Page 1
ral
ssBioMed CentBMC Public Health
Open AcceStudy protocol
Impact of an informed choice invitation on uptake of screening for
diabetes in primary care (DICISION): trial protocol
Eleanor Mann1, A Toby Prevost2, Simon Griffin3, Ian Kellar2,
Stephen Sutton2, Michael Parker4, Simon Sanderson2, Ann Louise Kinmonth2
and Theresa M Marteau*1
Address: 1Psychology Department (at Guy's), Health Psychology Section, Psychology and Genetics Research Group, 5th Floor Bermondsey Wing,
Guy's Campus, London, SE1 9RT, UK, 2Department of Public Health and Primary Care, University of Cambridge, Forvie Site, Robinson Way,
Cambridge, CB2 0SR, UK, 3MRC Epidemiology Unit, Institute of Metabolic Science, Box 285, Addenbrooke's Hospital, Hills Road, Cambridge,
CB2 0QQ, UK and 4The Ethox Centre, Division of Public Health and Primary Health Care, University of Oxford, Badenoch Building, Old Road
Campus, Headington, Oxford, OX3 7LF, UK
Email: Eleanor Mann - eleanor.mann@iop.kcl.ac.uk; A Toby Prevost - toby.prevost@phpc.cam.ac.uk; Simon Griffin - simon.griffin@mrc-
epid.cam.ac.uk; Ian Kellar - ik261@medschl.cam.ac.uk; Stephen Sutton - srs34@medschl.cam.ac.uk;
Michael Parker - michael.parker@ethox.ox.ac.uk; Simon Sanderson - simon.sanderson@phgfoundation.org;
Ann Louise Kinmonth - alk25@medschl.cam.ac.uk; Theresa M Marteau* - theresa.marteau@kcl.ac.uk
* Corresponding author
Abstract
Background: Screening invitations have traditionally been brief, providing information only about
population benefits. Presenting information about the limited individual benefits and potential
harms of screening to inform choice may reduce attendance, particularly in the more socially
deprived. At the same time, amongst those who attend, it might increase motivation to change
behavior to reduce risks. This trial assesses the impact on attendance and motivation to change
behavior of an invitation that facilitates informed choices about participating in diabetes screening
in general practice. Three hypotheses are tested:
1. Attendance at screening for diabetes is lower following an informed choice compared with a
standard invitation.
2. There is an interaction between the type of invitation and social deprivation: attendance
following an informed choice compared with a standard invitation is lower in those who are more
rather than less socially deprived.
3. Amongst those who attend for screening, intentions to change behavior to reduce risks of
complications in those subsequently diagnosed with diabetes are stronger following an informed
choice invitation compared with a standard invitation.
Method/Design: 1500 people aged 40–69 years without known diabetes but at high risk are
identified from four general practice registers in the east of England. 1200 participants are
randomized by households to receive one of two invitations to attend for diabetes screening at
their general practices. The intervention invitation is designed to facilitate informed choices, and
Published: 20 February 2009
BMC Public Health 2009, 9:63 doi:10.1186/1471-2458-9-63
Received: 23 December 2008
Accepted: 20 February 2009
This article is available from: http://www.biomedcentral.com/1471-2458/9/63
© 2009 Mann et al; licensee BioMed Central Ltd.
This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0),
which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.Page 1 of 12
(page number not for citation purposes)
comprises detailed information and a decision aid. A comparison invitation is based on those
currently in use. Screening involves a finger-prick blood glucose test. The primary outcome is
attendance for diabetes screening. The secondary outcome is intention to change health related
Page 2
BMC Public Health 2009, 9:63 http://www.biomedcentral.com/1471-2458/9/63
behaviors in those attenders diagnosed with diabetes. A sample size of 1200 ensures 90% power
to detect a 10% difference in attendance between arms, and in an estimated 780 attenders, 80%
power to detect a 0.2 sd difference in intention between arms.
Discussion: The DICISION trial is a rigorous pragmatic denominator based clinical trial of an
informed choice invitation to diabetes screening, which addresses some key limitations of previous
trials.
Trial registration: Current Controlled Trials ISRCTN73125647
Background
Invitations to attend for screening traditionally provide
information about its population benefits and aim to
achieve high rates of uptake [1]. They do not usually refer to
the likelihood of health benefits for the individual, which are
low [2], nor to the possibility of adverse effects. This neglect
reflects a greater concern with potential public health bene-
fits than with individual autonomy. There has, however,
been a policy change in the UK and elsewhere [3,4] towards
a view that participation in screening programmes should
reflect individual choices informed about both the nature
and frequency of possible individual benefits and harms of
screening. Invitations to participate in screening do not rou-
tinely reflect this policy change. This may reflect a reluctance
of those organizing screening programmes to implement a
policy change that may privilege concern for informed
choice to the neglect of achieving the public health benefits
of screening [5]. Giving information about the type and fre-
quencies of individual benefits and burdens could deter
some people from participating in screening programmes [6-
9]. We therefore test the hypothesis that facilitating informed
choice results in lower screening attendance than when
informed choices are not facilitated.
