Helping pregnant women stop smoking with a dual tasking treatment

Smoking in pregnancy is associated with many adverse health outcomes for the
unborn child including low birthweight (Shah & Bracken, 2000), stillbirth
(Flenady et al., 2011), premature delivery (Pollack, Lantz, & Frohna, 2000),
cardiovascular/heart defects, muscoskeletal defects, missing limbs/fingers,
clubfoot, craniosynostosis, facial defects, eye defects, orofacial clefts,
gastrointestinal defects, gastroschisis, anal atresia, hernia and undescended
testes (Hackshaw, Rodeck, & Boniface, 2011). However, despite the adverse
outcomes, fairly high smoking prevalences of 12*25% (US), 13*36% (Europe) and
13*31% (other countries) are reported in the period 1982-2001 (Schneider &
Schütz, 2008). In a recent study done in the Netherlands, over the period
2001-2015 12,2% of pregnant women indicated having smoked during pregnancy
(Lanting et al., 2015). A Dutch pregnant woman who was having trouble quitting
smoking elaborates in a recent article from Nederlandse Omroep Stichting (NOS)
about the help she received: *I was sent home with a flyer and the advice I
should do it on my own* (Nederlandse Omroep Stichting, 2017). The interview was
done to advertise a recent Dutch campaign launched by state secretary Van Rijn
to reduce the percentage of pregnant smokers (Nederlandse Omroep Stichting,
2017).
Several treatments to aid in smoking cessation during pregnancy have
been researched. Lumley et al. (2009) found that all smoking cessation
treatments (cognitive behavioural therapy/motivational interviewing, advice and
counselling based on assessment of stages of change, interventions using
feedback on fetal health status/mother*s health, provision of rewards and
incentives, provision of pharmacotherapies and other therapies, for example
hypnosis) compared to no treatment, show a small but statistically significant
improvement in smoking cessation outcomes in late pregnancy (RR = 0.94, 95%
confidence interval (CI) 0.93-0.96). Thus in the intervention condition 6% less
women continued to smoke (Lumley et al., 2009). Because most smoking cessation
programs offer multiple, partly overlapping modules in their treatment, main
components were identified in Lumley*s study to allow for subgroup analyses.
Subgroup analyses comparing the main intervention strategies showed that
contingency management showed a significantly larger effect (RR = .76, 95% CI
0.71 to 0.81) than Cognitive Behavioural Therapy (CBT: RR = .95, 95% CI 0.93 to
0.97) and Nicotine Replacement Therapy (NRT) (RR 0.95, 95% CI 0.92 to 0.98).
Interventions using a main component of *stages of change* showed a significant
reduction in smoking outcomes in late pregnancy (RR = .99 95% CI 0.97 to 1.00).
Interventions using feedback were found not to be significantly effective (RR =
.92, CI 0.84 to 1.02). However, great heterogeneity makes it difficult to
interpret the results of Lumley et al. (2009) at a general level. Two other
meta-analyses have looked at NRT alone (Coleman, Chamberlain, Cooper, &
Leonardi*Bee, 2011; Trivedi, 2013) and concluded that there were no significant
treatment effects compared to placebo (respectively RR = 1.76, 95% CI:
0.92-3.36; RR = 1.33, 95% CI 0.93*1.91). A narrative review focusing on partner
focused support (Hemsing, Greaves, O*Leary, Chan, & Okoli, 2012), found that 7
out of 9 trials found no effect of the intervention on smoking cessation.
Lastly, Greaves et al. (2011) outlined the evidence for contingency based
interventions and concluded the intervention to be effective on the short term.
In conclusion, the evidence for successful smoking cessation interventions in a
pregnant population is not very compelling.
In addition to most treatments not having the desired effect, the current array
of treatments has significant drawbacks associated with them. For example, for
NRT researchers have their doubts about safety for the unborn child (Coleman et
al., 2011; Trivedi, 2013). Still, clinical guidelines in the Netherlands
recommend using NRT when patient smoke more than 10 cigarettes a day, even when
they are pregnant (Kwaliteitsinstituut voor de Gezondsheidszorg CBO, 2009).
Drawbacks of other (effective) interventions are for example the need for
specialized personnel, significant time investment (cognitive behavioural
therapy), ethical and financial considerations, as well as societal
acceptability (contingency management) (Greaves et al., 2011). Because of these
drawbacks, the lack of efficacy within this population and the specific needs
within the (Dutch) healthcare system, there is need for a safe, acceptable and
cost and time inexpensive new treatment.
As in most addictions, two mutually interconnected emotional and behavioural
components can also be identified in smoking addiction: first there is the
emotional aspect of wanting, or even needing the cigarette. If this component
stays unanswered to long, the addicted subject experiences craving. The second
aspect is the behavioural component: smoking cigarettes. Psychological
anti-smoking interventions target either one of these components or both. The
current study focuses on the first aspect of (cigarette) cravings, since
research has shown that it*s an important pathway to smoking relapse (Zhou et
al., 2009), especially post quit craving (Killen & Fortmann, 1997; Wray, Gass,
& Tiffany, 2013).

