Blackwell Science, LtdOxford, UKADDAddiction0965-2140 2006 Society for the Study of Addiction

101
Original Article
Helping smokers to choose cessation methods
Marc C. Willemsen
et al.

RESEARCH REPORT

doi:10.1111/j.1360-0443.2005.01349.x

Helping smokers to decide on the use of efficacious

smoking cessation methods: a randomized controlled
trial of a decision aid
Marc C. Willemsen, Marieke Wiebing, Andre van Emst & Grieto Zeeman
STIVORO for a Smoke-free Future, The Hague, the Netherlands

ABSTRACT
Aims Most smokers attempt to stop smoking without using help. We evaluated the efficacy of a decision aid to
motivate quitters to use efficacious treatment. Setting and participants A total of 1014 were recruited from a
convenience sample of 3391 smokers who intended to quit smoking within 6 months. Design and intervention
Smokers were assigned randomly to either receive the decision aid or no intervention. The decision aid was expected
to motivate quitters to use efficacious cessation methods and contained neutral information on treatment methods, distinguishing between efficacious and non-efficacious treatments. Measurements Baseline questionnaire and follow-ups were used 2 weeks and 6 months after the start of the intervention. Findings The decision aid increased
knowledge of cessation methods and induced a more positive attitude towards these methods. Furthermore, 45%
reported increased confidence about being able to quit and 43% said it helped them to choose between treatments.
However, no clear effect on usage of treatment aids was found, but the intervention group had more quit attempts
(OR = 1.52, 95% CI 1.142.02) and higher point prevalence abstinence at 6-month follow-up (20.2% versus 13.6%;
OR = 1.51, 95% CI = 1.072.11). Conclusions An aid to help smokers decide to use efficacious treatment when
attempting to quit smoking had a positive effect on smoking cessation, while failing to increase the usage of efficacious
treatment. This finding lends support to the notion that the mere promotion of efficacious treatments for tobacco addiction might increase the number of quit attempts, irrespective of the actual usage of treatment.
Keywords

decision aid, randomized trial, smoking cessation, treatment.

Correspondence to: Marc C. Willemsen, STIVORO for a Smoke-free Future, PO Box 16070, 2500 BB, The Hague, the Netherlands,
E-mail: mwillemsen@stivoro.nl
Submitted 15 April 2005; initial review completed 22 September 2005; final version accepted 28 September 2005

RESEARCH REPORT
INTRODUCTION
Tobacco use is still one of the most widespread and
harmful of addictive behaviours. In the 25 countries of
the European Union alone, 656 000 people die every
year as a direct or indirect consequence of smoking,
making tobacco addiction the most important avoidable
cause of disease and premature death (European Commission 2004). Randomized trials have shown consistently that quitters who use evidence-based therapy
significantly improve their chance of success compared
to no therapy or placebo (Fiore et al. 2000). Among
these therapies are self-help methods, physician advice,
telephone counselling, cognitive behavioural therapy,

nicotine replacement therapy and non-nicotine medication such as bupropion.

Clinical guidelines for treating tobacco dependence
that describe the range of efficacious therapies have
recently become available in various developed countries
(Raw, McNeill & West 1998; National Health Committee
1999; Fiore et al. 2000). These guidelines recommend
that every smoking patient should be offered tobacco
dependence treatments. It is believed that increasing the
use of effective smoking cessation aids could, in principle,
have a substantial public health impact (West et al. 2005).
Nowadays, smokers who want to quit smoking can
choose from a variety of efficacious therapies. Indeed, the
use of treatments has increased, for example from 8% in

2006 The Authors. Journal compilation 2006 Society for the Study of Addiction

Addiction, 101, 441449

442

Marc C. Willemsen et al.

1986 to 20% in 1996 in the United States (Zhu et al.

