Access to this full-text is provided by Springer Nature.
Content available from Journal of Gambling Studies
This content is subject to copyright. Terms and conditions apply.
ORIGINAL PAPER
Accepted: 19 August 2023 / Published online: 2 September 2023
© The Author(s) 2023
Nicki A. Dowling
nicki.dowling@deakin.edu.au
1 School of Psychology, Deakin University, Geelong, Australia
2 Melbourne Graduate School of Education, University of Melbourne, Parkville, Australia
3 Department of Psychology and Neuroscience, Auckland University of Technology, Auckland,
New Zealand
Applying the Just-In-Time Adaptive Intervention Framework
to the Development of Gambling Interventions
Nicki A.Dowling1,2 · Simone N.Rodda1,3· Stephanie S.Merkouris1
Journal of Gambling Studies (2024) 40:717–747
https://doi.org/10.1007/s10899-023-10250-x
Abstract
Just-In-Time Adaptive Interventions (JITAIs) are emerging “push” mHealth interventions
that provide the right type, timing, and amount of support to address the dynamically-
changing needs for each individual. Although JITAIs are well-suited to the delivery of
interventions for the addictions, few are available to support gambling behaviour change.
We therefore developed GamblingLess: In-The-Moment and Gambling Habit Hacker, two
smartphone-delivered JITAIs that dier with respect to their target populations, theoretical
underpinnings, and decision rules. We aim to describe the decisions, methods, and tools
we used to design these two treatments, with a view to providing guidance to addiction
researchers who wish to develop JITAIs in the future. Specically, we describe how we
applied a comprehensive, organising scientic framework to dene the problem, dene
just-in-time in the context of the identied problem, and formulate the adaptation strate-
gies. While JITAIs appear to be a promising design in addiction intervention science, we
describe several key challenges that arose during development, particularly in relation to
applying micro-randomised trials to their evaluation, and oer recommendations for future
research. Issues including evaluation considerations, integrating on-demand intervention
content, intervention optimisation, combining active and passive assessments, incorporat-
ing human facilitation, adding cost-eectiveness evaluations, and redevelopment as trans-
diagnostic interventions are discussed.
Keywords Mobile health · Just-in-time adaptive intervention · Ecological momentary
intervention · Microrandomised trial · Gambling · Treatment
1 3
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Journal of Gambling Studies (2024) 40:717–747
Introduction
Mobile health (mHealth) interventions, which use mobile or wireless technologies to pro-
mote health (World Health Organization [WHO], 2017), can extend the provision of sup-
port for changing health behaviours beyond that provided by standard treatments. mHealth
interventions have many advantages, such as their accessibility, availability, convenience,
anonymity, transportability, and cost-eectiveness (Bakker et al., 2016; Carpenter et al.,
2020; Heron and Smyth, 2010; Kim et al., 2019; Klasnja & Pratt, 2012; Walton et al., 2018).
They also have a high potential as low-burden and scalable interventions that can accurately
record data and be translated to the real-world (Bakker et al., 2016; Carpenter et al., 2020;
Heron and Smyth, 2010; Kim et al., 2019; Klasnja & Pratt, 2012; Walton et al., 2018). This
modality also oers the unique potential to meet the needs of populations that are under-
served by traditional treatments, including people who are not able or willing to engage in
other treatments, by reducing geographic, nancial, or social help-seeking barriers (Bak-
ker et al., 2016; Heron and Smyth, 2010; Kim et al., 2019). mHealth interventions, which
most often involve health or motivational messages, reminders, or support, can supplement
traditional treatments or be employed as stand-alone treatments (Heron & Smyth, 2010).
They can be “pull” interventions that are initiated by individuals when they want support,
or “push” interventions that are initiated by computerised intervention protocols to decide
when and how support should be oered (Klasnja et al., 2015; Walton et al., 2018).
Just-In-Time Adaptive Interventions
Just-In-Time Adaptive Interventions (JITAIs) are a suite of increasingly popular “push”
mHealth intervention designs that tailor the type, timing, and amount of support toaddress
each individual’s dynamically-changing needs (Carpenter et al., 2020; Nahum-Shani et al.,
2015; Nahum-Shani et al., 2014; Nahum-Shani et al., 2018; Walton et al., 2018). Just-in-time
support refers to providing the right type, timing, or amount of support, while adaptive refers
to the use of dynamic information to repeatedly deliver this support to maximise outcomes
(Collins et al., 2004; Nahum-Shani et al., 2015; Nahum-Shani et al., 2014; Nahum-Shani et
al., 2018; Wang and Miller, 2020). Across various disciplines, mHealth interventions that
share similar just-in-time and adaptive components have also been described as dynamic
tailoring (Krebs et al., 2010), intelligent real-time therapy (Kelly et al., 2012), and individu-
ally and dynamically-tailored ecological momentary interventions (Heron & Smyth, 2010).
