ABI, VC and mortality 325

Downloaded from http://ndt.oxfordjournals.org/ by Gabriela Elena Voicu on March 27, 2012

Nephrol Dial Transplant (2012) 27: 325–331
doi: 10.1093/ndt/gfr266
Advance Access publication 3 June 2011

Patient stories about their dialysis experience biases others’ choices

regardless of doctor’s advice: an experimental study

Anna E. Winterbottom1, Hilary L. Bekker1, Mark Conner2 and Andrew F. Mooney3

1
Leeds Institute of Health Sciences, University of Leeds, Leeds, UK, 2Institute of Psychological Sciences, University of Leeds, Leeds,
UK and 3Adult Renal Services, St James University Hospital, Leeds, UK
Correspondence and offprint requests to: Anna E. Winterbottom; E-mail: anna@winterbottom.co.uk

Abstract Results. In both studies, participants were more likely to

Background. Renal services provide resources to support
patients in making informed choices about their dialysis
modality. Many encourage new patients to talk with those
already experiencing dialysis. It is unclear if these stories
help or hinder patients’ decisions, and few studies have
been conducted into their effects. We present two studies
comparing the impact of patient and doctor stories on hypo-
thetical dialysis modality choices among an experimental
population.
Methods. In total, 1694 participants viewed online informa-
tion about haemodialysis and continuous cycling peritoneal
dialysis and completed a questionnaire. In Study 1, using
actors, treatment information was varied by presenter
(Doctor, Patient), order of presenter (Patient first, Doctor
first) and mode of delivery (written, video). Information in
Study 2 was varied (using actors) by presenter (Doctor,
Patient), order of presenter (Patient first, Doctor first),
inclusion of a decision table (no table, before story, after
story) and sex of the ‘patient’ (male, female) and ‘Doctor’
(male, female). Information was controlled to ensure
comparable content and comprehensibility.