2000), and from 18% in 1991 to 24% in 2000 in the
Netherlands (Willemsen, Wagena & van Schayck 2003).
Furthermore, recent data from the Community Intervention Trial for Smoking Cessation (COMMIT) trial show
that the prevalence in the United States of ever having
used assisted methods to quit increased from 45% to 60%
between 1993 and 2001 (Hyland et al. 2004). These
increases are probably caused by the fact that more stopsmoking methods became available in the 1990s, especially a wider range of pharmacological treatments. However, despite these increases, the current situation is still
that the number of smokers who attempt to quit on their
own far outweighs the number who use some form of
efficacious therapy. Effective treatments for tobacco
dependence are still widely underused (Cokkinides et al.
2005). For example, in the Netherlands 64% of smokers
who made a quit attempt in 2004 had not used any
method (Zeegers, Segaar & Willemsen 2005). Nicotine
therapy was used by only 13% of quitters (8% used the
gum) and only 4% used bupropion. Popular methods
were non-efficacious methods such as laser therapy (8%)
and the Allen Carr method (book or 1-day course; 6%).
Various strategies have been suggested to improve the
usage of efficacious cessation aids. One strategy is by
reimbursing treatment costs. Some recent studies showed
that this results in modest but significant improvements
in usage (Curry et al. 1998; Schauffler et al. 2001; West
et al. 2005), whereas others reported no benefit (Boyle
et al. 2002). In the United Kingdom, free formal smoking
cessation services have been made widely available to
smokers, although they are still much underused (Britton
2004). Another strategy is through intermediaries. For
example, family physicians can advise their smoking
patients to quit smoking and inform them about the most
beneficial treatment. However, many patients who smoke
still do not receive adequate advice from physicians on
tobacco cessation methods, despite the emergence of clinical guidelines (Cabana et al. 1999; Schnoll & Engstrom
2004). A third strategy is by offering cessation services in
an unsolicited manner, for example through direct marketing (Paul et al. 2004). Little is known about the efficacy of this strategy. A fourth strategy is by educating
smokers about the benefits of treatment. One reason why
few smokers use treatment may be that they are unaware
of the available range of treatments and do not know their
potential to increase success rates (Hammond et al.
2004). Recent data show that many smokers have misconceptions about the efficacy and safety of nicotine
replacement therapy (Etter & Perneger 2001; Bansal
et al. 2004). Smokers who were more knowledgeable
about the safety and efficacy of nicotine medications were
more likely to report ever having used them (Bansal et al.
2004).

To improve awareness of treatment choices and the

potential benefits of using these treatments, we developed
a decision aid called Starters Kit, which provides smokers with detailed information about the availability and
efficacy of a wide range of treatment methods and helps
them to make a choice. Although the field of consumer
decision support interventions has grown rapidly in the
last 20 years (Estabrooks et al. 2001), they have not yet
been applied to tobacco addiction treatment.
The efficacy of the decision aid was evaluated by
exposing smokers to the aid and comparing their reactions and behaviour change with those of a group receiving no intervention. We hypothesized that the decision
aid would increase smokers knowledge of smoking cessation therapies, make their attitude towards using them
more positive, increase their expectation of success when
using cessation methods and increase their intention to
use treatment in future quit attempts. Furthermore, we
expected that exposure to the decision aid would result
in more quit attempts and in greater use of evidencebased cessation methods among smokers making a quit
attempt, which in turn would result in higher success
rates of smoking cessation.
METHODS
Design and participants
We carried out a randomized controlled trial among participants drawn from an internet-based database of more
than 35 000 households. These respondents participate
in various studies by TNS NIPO, a large market research
company, and are used to examine, test and evaluate
commercial and non-commercial products at home, for
which they receive a financial reward. First, all households received a first screening questionnaire. This identified 3391 smokers who intended to quit smoking within
the next 6 months. These were sent a further screening
questionnaire through the internet to identify and recruit
participants into the study. The response to this questionnaire was 87% (2955 respondents). The result from this
questionnaire was that a number of potential participants were excluded. Excluded were participants who no
longer smoked (9.4%), those who no longer intended to
quit smoking within 6 months (46.4%), those who
reported they already knew the decision aid (in response
to a colour picture that was presented to them) (6.4%),
and those who were unable to or did not want to participate in an evaluation of an information kit that they
were told they would receive by mail within a few weeks
(without giving specifics about its content, nor stating
that it was about smoking cessation) (3.4%). Thus, a total
of 1014 respondents (34.4%) were enrolled into the
study. Of these, 500 were randomized to the experimental
and 514 to the control group. Participants in the experi-

2006 The Authors. Journal compilation 2006 Society for the Study of Addiction

Addiction, 101, 441449

Helping smokers to choose cessation methods

mental group received the decision aid by post. Participants in the control group received no intervention. The
fieldwork for the study was conducted in April/May 2003
(pre-test and first post-test) and October/November 2003
(post-test).
Intervention
The decision aid was designed to motivate smokers to
use an efficacious cessation method when making a
quit attempt and to help them make an informed decision about the treatment method to use. The decision
aid was an 8-cm (height) 23-cm (length) 23-cm
(width) cardboard box, with the logos of STIVORO and
the Dutch Cancer Foundation printed on its side. The
box was called Starters Kit and contained a number
of items. The first was a booklet describing all major
treatment methods available in the Netherlands. A
distinction was made between category A treatments
(i.e. self-help manual, computerized tailored advice, telephone counselling, group counselling, physician advice,
behavioural therapy, nicotine replacement therapy,
bupropion) and category B treatments (e.g. the Allen
Carr method, hypnosis, acupuncture and soft laser
therapy). This distinction was introduced by saying
Research has shown that some treatments work better
than others, but not all treatments have been thoroughly studied yet. Category A represented all forms of
treatment of which we now know for sure that they are
effective. The distinction between the A and B categories was in line with the Dutch clinical guidelines for
smoking cessation (van Weel et al. 2005). A short objective description was given of each treatment, using
neutral wording, supported with information on the
intensity of the treatment (number and frequency of
contacts) the type of contact (written, telephone, faceto-face), the length of the treatment period, the financial
costs and whether these could be reimbursed by health
insurers. Furthermore, the kit contained a video showing Dutch celebrities as well as unknown Dutch exsmokers who described how they had successfully quit
smoking using specific treatments. It was expected that
by observing these models, through a process of vicarious learning (Bandura 1986), smokers would learn
which cessation method could work best for them. The
video also featured a tobacco control expert giving independent and objective information on how various
treatment methods work and what may be expected
from them. Finally, the kit contained a number of samples of category A treatment methods, to function as a
cue to action for smokers to actually make the important step of applying for a cessation method. The following items were included for this purpose: a postcard
with which a smoking cessation self-help manual could
be ordered, a leaflet about how to apply for group coun-