A further concern is that invitations that support informed
choice may reduce uptake differentially across social groups
resulting in lower uptake amongst the more socially
deprived. These groups are also likely to be those in poorest
health and hence those in whom the benefits of screening
might be greatest [10]. Given that participation in screening
programmes is already lower in these groups [11], an
informed choice policy could increase the health gap
between socially more and less advantaged groups in one
of two ways. Information on the limitations of screening
may have greater impact upon the more socially deprived
because they are less aware of such limitations [12-14].
Information about the possible longer term benefits and
more immediate harms of screening may also be more
demotivating among this group which is known to be ori-
ented more towards the present than to the future, both
generally and in relation to their health [15,16]. Indeed,
one of the few studies to examine the impact of risk infor-
but decreased attendance in those with lower levels [17].
Consequently, we predict that an informed choice invita-
tion will differentially affect attendance among individuals
who are more socially deprived compared with those who
are less socially deprived.
Informed choice invitations attempt to enhance individual
autonomy. This may lead to increased motivation to reduce
identified risk among those accepting the invitation for
screening [18-22]. According to Self Determination Theory
[22] autonomous decisions are more intrinsically motivated;
that is, the individual has greater personal interest in the
behavior, thus the decision is more likely to lead to action
[23]. In the case of diabetes, the most effective way of reduc-
ing risk is by changing behavior (improving diet, increasing
physical activity, taking medication) [24,25]. Therefore, we
predict stronger intentions to engage in these risk reducing
behaviors in those attending for diabetes screening after
receiving an informed choice invitation than in those who
received a standard invitation.
The ethical dilemma central to the proposed research con-
cerns the potential conflict between two fundamental
moral principles guiding health care practice and policy:
on the one hand, patient-centered practice, which privi-
leges the principle of respect for individual autonomy;
and on the other, the promotion of public health benefits,
which privileges achieving the greatest overall benefit. The
current trial examines whether such an ethical conflict fol-
lows the implementation of an informed choice policy in
the context of screening.
Achieving informed choices in screening
Informed choices or decisions can be considered to have
two theoretical core characteristics: first, they should be
informed by best current evidence; and second, they
should reflect the decision-maker's values [26,27]. In
making an informed choice people are neither deceived
nor coerced [28]. Individuals first need good quality infor-
mation that can be read and understood across a wide
range of literacy as a necessary, although not sufficient,
basis for an informed choice.Page 2 of 12
(page number not for citation purposes)
mation across social groups found that a risk counseling
intervention designed to increase attendance for mammog-
raphy had no impact on those with high levels of education
Decision aids aimed at conveying written information
and structuring the decision-making process can be used
Page 3
BMC Public Health 2009, 9:63 http://www.biomedcentral.com/1471-2458/9/63
to facilitate informed choices about a range of health
interventions including screening. The few published
evaluations of self-administered decision aids provide
some evidence that knowing about and evaluating the
personal importance of the benefits and risks associated
with a particular choice can help individuals to clarify
their values. This improves the quality of their subsequent
decisions [29]. Thus, for example, the use of a decision aid
supporting women's decisions to use hormone replace-
ment therapy resulted in more accurate perceptions of
breast cancer risks, greater confidence to make a decision
and more satisfaction with the decision made [30]. How-
ever, the evidence is limited and mixed. Four of the 10
studies reviewed by O'Connor and colleagues [29] found
significant improvements in decision quality, as meas-
ured by the decisional conflict scale [31]; the remaining
six studies found no significant effects of a decision aid in
improving the decision making process.
The impact of decision aids on screening uptake has also
been mixed and small. Krist, Woolf, Johnson, & Kerns
[32] found requests for prostate cancer screening tests
were lower in those using a decision aid, whereas Mathieu
and colleagues [33] reported no difference in breast can-
cer screening uptake in women using a decision aid com-
pared to those not using the aid. Similarly, Trevena, Irwig,
& Barratt [34] found no impact of a decision aid on rates
of self reported use of colorectal screening kits.