JITAIs aim to prevent negative health outcomes and/or promote positive health behav-
iours (Klasnja et al., 2015; Nahum-Shani et al., 2014). They are developed to provide sup-
port when people are: (a) susceptible to negative health outcomes (states of vulnerability)
and/or positive health behaviour change (states of opportunity); and (b) able and/or willing
to receive and employ the support (states of receptivity) (Nahum-Shani et al., 2015; Nahum-
Shani et al., 2018). In everyday settings, these states can rapidly emerge across individuals
and over time within individuals (Shiman, 2009; Shiman et al., 2008; Stone & Shiman,
1994). JITAIs leverage mobile or wireless technologies, including smartphone-embedded or
wearable sensors and smartphone-delivered ecological momentary assessments (EMAs), to
continuously monitor these dynamically-changing internal states and situational contexts in
real-time to identify the type and timing of providing support, while attempting to maximise
uptake and impact and minimise burden, disruption, and habituation (Carpenter et al., 2020;
1 3
718
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Journal of Gambling Studies (2024) 40:717–747
Klasnja et al., 2015; Nahum-Shani et al., 2015; Nahum-Shani et al., 2014; Nahum-Shani et
al., 2018).
There is empirical evidence that JITAIs are eective in changing behaviour across sev-
eral health domains, including healthy diet, post-traumatic stress, depression, anxiety/stress,
pain, bipolar disorder, weight loss, addiction, diabetes management, and physical activity
(Carpenter et al., 2020; Heron & Smyth, 2010; Nahum-Shani et al., 2014, 2018; Wang &
Miller, 2020). A systematic review and meta-analysis of 33 empirical JITAI studies con-
ducted from 2008 across health domains revealed moderate-to-large eect sizes for JITAIs
relative to both waitlist-control conditions (k = 9, Hedge’s g = 1.65) and non-JITAI treat-
ment conditions (k = 21, Hedge’s g = 0.89; Wang and Miller, 2020). JITAIs are well-suited
to the delivery of interventions across the addictions, given that use episodes or lapses are
precipitated by discrete but uctuating states (e.g., motivation, urges or cravings) or events
(e.g., high-risk situations; Carpenter et al., 2020; Goldstein et al., 2017; Heron and Smyth,
2010; Witkiewitz and Marlatt, 2004). Indeed, there is evidence that JITAIs are feasible,
acceptable, credible, and eective in addressing smoking (Brendryen et al., 2008; Bren-
dryen & Kraft, 2008; Businelle et al., 2016; Free et al., 2011; Hebert et al., 2020; Naughton
et al., 2016; Riley et al., 2008; Rodgers et al., 2005; Vidrine et al., 2006), binge drinking
(Suoletto et al., 2012), heavy drinking (Weitzel et al., 2007), heavy drinking and smoking
(Witkiewitz et al., 2014), and alcohol use disorders (Gonzalez & Dulin, 2015; Gustafson et
al., 2014; Moody et al., 2018).
Gambling Just-In-Time Adaptive Interventions
Despite growing evidence of their ecacy, only a small number of JITAIs have been devel-
oped to support gambling behaviour change. Two smartphone apps, Smartphone-based
Problem Gambling Evaluation and Technology Testing Initiative (SPGeTTI; Humphrey et
al., 2019) and Don’t Go There (Coral et al., 2020), employ geolocation sensors (GPS, gyro-
scopes, accelerometers, and magnetometers) to deliver notications when they detect that
individuals are close to land-based gambling venues. SPGeTTI also includes on-demand
intervention content (gambling diary, self-monitoring tips for relapse prevention, and con-
tacts for help services), while Don’t Go There includes a feature that enables an elected
health professional to access the user’s information. Low rates of recruitment precluded a
planned randomised controlled trial (RCT) evaluating SPGeTTI, whereby only four partici-
pants completed the study. Focus group interviews with a separately recruited sample of 20
gamblers revealed a high interest in the use of JITAIs for intervention delivery, but specic
issues with the SPGeTTI app, including excessive battery drainage. Don’t Go There is cur-
rently in the development stage, with a usability study planned.