443

selling, a brochure on nicotine gum, information on

how to apply for telephone counselling, a questionnaire
that smokers could complete and return in order to
receive computer-tailored advice on smoking cessation
and information on a relapse prevention programme
called after care. This programme consisted of seven emails or postcards with information and advice on how
to prevent relapse, which smokers could receive during
the first 3 months after their stated quit date. Smokers
could apply for this service by sending in a card or by
registering through the internet. The production costs
of the box was 3.59 (excluding VAT, with a circulation
of 200 000).
Measurements
The first follow-up measurement was conducted 2 weeks
after participants in the intervention condition had
received the decision aid. The second follow-up took place
6 months after they had received the intervention. All
measurements were conducted through the internet.
Five behavioural and six intermediary psychological outcome measures were used, and self-reported measures of
effects were also included.
Self-reported effects
Four items were included to measure self-perceived
changes in (1) attitude towards the use of tobacco dependence treatments (Did the starters kit make you think
more positive or more negative about the use of smoking
cessation aids or treatment?); (2) motivation to quit
smoking (Did the starters kit increase your motivation
to quit smoking?); (3) confidence about being able to
quit smoking (To what extent did the starters kit
increase or reduce your confidence in the success of your
next quit attempt?); and (4) the degree to which the quit
kit helped smokers make a choice between treatment
methods (To what extent did the starters kit help you
choose the most suitable treatment aid?). Answers
ranged from very helpful to not helpful at all, on a fivepoint scale. The first three items were scored on a fivepoint bipolar scale ranging from much more to much
less.
Behavioural outcomes
At both follow-up measurements, we assessed whether
the respondents had made a quit attempt since the first
measurement, whether they had used any of the A or B
category treatments for this attempt, and the 7-day point
prevalence (i.e. not having smoked in the past 7 days, not
even a puff). At the second follow-up, we also measured
continuous abstinence. Participants were assumed to be
continuously abstinent when they were non-smokers at
the 6-month follow-up (7-day point prevalence criterion), said that they had quit smoking in either April or

2006 The Authors. Journal compilation 2006 Society for the Study of Addiction

Addiction, 101, 441449

444

Marc C. Willemsen et al.

May 2003 and indicated that they had not smoked since
this quit attempt.
Intermediary psychological outcomes
Knowledge of treatments was measured by asking participants to select from a list of 17 cessation methods
whichif anythey knew. Three indices were constructed: one for the number of pharmacological products they knew (range 06), one for the number of other
A category treatments (range 06) and one for the
number of B category treatments (range 05). Attitude
towards the use of treatment was assessed by asking:
How do you feel about the use of smoking cessation aids
or about receiving support from professionals when trying to quit? [five-point scale from very negative (+2) to
very positive (2)]. We measured respondents expectation of success when using a treatment method by asking Do you think your chance of success would become
greater or smaller by using smoking cessation aids or
receiving support from professionals? [five-point scale
from much greater (+2) to much smaller (2)]. Selfefficacy expectation about being able to quit (Imagine
that you quit smoking. Do you expect you will be able
to refrain from smoking in every situation that may
occur?) was measured on a five-point scale from
certainly yes (+2) to certainly no (2) (Mudde, Kok &
Strecher 1995). General intention to use treatment in
future quit attempts was measured with the question
Would you use treatment if you quit smoking?
[certainly yes (+2) to certainly not (2)]. A more specific
intention to use specific treatment in the future was
measured by asking participants to select from a list of
17 cessation methods whichif anythey would use in
their next quit attempt. Three dichotomous (yes/no)
items were constructed indicating whether they
intended to use any of the pharmacological aids, any of
the other category A methods, or any of the B category
aids.
Potential confounders and effect modifiers
Several variables were included as additional possible
confounders and effect modifiers. The first group consisted of demographic variables: gender, age and socioeconomic status (SES). SES (five categories) was constructed by combining educational level and the (most
recent) profession of the head of the household. The second group related to smoking history: number of cigarettes smoked per day, nicotine dependency (as measured
with the time to first cigarette item) (Heatherton et al.
1991), number of quit attempts in the past, smoking status of partner and stage of change (contemplation or
preparation) (Prochaska & DiClemente 1983). The third
group consisted of psychological background variables
that were found in previous studies to predict successful