The Author 2011. Published by Oxford University Press on behalf of ERA-EDTA. All rights reserved.
For Permissions, please e-mail: journals.permissions@oup.com
choose the dialysis modality presented by the patient rather
than that presented by the doctor. There was no effect for
mode of delivery (video versus written) or inclusion of a
decision table.
Conclusions. As ‘new’ patients were making choices
based on past patient experience of those already on dial-
ysis, we recommend caution to services using patient sto-
ries about dialysis to support those new to the dialysis in
delivering support to those who are new to the decision
making process for dialysis modality.
Keywords: decision making; dialysis; patient stories
Introduction
Uptake of dialysis modalities vary considerably geograph-
ically around the world and uptake of different modalities
show changes over time within those areas [1, 2, 3]. The
factors involved in choosing a dialysis modality are com-
plex and depend upon clinical need, organizational
326
structure, professional views and patient preferences. How-
ever, national and international guidelines exist which lay
out the level of information that patients should receive
prior to choosing a dialysis modality [4, 5]. In general,
patients are encouraged to be involved in choosing their
dialysis modality, so that their choices reflect their prefer-
ences [6]. To do so, patients require comprehensive good
quality information on which to base their decision. Renal
units put considerable effort into providing information to
support patients using their services and to help them make
decisions about renal replacement therapies [7]. However,
despite clear and concise recommendations of the most
effective methods of supporting patient decision making
[8], most of these resources are developed without aware-
ness of evidence from the decision sciences about how the
presentation of information influences people’s choices [7,
8]. This expertise is useful in designing patient resources to
ensure their structure and content are sufficiently proactive
to engage patients in making informed decisions about
dialysis modality [9, 10]. It is also useful in understanding
what everyday service factors may be influencing inadver-
tently patients’ judgements and decisions [11].
Ideally, we would like our patients to make an informed
decision about dialysis modality, i.e. attending to the ad-
vantages and disadvantages of both options, deliberating
about how they feel about the consequences of the options
and making a choice based on a trade-off between these
evaluations [12]. From the decision sciences, we know that
people make decisions by consciously attending to all the
details of the choice and applying this type of systematic
processing strategy. However, people also make decisions
using quick intuitive strategies which focus on only a sub-
section of the choice details and this is known as heuristic
processing [13]. We all use both heuristic and systematic
strategies to make decisions. Asking someone whose opin-
ion we trust is a quicker and easier way to make a decision
and usually results in a satisfactory choice; evaluating all
the pros and cons is time consuming and stressful but en-
sures we anticipate more fully the consequences of our
choice [11, 14]. Patients with chronic kidney disease make
dialysis choices in the same way as they make everyday
decisions. Patients making decisions unaided are more
likely to make treatment choices using a heuristic strategy
[13, 15, 16].
The use of patient narratives about illness and/or treat-
ment experience is commonplace, e.g. patient support
groups, information leaflets, and adverts. Patients find these
experiences useful when making sense of, and coping with,
their illness. More recently, service providers have used
these patient stories as a package to deliver information
about treatment options [17]. The belief is that delivering
information within a patient story makes it more accessible
to patients, so increasing the likelihood of patients being
informed about their illness and treatment choices. How-
ever, providing information in the form of a patient story is
likely to encourage patients to employ a heuristic strategy
when processing the information [17]. A recent review
found information presented in the first person (e.g. the
patient’s account) was more than twice as likely to influ-
ence other people’s choices than information presented in
the third person (e.g. the doctor talking about the patient’s
A.E. Winterbottom et al.
experience) [17]. Applied to the chronic kidney disease
context, this would mean patients new to dialysis modality
decisions making their choice by judging another patient
and/or their experience rather than engaging systematically
with the extensive dialysis modality information provided
by services.
Several techniques may reduce the effects of the biases
inherent in the way information is presented and encourage
patients to make treatment choices using a systematic or
informed decision making strategy [12]. For example, en-
suring balanced information about all treatment options,
using figures to illustrate the decision options and conse-
quences, using tables to summarize decision options and
attributes, presenting risk figures as natural frequencies
rather than as odds or verbal descriptors, using questions
to elicit patient values about consequences and so on [11,
18–20]. Currently, there is little evidence, indicating which
of these techniques, and/or a combination of techniques, is
Downloaded from http://ndt.oxfordjournals.org/ by Gabriela Elena Voicu on March 27, 2012
most effective in helping patients who are new to the de-
cision making process for dialysis modality to make an
informed choice.
In this paper, we present two complementary studies
conducted with the primary aim of investigating the impact
of another’s dialysis story on the new patient’s decision
about dialysis modality. Given the widespread use of
patient stories in current renal practise, we carried out these
proof of concept studies within a general population mak-
ing hypothetical dialysis modality choices and so testing
four hypotheses. The first study’s objective was to inves-
tigate Hypothesis 1—whether presenting dialysis treatment
information in either a patient or doctor story biases par-
ticipants’ decisions. We hypothesized that participants
would find the patient more persuasive than the doctor’s
story. In addition, we tested Hypothesis 2—whether video-
taped information was more persuasive than written infor-
mation and Hypothesis 3—whether the storyteller’s sex
affected participants’ choices. The second study’s objective
was to replicate the findings of Study 1 and to test Hypoth-
esis 4—the effectiveness of a decision attribute table to
enable participants’ to make more informed choices. We
hypothesized that a table summarizing the risks and bene-
fits of the two dialysis options by their attributes would
reduce the cognitive demands on participants to remember
the details and reduce the biasing effect of the stories on
decision making.
Materials and methods
Both studies used similar experimental designs and methods delivered via
the Internet to answer the above hypotheses. These are presented in two
separate sections.
Study 1
Design. An experimental study with nine conditions arranged in a 2 3
2 3 2 mixed group design plus a control group was employed. The
independent between group variables were format of the information (writ-
ten or videotaped), source of information (patient, doctor) and order of
presenting the information [haemodialysis (HD) first or continuous cycling
peritoneal dialysis (CCPD) first]. Participants were randomly allocated to
conditions by a computer program. Figure 1 summarizes the structure of
the study.
Those randomized to the control group were provided with only basic
information and no stories. In the eight experimental conditions, partic-
ipants read two short paragraphs containing basic information about HD
Patient stories bias others’ dialysis choices
and CCPD (Figure 2) and also viewed two stories, one presented by a
patient and another by a doctor. One of the stories contained information
about HD and the other CCPD. One story was presented by a male actor
and the other by a female actor.
In half of the experimental conditions (n ¼ 4), stories were presented as
video clips, the remaining groups (n ¼ 4) were presented as written stories.
The order of the presentation of the stories was varied; either the patient or
doctor presented the information first. Who presented the information was
also varied, approximately half the participants viewed the doctor pre-
sented information about HD (and the patient presented information about
CCPD), while others viewed the patient presented information about
HD (and the doctor presented information about CCPD; see Figure 1).
Sample. There was insufficient evidence from previous research to cal-
culate effect size. Power analysis to determine sample size was therefore
based on average estimates of effect size, power and statistical significance
(alpha; 21). Sample size was determined using software (G*Power version