Two other smartphone apps collect dynamic information from EMAs to initiate the
delivery of real-time adaptive interventions. Jeu-contrôle, which is a publicly available
smartphone app that has not yet been subject to evaluation, employs EMAs to provide per-
sonalised feedback to support time and expenditure limit adherence (Khazaal et al., 2017).
GamblingLess: Curb Your Urge is a smartphone app-delivered intervention based on the
relapse prevention model, which aims to prevent subsequent gambling episodes by reduc-
ing gambling cravings (Hawker et al., 2021; Merkouris et al., 2020). This individually-
and dynamically-tailored EMI, which was adapted from GamblingLess, an evidence-based
online self-directed program (Dowling et al., 2018, 2021; Hawker et al., 2021; Humphrey et
1 3
719
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Journal of Gambling Studies (2024) 40:717–747
al., 2020; Humphrey et al., 2022; Merkouris et al., 2020; Merkouris et al., 2017; S. N. Rodda
et al., 2019), tailors gambling craving management activities in response to repeated EMAs
measuring craving intensity. These intervention activities are also available ‘on-demand’.
Usability testing revealed that 29 key stakeholders (consumers, gambling clinicians, and
gambling researchers) (Hawker et al., 2021; Merkouris et al., 2020) highly rated the inter-
vention content, helpfulness, acceptability, and usability. In a pilot study (Hawker et al.,
2021), participants demonstrated a more than 70% reduction in the average number of gam-
bling episodes and cravings during the 4-week intervention period, as well as a 10% reduc-
tion in craving intensity immediately after a treatment activity. At the post-intervention and
one-month follow-up evaluations, participants reported signicant medium-to-large reduc-
tions in gambling symptom severity, gambling frequency and expenditure, cravings, and
self-ecacy. In an evaluation of the clinical impact of the JITAI, just under half of all par-
ticipants (48%) reported either recovery or improvement in the severity of their gambling
symptoms at the follow-up evaluation.
Review Manuscript Aims
We have recently developed two theoretically-informed and evidence-based JITAIs. The
rst, GamblingLess: In-The-Moment (Dowling et al., 2022), is part of a suite of gambling
online and mHealth interventions that builds on the pilot trial data of GamblingLess: Curb
Your Urge (Hawker et al., 2021; Merkouris et al., 2020), while the second, Gambling Habit
Hacker (Rodda et al., 2022), forms part of a suite of implementation support interventions
based on lived experience across the addictions (Brittain et al., 2021; Park et al., 2020;
Rodda et al., 2020; S. N. Rodda, N. Booth, Rodda et al., 2018a, c, d). In response to a call
for continued communication regarding the need to develop and evaluate JITAIs (Goldstein
et al., 2017), this review manuscript aims to describe the decisions, methods, and tools we
used to design these two JITAIs. This review manuscript complements the protocol papers
for these JITAIs (Dowling et al., 2022; Rodda et al., 2022), which describe their initial
evaluation and optimisation protocols. In contrast, this manuscript describes the steps we
took to develop GamblingLess: In-The-Moment and Gambling Habit Hacker and how and
why we made the decisions we did, with a view that sharing our approach will providing
guidance and encouragement to addiction researchers who wish to develop JITAIs in the
future (Goldstein et al., 2017).
We modelled this review manuscript on a similar paper authored by Goldstein et al.
(2017), which describes their approach when developing a JITAI targeting lapses follow-
ing a weight control diet. Like Goldstein et al. (2017), we describe how and why we made
certain decisions when we applied the comprehensive, organising scientic framework
developed by Nahum-Shani and colleagues (2015; 2014; 2018) to guide the JITAI design.
In this framework, four components are described: (1) decision points (timepoints at which
intervention delivery decisions are made); (2) intervention options (possible types, doses,
timings, and delivery modes of the support that may be provided at each decision point);
(3) tailoring variables (information relating to a person’s ecological context or internal state
that is employed to identify when and/or how intervention options are delivered); and (4)
decision rules (rules that specify which intervention option is to be oered, and when, for
each individual at each level of the tailoring variables). These components are primarily
guided by the distal outcome (long-term intervention goal), but also by multiple proximal
1 3
720
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Journal of Gambling Studies (2024) 40:717–747
outcomes (short-term intervention goals) (Nahum-Shani et al., 2018). Nahum-Shani et al.