quitting: perceived social pressure to quit smoking (How

often do people around you say that you should quit
smoking?, measured on a four-point scale ranging from
very often to never) and self-evaluative consequences of
quitting (three items, for example If I quit smoking, I will
be pleased with myself , Cronbachs alpha = 0.73)
(Dijkstra & De Vries 2001).
Statistical analyses
Analysis of variance (ANOVA) and Pearsons 2 test were
used to test for baseline differences between the two study
groups (all tests two-sided). Differences in baseline scores
between respondents who participated in the entire
experiment and participants who dropped out were
assessed by means of logistic regression analysis, with
dropout (yes/no) as the dependent variable and study
group, smoking history, demographics and potential psychological confounders as predictors.
The effect of intervention on behavioural outcome
measures was examined with logistic regression analyses, with relevant baseline factors as covariates. Respondents who were lost to follow-up were included as
smokers (intention-to-treat procedure).
For each intermediary psychological outcome variable, we examined changes between pre-test and first
follow-up, both in the intervention and the control
group. The paired-sample t-test was used for knowledge,
attitude and success expectations (ordinal scales). The
McNemar test was used for changes in intention scores
(dichotomous variable). For these analyses, we selected
all respondents who provided complete data both at pretest and first post-test. To assess whether increases in psychological outcome variables between pre-test and first
follow-up differed significantly between intervention
groups, difference scores (first follow-up score minus
pre-test score) were constructed. These were compared
between the intervention and control groups by means
of t-test (ordinal variables) or 2 analysis (categorical
variables).
In order to examine whether specific characteristics of
the smokers modified the effect of treatment on quitting,
potential effect modifiers (demographic, smoking history
and psychological background variables) were examined
in a series of stepwise logistic regression analyses (step 1:
treatment and effect modifiers, step 2: interaction terms)
with behaviour changes (i.e. making a quit attempt and
being abstinent for 7 days at 6-month follow-up) as
dependent variables. Treatment and effect modifiers were
entered first, followed by interaction terms (treatment
effect). Interaction terms were then removed using a
backward eliminating procedure (P(removal) = 0.10). If
any interactions remained, the effect of treatment was
stratified to this variable in a logistic regression analysis
while controlling for relevant confounders.

2006 The Authors. Journal compilation 2006 Society for the Study of Addiction

Addiction, 101, 441449

Helping smokers to choose cessation methods

RESULTS
There were no significant differences between the two
study groups in the characteristics of participants
(Table 1). However, a baseline comparison between the
intervention and control groups in terms of potential psychological confounders revealed that respondents in the
intervention condition differed in two items from those
who were randomized to the control condition: they had
higher general self-efficacy expectations towards smoking cessation (mean scores were 0.0 versus 0.17;

Table 1 Demographic and smoking history characteristics of

the study population at the pre-test measurement.

Variable

Intervention group
(n = 500)

Control group
(n = 514)

Sex (%)
Male
Female

53.4
46.6

54.1
45.9

Age (years) (%)

1824
2534
3544
4554
5565
65+

4.2
22.2
31.6
25.0
11.6
5.4

3.1
22.6
30.4
25.5
14.2
4.3

Socio-economic status (%)

A (high)
Bb
Bo
C/D (low)

17.1
37.8
19.5
25.5

14.8
40.2
21.2
23.8

Daily smoker (%)

86.2

87.7

Number of cigarettes/day (%)

12
6.6
37
14.0
812
15.0
1317
15.4
1822
27.6
2327
10.0
> 28
11.4

6.2
14.4
14.6
17.1
26.3
12.6
8.8

Number of previous quit

attempts (mean, SD)

2.7 (1.44)

445

t = 2.50; P < 0.05) and were less likely to intend to use

any of the category A treatments in future quit attempts
(11.8% versus 15.6%; 2(1) = 3.04; P = 0.08). Because
these two variables were also associated with behaviour
change (quit attempt, 7-day point prevalence of quitting,
continuous abstinence), we included them as covariates
in all logistic regression analyses involving comparisons
between the intervention and control.
Of the respondents who participated in the baseline
measurement, 9.6% were lost to the first follow-up and
11.8% to the second follow-up (including non-response
to the first follow-up). Attrition analyses revealed that
loss to follow-up was not significantly (P < 0.05) associated with any of the baseline variables, except with the
experimental manipulation: participants in the intervention group were more likely to be missing at the first
follow-up (85% versus 96%), but not at the second
follow-up.
All participants in the intervention group said they
had received the decision aid; 86.1% reported to have
actually read the booklet, and 59.8% had watched the
video, while 7.5% had applied for the relapse prevention
programme.
Self-perceived effects
Fifty-six per cent said that the decision aid had given
them a more positive opinion about the use of cessation
treatment, while 9% had become more negative and
35% chose the neutral answer. In response to the question of whether the decision aid had increased their
motivation to quit smoking, 69% answered affirmatively,
29% neutral and 2% negatively. Self-reported confidence about being able to quit smoking increased as a
result of the decision aid for 45% of the respondents
(52% neutral, 3% became less confident). Finally, 43%
said that it had helped them to choose the most suitable
treatment aid (35% neutral, 22% said that it had not
helped them).
Behaviour change