3.0.10). Effect size, the smallest difference expected to be detected by the
study, was estimated to be a moderate size for F and was calculated at r ¼
0.25 [21]. Power, the ability of a statistical test to detect a statistically
significant effect where there is an effect, was set at 0.8 [22]. Alpha, the
statistical significance level, was set at 0.05 [22]. As sample size was an a
priori calculation it was not possible to know the degrees of freedom. This
was estimated as being 45 (approximate numbers in each group—5).
Required sample size was calculated using G*power as a total of 492
across the eight conditions.
On screen, read information about
renal failure and dialysis
n=647 – complete data sets
Continue Complete questionnaires; end of study
n=578 (Control group)
n=69
Read Transcripts of Watch videos of
1. Male or Female 1. Male or Female
2. Doctor or patient 2. Doctor or patient
3. Supporting HD or CCPD 3. Supporting HD or CCPD
Followed by corollary Followed by corollary
(i.e. reverse of 1, 2 & 3) (i.e. reverse of 1, 2 & 3)
n=293 n=285
Complete questionnaires; end of
study (Study group)
n=578
Fig. 1. Procedure for Study 1.
Haemodialysis is carried out three times a week in hospital. Nurses supervise the
treatment. Each treatment lasts four hours. To have the treatment you need
an operation. You will need a local anaesthetic. You will be awake during the
procedure. The operation connects a vein and an artery in your arm to create a
fistula. Blood is removed from the fistula and passed round a dialysis machine.
This cleans and removes excess water and waste products from your body. The
blood is then returned to the body.
Continuous cycling peritoneal dialysis (CCPD) is done each night at home. You
carry it out yourself. Each treatment lasts eight hours. To have the treatment you
need an operation. You will need a general anaesthetic. You will be asleep
during the procedure. The operation means having a tube put in your abdomen
called a catheter. The space around your organs in your abdomen is filled with
fluid. This removes excess water and waste products from your body. The fluid is
then drained out of the body.
Fig. 2. Example of basic treatment information provided to all participants.
327
As our understanding of the biasing effect of narrative information or
stories is in its infancy, we felt it unethical to manipulate established
patient resources in a renal population without evidence. The sample,
therefore, consisted of 784 undergraduate and postgraduate students and
staff, exceeding the number required by the sample size calculation. These
were recruited from nine Universities, Psychology departments at every
UK University and two email discussion groups for postgraduate Psychol-
ogy students.
Development of study materials. Four hypothetical decision scenarios
were developed to describe the two dialysis treatments: HD and CCPD,
from the perspective of both doctor and patient (Figure 3). The treatment
information was based on previous research examining the quality of renal
patient information [7], the experience of patients who were in the process
of making/or had recently made their treatment decisions (Winterbottom
A, Bekker HL, Conner M, Mooney A. unpublished results), and guidance
from a renal consultant (A.M.). The story scripts were designed to provide
realistic treatment information comparable to that provided in practise. The
scripts were developed, so the content was of a similar quality in terms of
the information they contained and language used. Although no specific
guidelines exist as to a suitable approach for analyzing the content of text
in these terms, steps were undertaken to ensure that each of the stories was
Downloaded from http://ndt.oxfordjournals.org/ by Gabriela Elena Voicu on March 27, 2012
balanced in its content, evaluated to see if the information was ‘framed’ by
the narrator, assessed the text for its comprehension using the Flesch read-
ing ease score, which is a validated measure of the ease with which a piece
of writing can be understood [23] and matched for the number of words
they contained. Furthermore, a pilot study assessed the stories suitability
for inclusion in the study based on their perceived quality, clarity and
persuasive content. A convenience sample of 24 postgraduate students
completed measures to assess the quality and similarity of the information
between transcripts. No differences were observed suggesting that they
were perceived to be of good quality and contain comparative information.
Measures. Participants were asked to make a hypothetical treatment
choice by indicating on a 20- point Likert scale to measure the degree to
which participants agreed with statements provided; in this study, a score
of 0 indicated ‘definitely wanted to choose HD’ and 20 indicated ‘defi-
nitely wanted to choose CCPD’ and 10 indicated indecision. Participants
completed a questionnaire with several measures of psychological param-
eters that could be relevant to the decision making process in this scenario
including: decisional conflict scale [24]; regret [25]; knowledge; risk per-
ception; need for cognition [26] and attitudes etc. The results from these
measures are reported elsewhere (Winterbottom A, Bekker HL, Conner M,
Mooney A. unpublished results).
Procedure. Ethical approval was granted by the Institute of Psycholog-
ical Sciences, University of Leeds, UK Research Ethics committee in
2007. Emails were sent out to key contacts (e.g. school secretaries) in
University departments at the beginning of the University term (October
2007) to target new and returning undergraduate and postgraduate students
Information about haemodialysis presented by a Doctor
My name is Dr White, I am 42 years old, and I have been a kidney consultant for
about 7 years now. As a kidney Doctor I have talked to many patients about the
type of dialysis treatment they wish to choose. Patients tend to read a lot of
information before they make their decision and often have the opportunity to talk to
other patients who are already receiving dialysis. Haemodialysis is a treatment that
is carried out three times a week in the hospital. Patients often find this treatment
most convenient for them as they do not have to have any dialysis equipment at
home. Having the treatment in hospital means that patients are not responsible for
carrying out the treatment themselves. Patients who have haemodialysis often
report that they like having the support of a nurse when they have their treatment.
Many of them also enjoy meeting with other kidney patients when they go for their
dialysis. It is often possible to arrange the treatment at a time of day that is suitable
for the patient. This means there is less disruption to a patient’s everyday routine. I
have talked to many patients who have been able to have an annual holiday. The
hospital can arrange for dialysis to take place at a local hospital abroad. Many
patients report they feel much better for starting treatment and feel positive about
the future.
Fig. 3. Example of information presented in each condition.
328
at nine Universities. Participants completed the study online. Participants
were provided with information about the study in an introductory email.
Participants clicked on a hyperlink and were connected to further informa-
tion and instructions. All participants who provided their email addresses
were entered in to a prize draw, with the chance of winning £150. Clicking
on the submit button at the bottom of the instruction page, indicated
consent to take part in the study. Clicking on this button also allocated
the participant to one of the experimental conditions. Participants then
worked on the study individually. Participants were not given the oppor-
tunity to save their work and were therefore required to complete the whole
of the study in one sitting. Questionnaires included in the study are avail-
able from the first author.
Analysis. Between groups analysis of variance was conducted to inves-
tigate differences in treatment choice based on whether the patient or
doctor presented the information. Post hoc tests were carried out to inves-
tigate differences between variables where the means were significantly
different.
Study 2—Method
Study 1 was modified in three ways:
A.E. Winterbottom et al.
All participants who made a hypothetical treatment choice
are included in these analyses (Study 1, n ¼ 578; Study 2,
n ¼ 1116). The majority of participants in both studies
were female, comparable in age and described their ethnic-
ity as ‘white’. Details of both studies were sent out to
similar student populations but only 5% per cent reported
completing both studies. See Table 2 for full sample
characteristics.
Differences in sample characteristics by those (not)
completing the study
The data set in both studies was examined to see if there
were any differences in the demographic characteristics of
those who completed the study and made a hypothetical
choice (n ¼ 1694) and those that did not (n ¼ 812). Stat-
istical analysis indicated that although non-completers
were significantly younger (M ¼ 22.8, SD ¼ 10.3) than

