related to falls, weight-loss and pressure
ulcers–more insight in risk assessment among nursing home
residents. Journal of clinical nursing. 2016.

6. Nassaji M, Askari Z, Ghorbani R. Cigarette smoking and
risk of pressure ulcer in adult intensive care unit patients.
International journal of nursing practice. 2014; 20:418–423.

7. Thomas DR. Role of nutrition in the treatment and
prevention of pressure ulcers. Nutrition in Clinical Practice.
2014; 29:466–472.

8. Armstrong D, Bortz P. An integrative review of pressure
relief in surgical patients. AORN journal. 2001; 73:645–648,
650–643, 656–647 passim.

9. Aronovitch SA. Intraoperatively acquired pressure ulcer
prevalence: a national study. Journal of wound, ostomy, and
continence nursing. 1999; 26:130–136.

10. Beckett AE. Are we doing enough to prevent patient injury
caused by positioning for surgery? Journal of perioperative
practice. 2010; 20:26–29.

11. Chalian AA, Kagan SH. Backside first in head and neck
surgery?: preventing pressure ulcers in extended length
surgeries. Head & neck. 2001; 23:25–28.

12. Bulfone G, Marzoli I, Quattrin R, Fabbro C, Palese A. A
longitudinal study of the incidence of pressure sores and the
associated risks and strategies adopted in Italian operating
theatres. Journal of perioperative practice. 2012; 22:50–56.

13. Frankel H, Sperry J, Kaplan L. Risk factors for pressure
ulcer development in a best practice surgical intensive care
unit. The American surgeon. 2007; 73:1215–1217.

14. Haleem S, Heinert G, Parker MJ. Pressure sores and hip
fractures. Injury. 2008; 39:219–223.

15. Norris R, Parker M. Diabetes mellitus and hip fracture: a
study of 5966 cases. Injury. 2011; 42:1313–1316.

16. Schultz A, Bien M, Dumond K, Brown K, Myers A.
Etiology and incidence of pressure ulcers in surgical
patients. AORN journal. 1999; 70:434, 437–440, 443–439.

17. Slowikowski GC, Funk M. Factors associated with pressure
ulcers in patients in a surgical intensive care unit. Journal of
wound, ostomy, and continence nursing. 2010; 37:619–626.

18. Tschannen D, Bates O, Talsma A, Guo Y. Patient-specific
and surgical characteristics in the development of pressure
ulcers. American journal of critical care. 2012; 21:116–125.

19. Zambonato BP, de Assis MC, Beghetto MG. [Association of
Braden subscales with the risk of development of pressure
ulcer]. Revista gaucha de enfermagem/EENFUFRGS. 2013;
34:21–28.

Oncotarget14524www.impactjournals.com/oncotarget

20. Stordeur S, Laurent S, Hoore Wd. The importance of repeated
risk assessment for pressure sores in cardiovascular surgery.
Journal of Cardiovascular Surgery. 1998; 39:343–350.

21. Pokorny ME, Koldjeski D, Swanson M. Skin care
intervention for patients having cardiac surgery. American
Journal of Critical Care. 2003; 12:535–544.

22. Ekström W, Al-Ani AN, Sääf M, Cederholm T, Ponzer S,
Hedström M. Health related quality of life, reoperation
rate and function in patients with diabetes mellitus and hip
fracture-A 2 year follow-up study. Injury. 2013; 44:769–775.

23. Aragón-Sánchez J, Hernández-Herrero MJ, Lázaro-
Martínez JL, Quintana-Marrero Y, Maynar-Moliner M,
Rabellino M, Cabrera-Galván JJ. In-hospital complications
and mortality following major lower extremity amputations
in a series of predominantly diabetic patients. International
Journal of Lower Extremity Wounds. 2010; 9:16–23.

24. Kang ZQ, Zhai XJ. The Association between Pre-existing
Diabetes Mellitus and Pressure Ulcers in Patients Following
Surgery: A Meta-analysis. Scientific reports. 2015; 5:13007.

25. Liu P, He W, Chen HL. Diabetes mellitus as a risk factor for
surgery-related pressure ulcers: a meta-analysis. Journal of
wound, ostomy, and continence nursing. 2012; 39:495–499.

26. Baumgarten M, Margolis D, Berlin JA, Strom BL, Garino J,
Kagan SH, Kavesh W, Carson JL. Risk factors for pressure
ulcers among elderly hip fracture patients. Wound repair
and regeneration. 2003; 11:96–103.

27. Lewicki LJ, Mion L, Splane KG, Samstag D, Secic M.
Patient risk factors for pressure ulcers during cardiac
surgery. AORN journal. 1997; 65:933–942.

28. Papantonio CT, Wallop JM, Kolodner KB. Sacral ulcers
following cardiac surgery: incidence and risks. Advances in
skin & wound care. 1994; 7:24–37.

29. Spittle M, Collins RJ, Connor H. The incidence of pressure
sores following lower limb amputations. Practical Diabetes
International. 2001; 18:57–61.

30. Ioannidis JP. How to make more published research true.
PLoS medicine. 2014; 11:e1001747.

31. Liberati A, Altman DG, Tetzlaff J, Mulrow C, Gotzsche PC,
Ioannidis JP, Clarke M, Devereaux PJ, Kleijnen J, Moher D.
The PRISMA statement for reporting systematic reviews
and meta-analyses of studies that evaluate health care
interventions: explanation and elaboration. Ann Intern Med.
2009; 151:W65–94.

32. DerSimonian R, Laird N. Meta-analysis in clinical trials.
Controlled clinical trials. 1986; 7:177–188.

33. Greenland S. Quantitative methods in the review of
epidemiologic literature. Epidemiologic reviews. 1987;
9:1–30.

34. Higgins JP, Thompson SG. Quantifying heterogeneity in a
meta-analysis. Statistics in medicine. 2002; 21:1539–1558.

35. Egger M, Davey Smith G, Schneider M, Minder C. Bias
in meta-analysis detected by a simple, graphical test. Bmj.
1997; 315:629–634.

36. Duval S, Tweedie R. Trim and fill: A simple funnel-plot-
based method of testing and adjusting for publication bias
in meta-analysis. Biometrics. 2000; 56:455–463.