Differences in findings between studies may relate to the
characteristics of the groups invited, the nature of the invi-
tation or decision aid, the precision and invasiveness of
the test, the prevalence, severity and treatability of the dis-
ease screened for, or the rigor of the study designs and
measures used to evaluate screening interventions. Trials
of informed choice in screening may be particularly vul-
nerable to methodological bias. Traditionally, screening
has been promoted by health professionals, so self report
measures of screening attendance might be inaccurate.
Furthermore, the consent process might select partici-
pants who are more likely to attend for screening [35]. For
example, people refusing consent to participate in a
screening trial may do so because they have no intention
of attending for screening. In a trial of colorectal cancer
screening, over 75% of general practice patients contacted
chose not to consent to take part in the trial [34]. By con-
trast Krist and colleagues [32] recruited men to a prostate
cancer trial taking place during a health examination that
the men had already scheduled. Only 5% did not consent.
In addition, there is evidence that questionnaires can alter
research outcomes [36-38]; baseline questionnaires may
prompt greater reflection on aspects of screening covered
in the measures, thereby altering responses.
have designed a rigorous pragmatic, population based ran-
domized trial evaluating the effect on uptake of screening
for type 2 diabetes of a validated invitation designed to
inform the choice to attend. This trial protocol is designed
to measure real screening decisions, in a primary care set-
ting, unbiased by prior consent to study participation. This
is the first trial of an informed choice decision aid for
screening which includes all these design features.
Screening for type 2 diabetes
Screening for type 2 diabetes and cardiovascular risk was
chosen for study as type 2 diabetes is a condition for
which behavior change (increasing physical activity,
improving diet adherence to medication) is a major com-
ponent of treatment [25], and for which at-risk popula-
tion-based screening programmes are now being
proposed and implemented [39-41]. Type 2 diabetes is a
serious condition that is commonly undiagnosed until
complications occur [42,43]. Existing evidence suggests
that the adverse consequences of screening are likely to be
limited [44,45] and that intensive treatment of clinically
diagnosed patients is beneficial [24].
Uncertainties remain in relation to the overall cost-effec-
tiveness of detection by screening and initiation of treat-
ment earlier in the trajectory of the disease [43]. Evidence
suggests that screening individuals at high risk of undiag-
nosed diabetes is most cost-effective [43,46], and that pre-
venting complications arising from diabetes through
behavior change in individuals in the early stages of the
disease is more effective than treatments for those compli-
cations [47,48]. Patients at higher risk of undiagnosed
diabetes can be identified using the Cambridge Risk Score
[42,49-51], which can be used to select patients to invite
for diabetes screening and to explore the impact of risk of
undiagnosed diabetes on lifestyle choices.
Study objective
The study objective is to estimate the impact upon attend-
ance for diabetes screening of an informed choice invita-
tion compared with a standard invitation. This is
examined both overall and stratified by social depriva-
tion. A secondary objective is to describe, amongst attend-
ers, intention to change health-related behaviors if
diabetes were subsequently detected. Intentions amongst
attenders receiving the informed choice and standard invi-
tations are compared.
The trial tests three principal hypotheses:
1. Attendance at screening for diabetes is lower following
an informed choice compared with a standard invitationPage 3 of 12
(page number not for citation purposes)
Given the uncertain effect of attempts to support informed
choice and to evaluate its effects on screening uptake we
2. There is an interaction between the type of invitation
and social deprivation: attendance following an informed
Page 4
BMC Public Health 2009, 9:63 http://www.biomedcentral.com/1471-2458/9/63
choice compared with a standard invitation is lower in
those who are more rather than less socially deprived.
3. Amongst those who attend for screening, intentions to
change behavior to reduce risks of complications in those
subsequently diagnosed with diabetes are stronger follow-
ing an informed choice invitation compared with a stand-
ard invitation
Methods
Design
The study design and participant recruitment (completed
July 2007) is shown in figure 1. DICISION is a rand-
omized controlled trial testing the impact of an invitation
designed to facilitate informed choice (referred to as the
informed choice invitation) on attendance for type 2 dia-
betes screening. The group that receives the informed
choice invitation is compared to a group that receives an
invitation typical of current practice (referred to as the
standard invitation). Randomization of individuals by
household clusters is undertaken from a central site fol-
lowing stratification by cluster size and general practice.
The trial is set in four general practices and participants are
recruited from practice lists. The informed choice invita-
tion is designed to be replicable in a primary care setting
using minimal resources. The design protects the primary
end point by measuring attendance before providing fur-
ther information about the trial or collecting question-
naire data that may have intervention effects. The trial is
managed jointly between Kings College London and the
University of Cambridge. COREC approval has been
given (REC: Cambridgeshire 1: 06/Q0104/17, 05th May,
2006), and R&D approval has been obtained from Greater
Peterborough PCP, and Suffolk West PCT.