2.7 (1.44)

Smoking status of partner (%)

No partner
20.8
Smoker
42.0
Ex-smoker
17.6
Never smoker
19.6

17.9
39.7
19.8
22.6

Stage of change (%)

Preparation
Contemplation

35.4
64.6

31.5
68.5

Time to first cigarette (%)

< 5 minutes
630 minutes
3160 minutes
> 60 minutes

16.0
38.8
16.2
29.0

16.7
38.1
17.9
27.2

No effects were found at the first follow-up. Table 2 shows

significant behaviour change between the baseline and
the second follow-up measurement. At the second followup, significant differences were observed between the two
experimental conditions in terms of having made a quit
attempt and 7-day abstinence. Logistic regression analyses were used to test whether these effects would hold
after correcting for the two confounders, showing that
those in the intervention group still had a greater
chance of making a quit attempt (OR = 1.52; 95%
CI = 1.142.02; P < 0.005) and being abstinent for
7 days (OR = 1.51; 95% CI = 1.072.11; P < 0.05). The
treatment effect on 7-day abstinence was modified by
stage of change (P < 0.001) in that exposure to the deci-

2006 The Authors. Journal compilation 2006 Society for the Study of Addiction

Addiction, 101, 441449

446

Marc C. Willemsen et al.

Table 2 Association between experimental condition and behavioural outcomes at 2-week follow-up and 6-month follow-up.
2-Week follow-up

Quit attempt
7-Day abstinence
Continuous abstinence

6-Month follow-up

Intervention
group (%)

Control
group (%)

Difference
(95% CI)

P-value

Intervention
group (%)

Control
group (%)

Difference
(95% CI)

P-value

14.0
8.6

11.9
6.4

2.1 (2.02.1)
2.2 (2.22.2)

NS
NS
NS

31.0
20.2
5.0

22.2
13.6
5.1

8.8 (3.414.2)
6.6 (2.011.2)
0.1 (1.92.1)

< 0.01
< 0.01
NS

Table 3 Self-reported use of treatment among respondents who made a quit attempt since the start of the experiment, measured at
the 6-month follow-up (proportions and 95% CI).

Type of treatment

Intervention
group
(n = 155)

Control
group
(n = 114)

Difference
(95% CI)

Any NRT
Bupropion
Any other category A method
Any category B method
Advice from family physician
No treatment used (willpower only)

13.5
7.1
9.0
13.5
3.9
36.8

7.9
10.5
10.5
14.0
8.8
42.1

5.6 (1.7 to 9.9)

3.4 (2.6 to 9.4)
1.5 (5.7 to 8.7)
0.5 (7.3 to 8.3)
4.9 (1.1 to 10.9)
5.3 (5.8 to 16.4)

Note: multiple answers were allowed.

sion aid had an effect among contemplators (OR = 2.62;

95% CI = 1.584.33) but not among smokers in the preparation stage (OR = 0.77; 95% CI = 0.471.26).

to use other category A treatments was significantly

increased in both groups.

Use of treatment aids

DISCUSSION

Table 3 shows which types of treatment were used by

smokers who had made a quit attempt. Those who
attempted to quit and those who did not make a quit
attempt did not differ significantly with respect to the
usage of treatment aids (two-sided 2 test). Analyses of
effect modification showed that exposure to the decision
aid resulted in a greater chance of making a quit attempt
among smokers in the contemplation stage (OR = 2.07;
95% CI = 1.362.96), but not among those in the preparation stage (OR = 1.01; 95% CI = 0.651.57).

The hypotheses were partly confirmed. As expected, the

decision aid was effective in increasing smokers knowledge of smoking cessation methods and in making their
attitude towards these methods more positive. Furthermore, 45% reported increased confidence in their own
ability to quit smoking and 43% said the kit had helped
them to make a choice between treatment options. In
view of these positive findings, it was surprising to find no
significant effect on actual usage of treatment aids. Exposure to the kit did not result in greater use of bupropion,
advice from a physician, nor other efficacious (category
A) treatments (nor of category B treatments, for that
matter). Nicotine replacement therapy (NRT) was used by
13.5 of quitters in the experimental group compared to
7.9% in the control group, but this difference was not significant, due possibly to insufficient statistical power.
Thus, it cannot be ruled out that this intervention
improves usage of NRT. More studies are needed involving
more smokers who make a quit attempt, to be able to conclude whether this type of intervention has a beneficial
effect on usage of efficacious cessation methods. For the
present study, however, it must be concluded that the
decision aid was not powerful enough to increase clear

Effects on psychological variables

Table 4 shows changes in intermediary psychological
variables from baseline measurement to 2-week followup. Knowledge about cessation treatments increased significantly in both study groups, and these increases were
significantly greater in the intervention group than the
control group. Attitude towards the use of cessation treatment became significantly more positive, and success
expectations significantly increased in the intervention
group, but not in the control group. General intention to
use cessation treatment did not show an increase in any
of the two experimental groups. However, the intention

2006 The Authors. Journal compilation 2006 Society for the Study of Addiction

Addiction, 101, 441449

Helping smokers to choose cessation methods

447

Table 4 Comparison of changes in psychological outcomes from pre-test (T0) to 2-week follow-up (T1) between intervention and
control groups.