(2015) organises these components into three areas: (1) dening the problem, (2) dening
what just-in-time means in the context of the problem, and (3) formulating the adaptation
strategy. We found Goldstein et al.’s (2017) application of this framework to a specic JITAI
incredibly helpful in informing our decision-making across the development and evaluation
phases for our JITAIs; and we oer this manuscript in the same spirit.
Our review manuscript rst provides an overview of each of our JITAIs, followed by
why we selected the behaviour change theories that guided their construction and how we
applied the scientic framework to guide their design. We conclude the manuscript with a
discussion of how and why we applied micro-randomised trials (MRTs) to enable the opti-
misation of these JITAIs, followed by the key logistical and methodological challenges we
faced during the development and evaluation phases, as well as considerations for future
research.
Overview of the JITAIs
In line with recommendations (Nahum-Shani et al., 2018), the development of both Gam-
blingLess: In-The-Moment and Gambling Habit Hacker involved a multidisciplinary col-
laboration with expertise drawn from clinical psychology, social psychology, biostatistics,
research design, implementation science, and technology development. Both apps will be
subject to 28-day MRTs, accompanied by within-group follow-up evaluations across a six-
month period and acceptability evaluations (Dowling et al., 2022; Rodda et al., 2022). Both
apps are available for download during the trial period on Android (Google Play Store) and
Apple (App Store) devices.
GamblingLess: In-The-Moment
GamblingLess: In-The-Moment is one digital oering in a suite of gambling mHealth inter-
ventions that are evidence-based and theoretically-informed. It is a smartphone app-deliv-
ered JITAI for people who want to quit or reduce their gambling. The aim of this JITAI is
to provide the type and amount of support required at times when people are cognitively
vulnerable (i.e., when they report high-intensity cravings, low self-ecacy, or positive gam-
bling outcome expectancies) to reduce the likelihood of a subsequent gambling episode. The
long-term goal is to reduce gambling symptom severity (distal outcome) via the short-term
goal of reducing the likelihood of gambling episodes (primary proximal outcome). This
reduction in the probability of gambling episodes is posited to be achieved through reduc-
tions in craving intensity, improvements in self-ecacy, or reductions in positive outcome
expectancies (secondary proximal outcomes). In this JITAI, we created decision rules that
specify that individuals who are available for treatment (i.e., in a state of receptivity) and
report a state of cognitive vulnerability (characterised by high craving intensity, lowered
self-ecacy, and high positive outcome expectancies: tailoring variables) in EMAs deliv-
ered at three semi-random times each day (decision points) are delivered tailored cogni-
tive-behavioural and third-wave intervention activities designed to address these cognitive
processes (intervention options). The JITAI is designed to be used as a standalone or adjunc-
1 3
721
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Journal of Gambling Studies (2024) 40:717–747
tive treatment when individuals are actively gambling or to prevent relapse during recovery.
Illustrative screenshots of GamblingLess: In-The-Moment are displayed in Fig. 1.
Gambling Habit Hacker
Gambling Habit Hacker is a digital oering in a suite of treatments for addictive behaviours
delivering implementation support that has been developed using lived experience research.
It is a smartphone app-delivered JITAI for people who want to improve their ability to
adhere to their gambling expenditure limits (i.e., goals). This JITAI aims to provide the type
of support required at times of goal vulnerability (low strength of intention [to adhere to
their gambling expenditure limits], low goal self-ecacy, low urge self-ecacy, and high-
risk situations) to enhance adherence to gambling expenditure limits. The long-term goal is
to reduce gambling expenditure (distal outcome) via the short-term goal of increased adher-
ence to gambling expenditure limits (primary proximal outcome). This increased adher-
ence to gambling expenditure limits is posited to be achieved via increased strength of
intention, increased goal self-ecacy, and increased urge self-ecacy (secondary proximal
outcomes). In this JITAI, we created decision rules that specify that participants who are
available for treatment (i.e., in a state of receptivity) and indicate that they are in a state of
goal vulnerability (characterised by low strength of goal intention, low goal self-ecacy,
low urge self-ecacy, or a high-risk situation: tailoring variables) in EMAs sent at three
semi-random times a day (decision points) are encouraged to engage in action and coping
planning activities designed to facilitate the use of behaviour change strategies (interven-
Fig. 1 Illustrative screenshots of GamblingLess: In-The-Moment: Welcome page, create an account,
check-in (EMA) “snooze”
1 3
722
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Journal of Gambling Studies (2024) 40:717–747
tion options). Although individuals undertake the planning activities within the app, the
implementation of the plan is conducted in the real world. This JITAI is intended for use
as a standalone intervention across the entire period of post-intentional action, inclusive of
longer term maintenance of behaviour change. Illustrative screenshots of Gambling Habit
Hacker are displayed in Fig. 2.