Downloaded from http://ndt.oxfordjournals.org/ by Gabriela Elena Voicu on March 27, 2012

(1) A decision attribute table was developed and presented before or after those who completed the questionnaire (M ¼ 24.7, SD ¼
the treatment information or not at all. 8.1); [t(765) ¼ 2.28, P ¼ 0.01], the magnitude of the differ-
(2) Sex of the narrator was confounded in Study 1. One story was pre- ence in means was very small (g2 ¼ 0.01) and indicates
sented by a male and another by a female actor. In Study 2, the two those who did not complete the questionnaire were
stories were presented either by two males, two females or a male and on average 1.9 years younger than those completing it.
a female narrator.
Statistical analysis revealed no difference in sex [v2(1,
(3) Stories were presented in videotaped format only; Study 1 found no
main effect for format influencing decision making.
n ¼ 767) ¼ 0.00, P ¼ 0.97] and ethnicity (Fisher’s exact
test, P ¼ 0.26) between completers and non-completers in
either study, but in Study 2, completers were significantly
Design. The experiment contained 48 conditions arranged in a 3 (deci-
sion table—no decision table, before story, after story) 3 2 (order of
presenter of information—patient first, doctor first) 3 2 (who presented
the information—doctor presents HD, Patient presents HD) 3 2 (sex of
the patient—male, female) 3 2 (sex of the doctor—male, female) be- Participants recruited
tween groups design. A summary of the study structure is presented in
n=1722
Figure 4.