Variables

Intervention (I) (n = 425)

Control (C) (n = 492)

T0

T0

T1

Comparison of
changes between
I and Ca

2.6 (1.4)
0.9 (1.0)
1.6 (1.3)
0.27 (0.8)
0.15 (1.0)
0.76 (0.8)

2.9 (1.3)***
1.081 (1.1)***
1.7 (1.2)
0.29 (0.8)
0.10 (1.0)
0.73 (0.8)

I >C***
I > C***
I > C***
I > C**
NS
I > C***

Number of treatment aids that respondents know (M, SD)

Pharmacological (06)
2.7 (1.5)
Other Cat A (06)
0.9 (1.1)
Cat B (05)
1.6 (1.2)
Attitude towards treatment (M, SD)
0.28 (0.9)
General self-efficacy expectation
0.01 (1.1)
Expectation of success when using
0.74 (0.9)
treatment (M, SD)
General intention to use treatment (M, SD)
0.36 1.2
Intention to use specific treatment
Pharmacological (%)
Other Cat A (%)
Cat B (%)

27.5
11.3
19.5

T1

3.7 (1.9)***
2.3 (2.0)***
2.4 (1.6)***
0.47 (0.9)***
0.01 (1.0)
0.95 (0.7)***
0.38 (1.2)
30.8
23.3***
21.2

0.36 (1.2)
7.0
15.9
15.9

0.29 (1.2)
31.1*
22.6***
15.7

NS
NS
NS
NS

*P < 0.05; **P < 0.01; ***P < 0.001. aResult of univariate analyses of variance to examine changes in ordinal variables and of logistic regression analyses
to examine changes in categorical variables (i.e. intention to use specific treatment).

overall usage of cessation methods. For example, smokers

were not educated on false perceptions about nicotine
replacement therapy, such as the notion that one might
become addicted to them, that they are harmful to ones
health or the idea that they are designed to make one feel
physically sick if smoking at the same time (Bansal et al.
2004). An intervention that included these components
might have a stronger effect on usage of NRT.
We found positive effects of the decision aid on smoking cessation. Smokers who received the decision aid had
a more than 50% greater chance of making a quit
attempt and of being 7-day abstinent after 6 months,
compared to a control group receiving no intervention.
The 7-day point prevalence quit rate was 20.2%, which is
quite high. A very interesting finding was that the effect
of treatment on making a quit attempt, as well as on 6month point prevalence quit rate, was restricted to smokers who were in the contemplation stage of change at
baseline. Those in the preparation stage did not benefit
from the decision aid. One explanation is that these quitters had already decided on the best way of quitting,
whereas those in the contemplation stage were still in the
process of gathering information about treatment. This is
in line with the trans-theoretical model, which states that
people in the contemplation stage are most open to
consciousness-raising techniques, i.e. they seek new
information and try to gain understanding and feed-back
about the problem behaviour (Prochaska, DiClemente &
Norcross 1992).
An important question remains as to why smokers in
the intervention condition were much more successful in
quitting smoking, because, compared to the control

group, they were not more likely to use efficacious

tobacco dependence treatment when attempting to quit
smoking. Apparently, exposure to the decision aid somehow made them decide to try to quit smoking on their
own. We may rule out the possibility that the decision aid
was used as a self-help method in and of itself, because the
decision aid itself did not contain any concrete self-help
information that smokers might have put into practice.
Furthermore, if the aid was used as a self-help quit
method one would expect that self-efficacy expectations
would increase after exposure to the decision aid, which
did not occur. Another explanation is that the information in the decision aid took away specific barriers to quit
smoking, more specifically any excuses participants
might have, such as not having enough information on
the costs, efficacy and availability of cessation treatment
methods. These were all presented to them in a very concise and clear manner, leaving very little room for this
excuse. If this explanation were true, the implication
would be that large-scale promotion of the existence of
efficacious smoking cessation methods, for example
through the mass media, could increase the number of
quit attempts, irrespective of the actual usage of these cessation methods. Possibly smokers learned from the decision aid that there is a wide range of treatments available,
and this new knowledge challenged them to try to quit,
knowing that if they should fail, they could still fall back
on more formal treatments. From the viewpoint of population impact, therefore, promoting the availability of
evidence-based cessation methods may be as important
as or perhaps even more important than the actual use of
these methods by smokers.

2006 The Authors. Journal compilation 2006 Society for the Study of Addiction

Addiction, 101, 441449

448

Marc C. Willemsen et al.