Eye Movement Desensitation and Reprocessing (EMDR). One treatment which has
received research attention in treatment of cravings and addictive behaviours
is EMDR. Originally stemming from trauma research, EMDR is a treatment which is
effective in treating traumatic memories. Currently, the best explanation for
EMDR comes from working memory theory (Andrade, Kavanagh, & Baddeley, 1997). It
states that by recalling a traumatic memory from long term memory, it becomes
labile in the working memory. By engaging in a second task (in the traditional
EMDR procedure fast horizontal eye movements by following the therapist*s hand)
while simultaneously thinking of the traumatic memory, competition is created
between the two tasks within the limited cognitive resources of the working
memory. By repeating this process several times, emotionality and vividness of
the traumatic memory are reduced. The implication is then that eye movements
are not necessary, but that sufficient working memory load in general is
critical for the intervention to be successful (Van den Hout & Engelhard,
2012). Therefore, procedures as EMDR and similar interventions taxing working
memory can be referred to as dual tasking procedures (van den Hout & Engelhard,
2012).
Perhaps encouraged by the fact that craving is shown to take up
cognitive resources (Kemps, Tiggemann, & Grigg, 2008) and its role in relapse
(Killen & Fortmann, 1997; Wray et al., 2013) and smoking (American Psychiatric
Association, 2013), researchers have tried using EMDR to treat memories
associated with high feelings of craving. Littel, van den Hout, and Engelhard
(2016) conducted a proof of principle study to research if EMDR has potential
in both desensitizing addiction related memory representations and imagery.
Students were asked to formulate a memory or emotional state in which they
experienced craving and smoked a cigarette. While recalling their memory, the
experimental group engaged in eye movements. Littel et al. (2016) found that
eye movements has potential in attenuating craving and associated imagery. A
second study conducting a proof of principle of EMDR in addiction was done by
Markus, de Weert-van Oene, Woud, Becker, and DeJong (2016). In their study
three specific personal memories were formulated that were ordered from high to
low craving and treated with eye movements. Positive, short-term results were
found on craving. However, these results were not contained at follow-up. In
addition, no results were found on smoking outcomes. Markus et al. (2016) call
for dose response studies in which exposure is increased and daily measurements
are made. Lastly, studies looking at interfering with craving by means of
visuospatial interference showed that craving for cigarettes decreased compared
to control in the lab (May, Andrade, Panabokke, & Kavanagh, 2010) and via
digital delivery by playing Tetris (Skorka-Brown, Andrade, Whalley and May,
2015). However, all of these studies were carried out in non-help seeking
groups.
The potential for a dual tasking treatment in smoking cessation in the
short term shows promise. Firstly, because of the behavioral nature of the
treatment, there is a minimization of adverse side effects (both to the unborn
child and mother). In addition, the cost effectiveness of an automatized dual
tasking treatment is of great significance in this population. Lastly, because
this population is likely very motivated to stop smoking, it will serve as an
ultimate test for the efficacy of a dual tasking intervention in smoking
cessation.
Thus, the current study investigates whether a dual tasking treatment
serves as a potential efficacious treatment for reduction of cigarette cravings
and smoking cessation in pregnancy. First, the experimenter identifies a number
of idiosyncratic situations memory which elicit high feelings of craving in the
pregnant woman. These idiosyncratic situations are then used throughout the
study to activate feelings of craving. While craving is activated, participants
play a puzzle game on the tablet (dual tasking).

Specifically, the first objective of the current study is to compare the
efficacy of a dual tasking treatment versus treatment as usual in the treatment
of smoking addiction in a pregnant population on smoking outcomes.

Secondary objectives:
Effect of the intervention on smoking (breathalyzer) nicotine dependence, and
craving as a mechanism. Additionally, by looking at patient characteristics,
such as perceived stress, social support and positive and negative affect, an
attempt is made to be able to see which combination of treatment works for
whom.

This study is an individual randomized controlled trail with 2 arms: Full dose
condition and Treatment as Usual+ (TAU+) only. The two groups are compared on
variables measured on T1 and T2 . This amounts to a 2x2 design with the between
factor being *condition* and the within factor *time*.

Justification of design
Because of the vulnerability of the group of pregnant women and their unborn
babies, the control group is set to an *upgraded* Treatment as Usual. Since the
gynecologists and midwives indicated that help is available but not included in
the standard package of care, this was deemed appropriate. Moreover, this
*upgraded* TAU controls for differing activation of the app between groups.