Development of study materials. In addition to Study 1 materials, a

decision attribute table (see Table 1) was developed from renal patient
information and in consultation with a renal consultant (A.M.) and expert On screen, read information about renal failure
in patient decision aids (H.L.B.). The decision attribute table included all and dialysis
of the treatment information in an alternative-by-attribute design, which
meant that the pros and cons of the information could be viewed side-by-
side at a glance. This was developed as a resource for the participants to
Review decision attribute
mimic a simple patient decision aid and was essential to test our hypothesis
that provision of such information reduces the likelihood of a heuristic table
decision being made about dialysis modality. n=387

Sample. Sample size calculations were determined by duplicating the

method described in Study 1. Sample size was determined using software
(G*Power version 3.0.10). Effect size (F) was estimated to be a medium Watch video of
size at 0.25 [22]. Power was set at 0.8 [23]. The level for statistical sig- 1. Male or Female
See no decision attribute
nificance (alpha) was set at 0.05 [23]. As sample size was an a priori 2. Doctor or patient
calculation, it was not possible to know the degrees of freedom. This table
3. Supporting HD or CCPD
was estimated as being 45 (approximate numbers in each). Required sam- n=377
Followed by corollary
ple size was calculated using G*power as a total sample of 495 across 48
(i.e. reverse of 1, 2 & 3)
conditions.
n=1116

Measures, procedure and analysis. The same measures, procedure and

analysis plan were adhered to as in Study 1, with the exception that more
Universities were contacted (n ¼ 25) to obtain the required sample size. Review decision attribute
table
n=352

Results
Complete questionnaires; end of study

Sample characteristics (Study group)

n=1116
Not all participants who began the study completed it
(Study 1, 578/784 (74%); Study 2, 1116/1722 (65%)). Fig. 4. Procedure for Study 2.
Patient stories bias others’ dialysis choices 329
Table 1. A decision table comparing HD and CCPD

HD CCPD

Life expectancy No difference between HD and CCPD No difference between HD and CCPD
Location of treatment In hospital, 3 times a week At home, every night (CCPD)
Length of treatment 4 h each session 8 h a night (CCPD)
Type of operation ‘Fistula’ created in the arm. 90% of patients have this Catheter put in the abdomen. 1% have this done under
done under local anaesthetic; 10% under general local anaesthetic; 99% under general anaesthetic.
anaesthetic.
Complications Blood infection, blood clots or narrowing in the fistula Infection at site of catheter, abdominal infection
Support by health professionals Nursing staff carry out the treatment in hospital 24-h telephone support and regular visits to your home
Fluid restriction Patients need to be fluid restricted Patients need to be fluid restricted
Dietary restrictions Some foods to be avoided Some foods to be avoided

Table 2. Participant characteristics (n ¼ 1694) 20

CCPD
Study 1 Study 2
Demographics 15
n % n %

Downloaded from http://ndt.oxfordjournals.org/ by Gabriela Elena Voicu on March 27, 2012

10.8
9.7*
Treatment
choice 10
Female 573 73.1 723 65
Age 24.8 (m) 7.8 (SD) 23.5 (m) 7.1 (SD)
Ethnicity—white 663 84.6 958 86 5
Education level—degree n/a n/a 553 50
or more HD