Theories of Behaviour Change
There is consensus in the literature that existing behaviour change theories are limited in
their ability to guide the construction of JITAIs as they fail to describe the dynamic pro-
cesses underlying states for receptivity, vulnerability, and opportunity (Nahum-Shani et al.,
2015; Nahum-Shani et al., 2014; Nahum-Shani et al., 2018; Riley et al., 2011; Spruijt-Metz
and Nilsen, 2014). Theories that explain the emergence of a state of vulnerability and/or
opportunity as a dynamic process involving the interaction of stable and transient factors
are most helpful, but still lack specication about the temporal relationships between fac-
tors in a manner that informs the type, timing, and amount of support provided (Klasnja et
al., 2015; Nahum-Shani et al., 2015; Nahum-Shani et al., 2014; Nahum-Shani et al., 2018).
GamblingLess: In-The-Moment
In the absence of such rened behaviour change theories, we employed the reformulated
relapse prevention model (Witkiewitz & Marlatt, 2004) as the guiding theoretical frame-
work for the development of GamblingLess: In-The-Moment. This model explains the like-
lihood of relapse resulting from the multidimensional, non-linear, and dynamic interactions
between various antecedents within high-risk situations. These antecedents include back-
Fig. 2 Illustrative screenshots of Gambling Habit Hacker app: Lived experience quote; individual strat-
egy menu; create an account; check-in (EMA); individual strategy page; action planning
1 3
723
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Journal of Gambling Studies (2024) 40:717–747
ground factors (e.g., family history, comorbid psychopathology, years of dependence, social
support), physiological states (e.g., physical withdrawal), cognitive processes (e.g., crav-
ing, self-ecacy, positive outcome expectancies, motivation, and the abstinence violation
eect), aective states, and coping skills. In this model, responses to high-risk situations
are related to distal and proximal precipitants, which operate within tonic processes and
phasic responses. Tonic processes, which are distal risks or stable background factors that
determine who is vulnerable for relapse, tend to accumulate to set the foundation for the
possibility of relapse. They set the initial threshold for relapse and often lead to the initia-
tion of a high-risk situation. In contrast, phasic responses, which are proximal or situational
processes uctuating over time and contexts, determine when relapse will occur. Momen-
tary coping responses, however, can also be considered to be phasic events that inuence
the degree to which a high-risk situation will result in a lapse. Feedback loops are included
in the model, whereby lapses may reciprocally impact on the same factors (i.e., cognitive
processes, aective states and coping behaviour) that contributed to them. The reformulated
relapse prevention model has received substantial empirical support across a range of addic-
tive behaviours (Menon & Kandasamy, 2018; Witkiewitz and Marlatt, 2004).
In this model, cognitive processes are conceptualised as both tonic processes and phasic
responses, whereby relatively stable cognitive processes, such as global self-ecacy and
outcome expectancies, may serve to function as tonic processes; while cognitive processes
that uctuate over time and contexts, such as cravings or momentary changes in self-e-
cacy and outcome expectancies, may serve to function as phasic responses. Considerable
cross-sectional evidence suggests that cravings, self-ecacy, and positive outcome expec-
tancies, are associated with problem gambling severity, gambling abstinence, and gambling
relapse; and that these cognitive processes improve following face-to-face and self-directed
interventions (See Dowling et al., 2022 for a review of this literature). Although there is less
empirical evidence relating to the degree to which these cognitive processes act as phasic
responses, several EMA studies (Dowling et al., 2020; Hawker et al., 2020) suggest that
cravings and transient changes in self-ecacy, but not transient changes in positive outcome
expectancies, are associated with the likelihood of subsequent gambling episodes in real-
time. Moreover, these studies suggest that all of these cognitive processes interact in real-
time with other factors explicated by the relapse prevention model, such as self-ecacy,
coping motives, cravings, high-risk positive reinforcement situations, positive emotional
states, and coping styles. These ndings suggest that these momentary cognitive processes
are potential intervention targets and mechanisms of change for JITAIs aiming to reduce
gambling behaviour. Accordingly, many of the JITAIs employed in addiction science have
successfully delivered intervention content tailored to contextual features highlighted by the
relapse prevention model to prevent episodes or lapses (Carpenter et al., 2020).