Because of the experimental nature of this intervention, it was deemed
appropriate to compare a maximum dose intervention to the above mentioned
control condition.

All interventions will be performed by the first researcher (Tom IJdema) or by
Master students Medical or Clinical psychology (from now on: test leaders).
They are all trained by the senior, EMDR schooled therapist Kees Korrelboom).

Treatment as Usual+ (TAU+): control
All participants will receive TAU+ when enrolled in the study. The test leader
will have a short talk about the downsides of smoking, one booklet (*Rookvrij
ook jij*) is handed out with anti-smoking tips and additional resources which
the pregnant woman may pursue. While the use of nicotine patches of gum and
e-cigarettes is not an exclusion criterion, this use is not encouraged.
However, its use will be registered in advance of and during treatment.
Participants assigned to the *TAU+ only* condition will receive an app in which
they receive facts about smoking. To ensure equal activation among conditions,
the prompts filled with information will follow the same schedule as the
prompts in the dual tasking app. Since the TAU folder contain mostly links and
telephone numbers for additional resources rather than information, additional
folders were used for the app. These are: "Rookvrij zwanger, dat bevalt beter"
(rougly translated to: *Smokefree pregnant, that feels better' from TRIMBOS
instituut) and "Wat je moet weten over Alcohol en Roken voor, tijdens en na de
zwangerschap" (roughly translated to: *what you should know about alcohol and
smoking before, during and after pregnancy* from TRIMBOS and STAP).

Full dose
In addition to receiving treatment as usual, participants assigned to the *full
dose intervention* will also receive therapist-assisted dual tasking treatment
during a f-t-f session. Treatment is given according to a standardized dual
tasking protocol targeting the desensitization of craving. In line with the
protocol, idiosyncratic memories and daily situations are targeted which are
associated with high feelings of craving.
The dual task intervention consists of playing a puzzle game on a touch
screen tablet. The researcher helps the patient to formulate three trigger
situations which the patient associates with high feelings of craving. A short
script is formulated. Next, three keywords are formulated to summarize the
trigger situation and these words are put into the software. Then, the patient
selects nine other general keywords associated with smoking from a list, which
are then put in the software as well. Then, the patient is stimulated to
visualize the memory and the associated craving as vividly as she can imagine.
Then, the participants fills in a manipulation check. Then she engages in
completing the pictures on the tablet. While playing the puzzle game, the
selected keywords appear in the top of the screen to keep stimulating the
participant to think of their craving and thus emphasizing the dual tasking
aspect of the intervention. After the participant has completed four sets of 45
seconds playing the puzzle game, the participant fills in the manipulation
check. Thereafter, the interviewer assesses the next trigger situation in the
same procedure described above and the puzzles are played in the same manner.
This is repeated until 3 trigger situations have been completed. See below for
a graphical representation of the intervention.
After the procedure, the mobile app (based on the same principles as in
the session) is uploaded on the patient*s personal mobile phone. The
personalized keywords are loaded into their app and instructions are given to
practice at home. After four weeks of practicing with the dual tasking app at
home, the patient will come to the hospital for an end evaluation. The app is
closed off for more practice after that.

For the current study the benefits greatly outweigh the risks. With regard to
burden: the participant is required to invest a minimum of 1 hour and 40
minutes (session 1: max 90 minutes, session 2: max 10 minutes). In addition,
questionnaires are conducted for a total of 20 minutes. Lastly, it is advised
to do the exercise on the app (for a total of 4 hours). The time investment is
significant should the participant choose to do all the exercises on the mobile
phone. However, these brief moments (3 minutes per exercise) are spaced out
throughout the days and 4 weeks of the intervention. The sessions in the
hospital are planned consecutively to a regular pregnancy consult at the
hospital to minimalize burden on the participant.

The risks for participating in the study are considered low. Researchers are
trained and supervised intensively by an experienced EMDR therapist. The
protocol is highly standardized. Secondly, the research will take place in the
hospital and midwive practices, where trained personnel is nearby should a
calamity arise. Thirdly, inducing craving in participants without presence of a
researcher could be considered a risk. However, previous research doesn*t
report risk associated with this treatment (Littel et al., 2016; Markus et al.,
2016). Lastly, the participants are free to stop with the study, or to seek out
additional treatment, without consequences for treatment in the hospital.

With regard to benefits, it was shown in several studies that there were
positive or buffering short term effects on smoking craving measures (Littel et
al., 2016; Markus et al., 2016; May et al., 2010; Skorka-Brown et al., 2015).
Combined with the fact that most existing treatments are neither safe,
effective or cost-effective in pregnant women, we believe it to be justifiable
to test the efficacy of the treatment in this population.