0
Patient presents PD second Patient presents HD second

better educated than non-completers. No such effect was Order of presentation

seen in Study 1.
Treatment choice for participants in the control group
For those who did not view stories (control group), there
was no significant difference in treatment choice. Forty-
nine per cent of participants chose HD and 51% chose
CCPD. For other conditions, choice varied between 44
and 56%.
Hypothesis 1—patients’ stories are more persuasive than
doctors’ stories
This hypothesis was supported. In Study 1, the hypothesis
that first person narratives (patients’ stories) are more per-
suasive than third person narrative (doctors’ stories) was
partially confirmed; there was an interaction between the
order of presentation of the information and who presented
the information on treatment choice [F(1,482) ¼ 10.16,
P ¼ 0.01; partial g2 ¼ 0.02; Figure 5]. Post hoc analyses
revealed that when the doctor presented the information
first, participants were more likely to choose the patient
option which was presented second. The converse, however,
was not true, when the doctor presented the information
second, participants remained undecided in their choice.
No other main or interaction effects approached significance.
In Study 2, a significant main effect was confirmed for
whether a doctor or a patient presented the information
about HD and CCPD on treatment choice [F(1,1022) ¼
4.63, P < 0.03; partial g2 ¼ 0.01]. An examination of the
means indicated that when the patient presented informa-
tion about CCPD, participants were more likely to choose
CCPD; when the Patient presented information about HD,
participants were more likely to choose HD (Figure 6). No
other main or interaction effects approached significance.
Fig. 5. Participants are more likely to choose the dialysis modality sup-
ported by the patient if the patient presents their information after the
doctor (Study 1). Participants received information as described and were
asked to record their choice on a Likert Scale (0 ¼ definitely choose HD,
20 ¼ definitely choose CCPD). In Study 1, a recency effect was observed
with participants more likely to choose the modality described by the
patient only if the patient presented second (*P ¼ 0.04).
Hypothesis 2—videotaped information is more
persuasive than written information
The hypothesis that videotaped information is more persua-
sive than written information was not supported (Study 1).
There was no influence for mode of delivery (written/
videotaped) information on treatment choice [F(1,482) ¼
0.53, P ¼ 0.47].
Hypothesis 3—sex of storyteller will affect dialysis
modality decision
The hypothesis that the sex of the storyteller would affect
dialysis decision making was not supported (Study 1).
There was neither an effect of the sex of the storyteller
per se [F(1,482)¼0.47, P ¼ 0.49], or for combination of
sexes of narrator and participant [F(1,482) ¼ 2.63, P ¼
0.11].
Hypothesis 4—decision attribute table will increase
likelihood of informed decision making
The hypothesis that inclusion of a decision attribute table
would enhance informed decision making (Study 2) was
not supported. There was no evidence to suggest that in-
clusion of a decision attribute table made the patient stories
less persuasive [F(2,1022) ¼ 0.13, P ¼ 0.88].
330
Fig. 6. Participants are more likely to choose the dialysis modality sup-
ported by the patient than that presented by the doctor (Study 2). Partic-
ipants received information as described and were asked to record their
choice on a Likert Scale (1 ¼ definitely choose HD, 20 ¼ definitely choose
CCPD). In Study 2, confounding variables of the sex of the presenter were
lost and additional testing of a decision attribute table was undertaken. In
this study, the recency effect was lost, and participants were more likely to
choose the modality described by the patient than that described by the
doctor (*P ¼ 0.01).
Discussion
This paper presents findings from two large carefully de-
signed complementary experimental studies, investigating
the influence of presenting dialysis modality information in
different ways on participants’ hypothetical choices. The
four main findings from the results are: first, using patient
stories of their dialysis experience (first person) as a way to
deliver information about treatment options biases partic-
ipants choices in a way that doctors talking about the ex-
perience of their patients (third person) does not. While
doctors are likely to present both options fairly, seeing a
patient talk about just one option is likely to influence the
patient’s decision. Second, delivering stories in written and
video format are equally influential in participants’ deci-
sion making about dialysis modality; third, the sex of the
narrator does not influence participant decision making;
fourth, the addition of a decision attribute table alone does
not ameliorate the biasing effect of patient stories and en-
courage participants to make more informed dialysis mo-
dality choices. These findings suggest participants are not
evaluating the details provided about dialysis options and
their consequences in order to inform their dialysis decision
but are making it by referring to another patient’s dialysis
choice. It is, therefore, unsurprising that a resource summa-
rizing information about dialysis options and attributes did
little to support participants’ decision making; participants