Practice recruitment
The study is conducted in four general practices recruited
from Cambridgeshire and surrounding counties, including
two with above average area deprivation scores (in or above
the middle quintile of the Index of Multiple Deprivation
2004). The practices are selected using a combination of
information on practice participation in the Wellcome
funded ADDITION trial (ISRCTN86769081), deprivation
scores, and local interest in participation in research.
14 practices are selected for initial contact by letter, to
invite participation and request a meeting between the
practice team and the research team. At this meeting,
research and clinical procedures are outlined, and mem-
bers of the practice team are provided with a written sum-
mary of the study and a Research Information Sheet for
Practices [52]. Before making a decision they are asked to
discuss the study with the rest of the practice team and
Participant recruitment
The sample is selected from practice registers. It includes
participants without known diabetes and in the top 25% of
risk of having prevalent undiagnosed diabetes defined by a
validated risk score. The Cambridge Risk Score [42] has
been validated as a pre-screening instrument for the identi-
fication of those at increased risk of having prevalent undi-
agnosed diabetes [42,49-51,53]. For example, Griffin and
colleagues [42] reported an area of 80% under a receiver-
operating characteristic curve. Risk scores are obtained
from a MIQUEST READ code based search of routinely-
held practice data (weight, height, BMI, age, gender, antihy-
pertensive and steroid medication). Risk scores have been
calculated for over 150,000 people in practices in Cam-
bridgeshire and surrounding counties participating in the
ADDITION trial [54]. It is estimated that sufficient practice
data are recorded to enable risk score calculations to be
undertaken for around 70% of patients.
Exclusion criteria
Patients meeting the following criteria are excluded from
the trial:
i. Diagnosed with diabetes since the medical record search
was performed.
ii. Pregnant or breast-feeding.
iii. Have a psychotic illness, such as schizophrenia, hypo-
mania, major depression, manic depression.
iv. Have an illness with a likely prognosis of less than one
year to death.
Consent process
Potential participants are sent a letter from the practice
with which they are registered informing them that the
practice is going to undertake screening for diabetes as
part of a research trial, and asking them to opt out if they
do not wish to be invited. Those wishing to opt out are
asked to return a freepost card to the practice.
When individuals attend for screening they are given the
trial "patient information sheet" and then asked for con-
sent to complete two questionnaires and for screening test
results to be seen by researchers. For those who do not
attend, information is given and consent obtained at the
time of sending a questionnaire by post.
Development and evaluation of the intervention materials
Standard invitation
The standard invitation is modeled on invitations com-
monly used to invite people for diabetes and coronaryPage 4 of 12
(page number not for citation purposes)
inform the trial co-ordinator of their decision within one
month. Four study practices and two reserve practices are
needed to ensure a sufficient sample size.
heart disease screening [11,55]. The text states that the par-
ticipant is being offered screening because they might have
a higher chance of developing type 2 diabetes, and that dia-
Page 5
BMC Public Health 2009, 9:63 http://www.biomedcentral.com/1471-2458/9/63
Page 5 of 12
(page number not for citation purposes)
Diagram of participant flow through the DICISION trialFigure 1
Diagram of participant flow through the DICISION trial.
FBG App’t
RBG Appointment
Arrive at appointment
Primary outcome: Attendance
Complete a questionnaire
Secondary outcome: Intention to change
lifestyle if diabetes were diagnosed
Other outcomes: SES, worry about
diabetes, expected results, ZTPI*
Sent invitation:
Take a RBG test
Normal result
Give health promotion
information
Positive result
FBG appointment
offered
Normal result:
Measure CVD risk
Give written record of
results and health
promotion information
Positive result:
Refer to
practice for
diagnostic test
Within 7 days
4 weeks later
2-3 weeks later
Take FBG test
Measure height weight,
BP and lipids
Send non-attender
questionnaire
Other outcome: intention
to change lifestyle
Other measures: SES,
worry about diabetes,
ZTPI*, decision satisfaction
Send time 2 questionnaire
Other outcome: intention to change lifestyle
Other measures: decision satisfaction, perceived
impact of results on diabetes status
Did not attend
Randomised
N = 1272
Excluded N = 228
- Criteria not met n = 43
- Opted out n = 183
- Opt out undelivered n = 2
Assessed for
eligibility
N = 1500
Allocated to receive the
standard invitation
N = 639
Allocated to receive the
informed choice invitation
N = 633
*Zimbarbo Time Perspective Inventory (short form)
End of preview.