Another explanation for the observed effectiveness of

the decision aid is that the decision aid worked as a motivating agent in and of itself, more or less irrespective of
the specific content. A qualitative pre-test that was conducted among smokers to assess the likeability and the
communication potential of the decision aid showed that
the aid was seen by smokers as original, credible and
trustworthy (as it featured the Dutch Cancer Foundation
as one of the sources). The design of the box had a very
active (step-by-step) and no-nonsense feel to it, which
may have motivated smokers to take action. Furthermore, the packages sturdy nature (a thick cardboard
box), its attractive exterior and the fact that it contained a
video may have made it difficult for many smokers to simply throw it away. Thus, by being present in the smokers
home for some time, it may have reminded them constantly that they should quit smoking. Further research
into the efficacy of decision aids in smoking cessation
should study these heuristic aspects more explicitly,
which may account partly for the observed effects on
behaviour.
A strength of the study was the low dropout from
the study and the fact that we worked with smokers
who did not self-select for the intervention. The study
population were Dutch smokers in either the contemplation or preparation stage of change and who were
not particularly interested in receiving information
about smoking cessation aids nor help with quitting.
This is in contrast to most experimental studies, which
use highly motivated self-selected smokers as study participants, often recruited through self-referral in clinics
or through newspaper advertisements. In this context,
the high quit rate that was obtained in our study is
remarkable, especially given the fact that treatment
effects were most pronounced among smokers who did
not intend to quit smoking within the next 4 weeks
(contemplation stage). These smokers typically do not
enter smoking cessation treatment trials. In the general
population of smokers, this group is larger than the
group of smokers in the preparation stage. An important conclusion is therefore that widespread diffusion of
decision aids such as the one presented in the present
study can potentially help reduce the burden of disease
caused by tobacco smoking. We recommend providing
the decision aid to all smokers contemplating quitting,
for example through family physicians, pharmacies and
other intermediaries.
A possible limitation of the study was the reliance on
self-reports of outcomes. We did not conduct biochemical
verification of smoking status, because this would have
affected our high participation and response rates.
In a study such as the present one, that features a lowintensity, population-based intervention trial involving
smokers who have no specific characteristics making it

plausible for them to deceive, biochemical validation is no

longer recommended (Benowitz et al. 2002). More specifically, we do not think that the respondents felt any
pressure to provide socially desirable answers, as they
received financial incentive for participating in the study
completely irrespective of the outcomes. Another possible
limitation of the study was the small sample size in the
group of smokers who made a quit attempt (n = 269),
leaving open the possibility that the intervention did
improve usage of NRT and perhaps even reduced the
usage of advice from family physicians. However, the total
number of respondents (n = 1.014) was sufficient to draw
firm conclusions about the effect of the intervention on
behavioural outcomes.
In sum, a decision aid designed to help smokers decide
to use efficacious treatment when attempting to quit
smoking was found to have a positive effect on smoking
cessation, while failing to increase the usage of treatment. This finding lends support to the notion that
promotion of the availability of efficacious treatment
methods for tobacco addiction may be as important or
even more important for many quitters than the actual
usage of these treatments.
Acknowledgements
This research was made possible by a grant from the
Dutch Health Research and Development Council
(ZON-MW).
Declaration of interest
The decision aid that is the subject of the study was developed by the same institute at which the research was carried out.
References
Bandura, A. (1986) Social Foundations of Thought and Action.
Englewood Cliffs, NJ: Prentice Hall.
Bansal, M. A., Cummings, K. M., Hyland, A. & Giovino, G. A.
(2004) Stop-smoking medications: who uses them, who misused them, and who is misinformed about them? Nicotine and
Tobacco Research, 6, S303S310.
Benowitz, N. L., Ahijevych, K., Jarvis, M. J., LeHouezec, J., Hansson, A., Lichtenstein, E. et al. (2002) Biochemical verification
of tobacco use and cessation. Nicotine and Tobacco Research, 4,
149159.
Boyle, R. G., Solbeg, L. I., Magnan, S., Davidson, G. & Alesci, N.
L. (2002) Does insurance coverage for drug therapy affect
smoking cessation? Health Affairs, November/December,
162168.
Britton, J. (2004) Smoking cessation services: use them or lose
them. Thorax, 59, 548549.
Cabana, M. D., Rand, C. S., Powe, N. R., Wu, A. W., Wilson, M.
H., Abboud, P.-A. C. et al. (1999) Why dont physicians follow
clinical practice guidelines? A framework for improvement.
JAMA, 282, 14581465.
Cokkinides, V. E., Ward, E., Jemal, A. & Thun, M. J. (2005)
Under-use of smoking cessation treatments: results from the