Gambling Habit Hacker
The guiding theory for Gambling Habit Hacker is the Health Action Planning Approach
(HAPA) (Schwarzer & Luszczynska, 2008), with delivery of intervention content aligned
with Self-Determination Theory (Deci & Ryan, 2008; Ntoumanis et al., 2021; Sheeran et al.,
2020). The HAPA is a social cognitive model that aims to address the theorised gap between
intention and behaviour (Schwarzer, 2008; Schwarzer and Luszczynska, 2008; Sutton,
2008). According to the model, behaviour change involves a continuous 2-phase process
1 3
724
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Journal of Gambling Studies (2024) 40:717–747
involving motivation and volition, whereby an individual regularly sets and reviews pri-
orities or goals and makes decisions on whether action is important. The motivation phase
involves intention formation through the realisation that behaviour needs to change, a belief
that change is worthwhile, prioritisation of change over other competing demands, and a
belief that the selected action can be implemented by the individual (task self-ecacy).
The volitional phase involves movement towards implementing behaviour change inten-
tions with specic implementation planning techniques, such as action planning and coping
planning. Action planning determines when, where, and how actions are taken (Sniehotta
et al., 2005). In contrast, coping planning specically addresses obstacles or barriers to
implementing the action (Gollwitzer, 1999; Sniehotta et al., 2005). Operationalised as if/
then planning, coping planning is used to link specic situations or events that may be
barriers to implementing the action (if) with a specic plan that could be implemented to
overcome the barrier (then). The purpose of advance planning is to prepare the individual to
respond to barriers automatically with reduced cognitive burden in-the-moment. In the voli-
tional phase, belief in one’s ability to maintain plans and cope with barriers that may arise
(maintenance self-ecacy) and one’s ability to regain control after a failure to cope with
barriers to action plan implementation (recovery self-ecacy) inuence the implementation
of intentions. Theoretically, both types of planning occur post-intention and prior to action
but it has been suggested that coping planning is more relevant following action planning
(Sniehotta et al., 2005). Meta-analytic evidence indicates that action and coping planning
are eective in reducing addictive behaviours, such as smoking and alcohol use (Malaguti
et al., 2020; McWilliams et al., 2019). There is also emerging evidence that gamblers can
easily develop action plans but that implementation barriers can reduce the success of these
plans (Rodda et al., 2020).
Finally, the overarching framework for delivery of Gambling Habit Hacker is Self-
Determination Theory (Deci & Ryan, 2008; Ntoumanis et al., 2021; Sheeran et al., 2020),
which highlights three components of motivation: autonomy, competence, and relatedness.
Gambling Habit Hacker aims to build competence through the identication of barriers
that could threaten action plans and goal adherence. We also made the decision to include
prompts to develop personalised goal setting and planning to enhance autonomy and include
real-world stories about the use of behaviour change strategies to foster relatedness.
Dening the Problem
Target Populations and Distal Outcomes
The rst step in designing a JITAI is to identify a target population and a distal outcome,
dened as the distal goal of the intervention, which is usually a primary clinical outcome
(Carpenter et al., 2020; Nahum-Shani et al., 2015; Nahum-Shani et al., 2014; Nahum-Shani
et al., 2018). Our two JITAIs have dierent target populations and distal outcomes to accom-
modate users across the continuum of gambling risk who may be experiencing harm from
their gambling. As recommended (Bakker et al., 2016), we wanted to capitalise on the high
accessibility of mHealth interventions to not only attract people with gambling problems,
but also target sub-clinical or at-risk gamblers, particularly because these gamblers account
1 3
725
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Journal of Gambling Studies (2024) 40:717–747
for the majority of population-level harm due to their higher prevalence (Browne et al.,
2017) and typically do not use treatment services (Bijker et al., 2022).
GamblingLess: In-The-Moment’s target population is people who want to reduce or
quit gambling. We anticipate that this app may attract relatively high-risk gamblers but we
also wanted lower-risk gamblers to benet from the intervention. In the relapse prevention
model, the distal outcome is relapse (i.e., a return to the previous problematic behaviour
pattern) (Witkiewitz & Marlatt, 2004), which we operationalised as the severity of gambling
symptoms.
Gambling Habit Hacker’s target population is people with lower severity gambling
problems who want to enhance their adherence to their gambling expenditure limits. In the
HAPA model, the distal outcome is a reduction in the intention-behaviour gap (Schwarzer,
2008; Schwarzer and Luszczynska, 2008; Sutton, 2008), which we operationalised as adher-
ence to gambling expenditure limits (i.e., actual gambling expenditure relative to planned
gambling expenditure). However, because we deemed it unfeasible to accurately collect
daily planned gambling expenditure over long periods of time, we pragmatically selected
gambling expenditure as the distal outcome. Both gambling expenditure and adherence
to gambling expenditure limits therefore guided the development of the remaining JITAI
components.