were not using a decision making strategy that required
conscious and systematic evaluation of details in order to
choose a treatment.
The results presented in this paper contribute to our
understanding of the complexity of modality choice among
end-stage renal disease patients. Although the delivery of
satisfactory pre-dialysis education is the subject of national
and international guidelines [4, 5], there continues to exist
wide variation in uptake of modalities geographically and
temporally. Our studies have evaluated aspects of the qual-
ity of patient information. However, it must be acknowl-
edged that this forms only part of effective delivery of
satisfactory preparation for dialysis [27] and effective de-
livery of patient information is also important, especially
A.E. Winterbottom et al.
among those presenting for dialysis late, where choice
might be particularly limited [27].
The strength, and also the limitation, of this research is
that it is an experimental study carried out with participants
who do not have chronic kidney disease nor are under
regular review by a supervising physician. The experimen-
tal design meant we were able to manipulate how the in-
formation about dialysis modality was presented in a
controlled environment and assesses its influence on par-
ticipants’ choices [31]. This experiment has meant we have
robust data demonstrating that (i) participants’ treatment
decisions are biased by another’s choice and (ii) informa-
tion only interventions are unlikely to engage participants
in informed decision making. What is less clear is the ex-
tent to which patients with chronic kidney disease making
dialysis modality decisions are biased by another patient’s
choice. Patients’ values and judgements may differ from
the general population [32, 33], despite patients having the
Downloaded from http://ndt.oxfordjournals.org/ by Gabriela Elena Voicu on March 27, 2012
same infrastructure with which to make decisions, i.e. heu-
ristic and systematic information processing strategies [34,
35]. Participants in this study were encouraged to complete
a free text section of the questionnaire, wherein their
thoughts and reasons for choosing dialysis modality
matched closely those reported by real patients on dialysis
in another study we conducted (Winterbottom A, Bekker
HL, Conner M, Mooney A. unpublished results). Evidence
also indicates that renal patients’ choices are influenced by
their peers [6, 36]. Although the extent to which people
make biased choices may vary dependant upon the conse-
quences of the decision, evidence suggests that this effect is
present regardless of whether it is a real or hypothetical
decision [37].
It is unclear whether patients with chronic kidney disease
having to choose a dialysis modality in order to extend life
encourages the use of heuristic and/or systematic strategies
and/or makes them more, less or equally persuaded by
another’s treatment choice. Almost equal numbers of par-
ticipants chose either HD or CCPD when no narratives
were present (control). This indicates that variations in up-
take rates are influenced by factors other than patients’
preferences. While some of these are known to services,
others are more hidden. If we want to encourage patients
to make an informed decision in accordance with their own
preferences, we need to address these factors.
We conclude that providing accessible information to
patients with chronic kidney disease is not sufficient to
enable them to make informed dialysis modality choices.
Furthermore, the way information is presented to patient’s
influences how they make their decision and can bias the
choices they make. Using a patient’s story about their dial-
ysis experience to deliver information about chronic kidney
disease and treatment choices will influence a patients’
choice in both video and written forms. Furthermore, the
failure of our simple decision attribute table to overcome
these biases emphasize that it is unclear what type of inter-
vention could ameliorate the influence of another patient’s
story on treatment choices. Services need to consider that
by providing opportunities for patients considering dialysis
modality to speak with others on dialysis may be discour-
aging them from making informed dialysis modality
decisions.
Patient stories bias others’ dialysis choices
Acknowledgements. The authors acknowledge the Economic and Social
Research Council (ESRC), who, in association with Baxter’s Healthcare,
funded the research as part of a PhD awarded to the first author (A.W.).
Transparency declaration. The first author (A.W.) undertook these re-
search studies as part of a PhD funded by a Economic and Social Research
Council (ESRC) CASE studentship and an unrestricted educational grant
from Baxter Healthcare Ltd. None of the authors have a financial interest
in the information provided in the manuscript.
Conflict of interest statement. None declared.
References
1. Renal Registry. Renal Registry Twelfth Annual Report. Bristol, UK:
The Renal Association; 2009
2. Anzdata. Australia and New Zealand Transplant and Dialysis Report.
The 32nd Annual Report. Anzdata Registry, 2009
3. ERA-EDTA Registry. ERA-EDTA Registry Annual Report. Amster-
dam, The Netherlands: Academic Medical Center, Department of
Medical Informatics, 2010
4. Department of Health. The National Services Framework for Renal
Services: Part One Dialysis and Transplantation. London, UK: The