2006 The Authors. Journal compilation 2006 Society for the Study of Addiction

Addiction, 101, 441449

Helping smokers to choose cessation methods

National Health Interview Survey. American Journal of Preventive Medicine, 28, 119122.
Curry, S. J., Grothaus, L. C., McAfee, T. & Pabiniak, C. (1998)
Use and cost effectiveness of smoking-cessation services under
four insurance plans in a health maintenance organization.
New England Journal of Medicine, 339, 673679.
Dijkstra, A. & de Vries, H. (2001) Do self-help interventions in
health education lead to cognitive changes, and do cognitive
changes lead to behavioural change? British Journal of Health
Psychology, 6, 121134.
Estabrooks, C., Goel, V., Thiel, E., Pinfold, P., Sawka, C. & Williams, I. (2001) Decision aids: are they worth it? A systematic
review. Journal of Health Services Research and Policy, 6, 170
182.
Etter, J. F. & Perneger, T. V. (2001) Attitudes toward nicotine
replacement therapy in smokers and ex-smokers in the general public. Clinical Pharmacology and Therapeutics, 69, 175
183.
European Commission. (2004) Tobacco or Health in the European
Union: Past, Present, and Future. ASPECT report. Luxembourg:
Office for Official Publications of the European Communities.
Fiore, M. C., Bailey, W. C., Cohen, S. J., Faith-Dorfman, S.,
Goldstein, M. G., Gritz, E. R. et al. (2000) Treating Tobacco Use
and Dependence Clinical Practice Guideline. Rockville, MD: US
Department of Health and Human Services Public Health
Service.
Hammond, D., McDonald, P. W., Fong, G. T. & Borland, R.
(2004) Do smokers know how to quit? Knowledge and perceived effectiveness of cessation assistance as predictors of cessation behaviour. Addiction, 99, 10421048.
Heatherton, T. F., Kozlowski, L. T., Frecker, R. C. & Fagerstrm,
K. O. (1991) The Fagerstrm test for nicotine dependence; a
revision of the Fagerstrm Tolerance Questionnaire. British
Journal of Addiction, 86, 11191127.
Hyland, A., Li, Q., Bauer, J. E., Giovino, G. A., Steger, C. & Cummings, K. M. (2004) Predictors of cessation in a cohort of current and former smokers followed over 13 years. Nicotine and
Tobacco Research, 6, S363S369.
Mudde, A. N., Kok, G. J. & Strecher, V. J. (1995) Self-efficacy as
a predictor for the cessation of smoking: methodological issues
and implications for smoking cessation programs. Psychology
and Health, 10, 353367.
National Health Committee (1999) Guidelines for Smoking Cessation. Wellington: Ministry of Health.

449

Paul, C. L., Wiggers, J., Daly, J. B., Green, S., Walsh, R. A.,
Knight, J. et al. (2004) Direct telemarketing of smoking cessation interventions: will smokers take the call? Addiction, 99,
907913.
Prochaska, J. O. & DiClemente, C. C. (1983) Stages and processes
of self-change of smoking: toward an integrative model of
change. Journal of Consulting and Clinical Psychology, 51, 390
395.
Prochaska, J. O., DiClemente, C. C. & Norcross, J. C. (1992) In
search of how people change: applications to addictive behaviors. American Psychologist, 47, 11021114.
Raw, M., McNeill, A. & West, R. (1998) Smoking cessation
guidelines for health professionals: a guide to effective smoking cessation interventions for the health care system. Thorax,
53, S1S19.
Schauffler, H. H., McMenamin, S., Olson, K., Boyce-Smith, G.,
Rideout, J. A. & Kamil, J. (2001) Variations in treatment benefits influence smoking cessation: results of a randomised controlled trial. Tobacco Control, 10, 175180.
Schnoll, R. A. & Engstrom, P. F. (2004) Tobacco control in the
physicians office: a matter of adequate training and resources.
Journal of the National Cancer Institute, 96, 553575.
Van Weel, C., Coebergh, J. W. W., Drenthen, T., Schippers, G. M.,
van Spiegel, P. I., Anderson, P. D. et al. (2005) De klinische
richtlijn: behandeling van tabaksverslaving [The practice
guideline: Treatment of tobacco dependence]. Nederlands Tijdschrift Voor Geneeskunde, 149, 1721.
West, R., DiMarino, M. E., Gitchell, J. & McNeill, A. (2005)
Impact of UK policy initiatives on use of medicines to aid
smoking cessation. Tobacco Control, 14, 166171.
Willemsen, M. C., Wagena, E. & van Schayck, O. (2003) De effectiviteit van stoppen-met-rokenmethoden die in Nederland
beschikbaar zijn: een systematische review op basis van
Cochrane-gegevens. [The efficacy of smoking-cessation methods available in the Netherlands: a systematic review based on
Cochrane data]. Nederlands Tijdschrift Voor Geneeskunde, 147,
922927.
Zeegers, T., Segaar, D. & Willemsen, M. (2005) Roken: de Harde
Feiten 2004 [Smoking: the Hard Facts 2004]. The Hague:
STIVORO.
Zhu, S.-H., Melcer, T., Sun, J., Rosbrook, B. & Pierce, J. P. (2000)
Smoking cessation with and without assistance: a populationbased analysis. American Journal of Preventive Medicine, 18,
305311.

2006 The Authors. Journal compilation 2006 Society for the Study of Addiction

Addiction, 101, 441449