Proximal Outcomes
Proximal outcomes are dened as short-term treatment goals and are evaluated straight after
the treatment is provided, with a view to evaluating the ecacy of the treatment (Carpen-
ter et al., 2020; Nahum-Shani et al., 2015; Nahum-Shani et al., 2014; Nahum-Shani et al.,
2018). Dening proximal outcomes can enhance the identication of appropriate decision
points, tailoring variables, decision rules, and intervention options (Nahum-Shani et al.,
2018). They can be: (a) intermediate versions of the distal outcome; (b) mediators of the
distal outcome (i.e., critical components in the pathways through which it is hypothesised
that the intervention inuences the distal outcome); and/or (c) outcomes relating to treat-
ment engagement (dened as motivational commitment/investment in the treatment pro-
cess) or intervention fatigue (dened as emotional or cognitive weariness associated with
intervention engagement) (Klasnja et al., 2015; Nahum-Shani et al., 2015; Nahum-Shani et
al., 2014; Nahum-Shani et al., 2018).
Consistent with the reformulated relapse prevention model, we selected gambling lapses
as the primary proximal outcome for GamblingLess: In-The-Moment, but operationalised
this intermediate version of the distal outcome as gambling episodes to reduce bias and
subjectivity in reporting (Goldstein et al., 2017). Moreover, we wanted to oer gamblers
the choice to select non-abstinence treatment goals, consistent with a harm minimisation
approach (Dowling & Smith, 2007; Dowling et al., 2009; Ladouceur, 2005). To this end,
should data allow, we may also explore whether the delivery of the intervention reduces the
probability of subsequent unplanned gambling episodes. We therefore employed a 30-day
Timeline Follow-Forward (an adaptation of the Timeline Follow-Back; Weinstock et al.,
2004) at pre-treatment to measure planned gambling behaviour. The primary analyses, how-
ever, will explore the eect of the treatment on the probability of any subsequent gambling
episode (Dowling et al., 2022). Should data allow, we may also use the Timeline Follow-
Forward data to explore whether the delivery of the intervention reduces the probability
1 3
726
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Journal of Gambling Studies (2024) 40:717–747
of subsequent planned and unplanned gambling expenditure, which is a related variable
that is not formally articulated by the relapse prevention model. Finally, although gambling
episodes best represent short-term progress towards reduced gambling symptom severity,
they are theoretically (Witkiewitz & Marlatt, 2004) and empirically (Dowling et al., 2020;
Hawker et al., 2020) associated with specic precipitating cognitive processes (cravings,
self-ecacy, and positive outcome expectancies) that could serve as tailoring variables.
We therefore selected these three mediating cognitive processes as secondary proximal out-
comes of reduced gambling symptom severity.
We selected adherence to gambling expenditure limits as the primary proximal outcome
for Gambling Habit Hacker, whereby any gambling expenditure limit can be selected to
accommodate both abstinence and non-abstinence treatment goals. We operationalised goal
adherence as being no higher than 10% more than the planned gambling expenditure, mea-
sured using the Timeline Follow-Forward. We may also explore the inuence of altering the
percentage of adherence as the primary proximal outcome (e.g., 20% exibility) or using
continuous measures of adherence. This approach is consistent with the HAPA model in that
it prompts individuals to form a clear intention prior to engaging in the period of behaviour
change. Because it also takes into account goal vulnerability (low strength of goal intention,
low goal self-ecacy, and low urge self-ecacy), we selected these three mediating pro-
cesses, which could serve as tailoring variables, as proximal outcomes of goal adherence.
Dening Just-In-Time in the Context of the Identied Problem
Decision Points
Decision points are the points in time when a treatment decision is made (Carpenter et al.,
2020; Nahum-Shani et al., 2015; Nahum-Shani et al., 2014; Nahum-Shani et al., 2018).
The identication of factors that signal states of vulnerability or opportunity for proximal
outcomes can help to select decision points (Nahum-Shani et al., 2015). Decision points can
be made at pre-specied time intervals, at specic times of the day, at specic days of the
week, or following random prompts for self-report data (Nahum-Shani et al., 2014, 2018).