Stationary Office, 2004
5. Covic A, Bammens B, Lobbedez T et al. Educating end-stage renal
disease patients on dialysis modality selection: a clinical advice from
the European Renal Best Practice (ERBP) Advisory Board. NDT Plus
2010; 3: 225–233
6. Morton RL, Tong A, Howard K et al. The views of patients and carers in
treatment decision making for chronic kidney disease: systematic review
and thematic synthesis of qualitative studies. Br Med J 2010; 340: c112
7. Winterbottom A, Conner M, Mooney A et al. Evaluating the quality
of patient information provided by renal units across the UK. Nephrol
Dial Transplant 2007; 22: 2291–2296
8. Elwyn G, O’Connor A, Stacey D et al. Developing a quality criteria
framework for patient decision aids: online international Delphi
consensus process. BMJ 2006; 333: 417
9. Mooney A, Winterbottom A, Bekker HL. Letter to the Editor: the
importance of expert education in enabling informed, activated
patients. Kidney Int 2009; 75: 1116–1117
10. Bekker HL, Winterbottom A, Mooney A. Patient information and
decision making processes (editorial). Br J Ren Med 2009; 14: 28–30
11. Bekker HL. The loss of reason in decision aid research: the patient as
decision maker not collaborator. Patient Educ Couns 2010; 78: 357–364
12. Bekker HL. Chpt. 7 using decision making theory to inform clinical
practice. In: Edwards A, Elwyn G, eds. Shared Decision-Making in
Health Care: Achieving Evidence-Based Patient Choice. 2nd edn.
Oxford University Press; 2009: 45–52
13. Payne J, Bettman J. Walking with the scarecrow: the information-
processing approach to decision research. In: Koehler D, Harvey N
(eds). Blackwell Handbook of Judgment and Decision Making. Mal-
den MA: Blackwell Publishing; 2007: 110–132
14. Larrick RP. Chapter 16: Debiasing. In: Koehler D, Harvey N (eds).
Blackwell Handbook of Judgment and Decision Making. Malden,
MA: Blackwell Publishing; 2007: 316–338
15. Baron J. Thinking and Deciding. 3rd edn. Cambridge, UK: Cambridge
University Press; 2008
16. Frisch D, Clemen RT. Beyond expected utility: rethinking behavioral
decision research. Psychol Bull 1994; 116: 46–54
331
17. Winterbottom A, Bekker HL, Conner MT et al. Does narrative in-
formation bias individual’s decision making? A systematic review.
Soc Sci Med 2008; 67: 2079–2088
18. Lipkus IM. Numeric, verbal, and visual formats of conveying health
risks: suggested best practices and future recommendations. Med
Decis Making 2007; 27: 696–713
19. O’Connor AM, Bennett CL, Stacey D et al. Decision aids for people
facing health treatment or screening decisions. Cochrane Database
Syst Rev 2009; , Issue 3, Art. No.: CD001431
20. Abhyankar P, Bekker HL, Summers B et al. Does framing trial par-
ticipation choices in different ways affect informed decision making:
an experimental study? 4th International Shared Decision Making
Abstracts. 2007
21. Connelly LM. Research consideration: power analysis and effect size.
MedSurgery Nurs 2008; 17: 35–40
22. Cohen J. A power primer. Psychol Bull 1992; 112: 155–159
23. Flesch R. A new readability yardstick. J Appl Psychol 1948; 32:
221–233
24. O’Connor AM. Validation of a decisional conflict scale. Med Decis
Making 1995; 15: 25–30
25. Brehaut JC, O’Connor AM, Wood TJ et al. Validation of a decision
Downloaded from http://ndt.oxfordjournals.org/ by Gabriela Elena Voicu on March 27, 2012
regret scale. Med Decis Making 2003; 23: 281–292
26. Cacioppo JT, Petty RE. The need for cognition. J Pers Soc Psychol
1982; 42: 116–131
27. Mehrotra R, Marsh D, Vonesh E et al. Patient education and access of
ESRD patients to renal replacement therapies beyond in-center hemo-
dialysis. Kidney Int 2005; 68: 378–390
28. Marrón B, Craver L, Remón C et al. ‘‘Reality and desire’’ in the
care of advanced chronic kidney disease. NDT Plus 2010; 3:
431–435
29. Marrón B, Martı́nez Ocaña JC, Salgueira M et al. Analysis of patient
flow into dialysis: role of education in choice of dialysis modality.
Perit Dial Int 2005; 25 (Suppl 3): S56–S59
30. Julián JC. Información y proceso de decisión en el tratamiento susti-
tutivo renal (TSR): el punto de vista del paciente renal. En ‘‘Libro
Blanco de la Diálisis Peritoneal en España’’ pp. 16–25
31. Robson C. Experimental Design and Statistics in Psychology. 3rd
edn. London, UK: The Penguin Group, 1994
32. Stiggelbout A. Assessing patients’ preferences. In: Chapman G,
Sonnenberg F, eds. Decision Making in Health Care: Theory, Psy-
chology and Applications. Cambridge, UK: Cambridge University
Press; 2000, pp. 289–312
33. Jansen SJT, Stiggelbout AM, Wakker PP et al. Unstable preferences:
a shift in valuation or an effect of the elicitation procedure? Med Decis
Making 2000; 20: 62–71
34. Chapman GB. The psychology of medical decision making. In:
Kohler DJ, Harvey N, eds. Blackwell Handbook of Judgment and
Decision Making. Malden, MA: Blackwell Publishing; 2007,
pp. 585–603
35. Wiseman DB, Levin IP. Comparing risky decision making under
conditions of real and hypothetical consequences. Organ Behav
Hum Decis Process 1996; 66: 241–250
36. Morton RL, Howard K, Webster AC et al. Patient information about
options for treatment (PINOT): a prospective national study of infor-
mation given to CKD Stage 5 patients. NDT Plus 2010; ; doi: 10.1093/
ndt/gfq555
37. Kuhberger A, Schulte-Mecklenbeck M, Perner P. Framing decisions:
hypothetical and real. Organ Behav Hum Decis Process 2002; 89:
1162–1175
Received for publication: 21.9.10; Accepted in revised form: 14.4.11