The extant theoretical and empirical evidence does not provide much insight into how
we should expect our proximal outcomes to be temporally related over time. The frequency
with which EMAs are delivered in previous JITAIs vary considerably, ranging from once
per week up to ve times per day (Heron & Smyth, 2010). Hence, we considered whether we
expected the process leading to our distal outcomes (reduced gambling symptom severity in
GamblingLess: In-The-Moment and adherence to gambling expenditure limits in Gambling
Habit Hacker), would develop over hours, days, weeks, months, or years (Nahum-Shani
et al., 2015). For GamblingLess: In-The-Moment, momentary changes in cognitive pro-
cesses (craving intensity, self-ecacy, and positive outcome expectancies) can reasonably
be expected to occur at any given minute, thereby potentially leading to immediate reactiv-
ity in the form of a gambling episode. Similarly, for Gambling Habit Hacker, strength of
intention, goal and urge self-ecacy, and whether an individual is in an internal or situ-
ational high-risk situation, can change quickly, thereby increasing risk for non-adherence
to gambling expenditure limits. However, decision points at every minute require frequent
assessments of these states of vulnerability to avoid missed opportunities for intervention
1 3
727
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Journal of Gambling Studies (2024) 40:717–747
provision. Moreover, our primary proximal outcomes (gambling episodes and non-adher-
ence to expenditure limits) occur less frequently (Dowling et al., 2020; Hawker et al., 2020),
suggesting that a less intensive EMA protocol may be required (Heron & Smyth, 2010; Kim
et al., 2019).
We therefore selected three decision points at random times during three pre-specied
periods each day: 8:30am-11:00am (morning), 1:00pm-3:30pm (afternoon), and 5:30pm-
8:00pm (evening). In making this decision, we attempted to balance the likelihood of
obscuring important temporal patterns in the secondary proximal outcomes with the degree
of assessment burden, cognitive overload, potential reactance, and risk of premature treat-
ment dropout posed by too frequent EMAs (Nahum-Shani et al., 2015; Nahum-Shani et al.,
2014; Nahum-Shani et al., 2018). We also considered participant availability and states of
receptivity, despite the fact that we may miss important opportunities for support by exclud-
ing night-time decision points (Klasnja et al., 2015; Nahum-Shani et al., 2015; Nahum-
Shani et al., 2014; Nahum-Shani et al., 2018). The use of semi-random EMA prompts across
each day will allow us to examine the degree to which the timing of intervention delivery
inuences intervention ecacy and engagement. Moreover, an evaluation of the frequency
and timing of the decision points in the acceptability evaluations will inform the limited
information we have in relation to how our proximal outcomes change over time.
Intervention Options
At any of our given decision points, intervention options are the range of potential treat-
ments that may be employed based on our tailoring variables and decision rules (see below)
(Carpenter et al., 2020; Goldstein et al., 2017; Nahum-Shani et al., 2015; Nahum-Shani et
al., 2014; Nahum-Shani et al., 2018). These can include dierent types of support (e.g.,
psychoeducation, feedback, reminders, tips, motivational messages, self-monitoring, goal-
setting, planning behaviour, glanceable displays, coping skills training), support delivery
modes (e.g., provision or availability of support), amount of support (e.g., dose or intensity),
or support delivery media (e.g., phone calls, text messages) (Bakker et al., 2016; Goldstein
et al., 2017; Heron and Smyth, 2010; Kim et al., 2019; Klasnja & Pratt, 2012; Nahum-
Shani et al., 2014; Nahum-Shani et al., 2018). These intervention options, which should be
designed for just-in-time delivery (i.e., precisely when people are in states of vulnerability
or opportunity), are sometimes referred to as EMIs (Heron & Smyth, 2010; Nahum-Shani et
al., 2014, 2018). These intervention options, which often target proximal outcomes, should
be theoretically- and empirically-driven (Nahum-Shani et al., 2015; Nahum-Shani et al.,
2014; Nahum-Shani et al., 2018).
The intervention options in GamblingLess: In-The-Moment were designed to target the
cognitive processes which signal a state of cognitive vulnerability (cravings, lowered self-
ecacy, and endorsement of positive outcome expectancies; secondary proximal outcomes)
that increase the probability of a subsequent gambling episode (primary proximal outcome).
The JITAI comprises 53 activities spanning three separate intervention modules: (1) Curb-
ing Cravings (comprising ten craving management activities); (2) Tackling Triggers (com-
prising 25 activities to enhance self-ecacy in ve high-risk situations: nancial pressures,
unpleasant emotions, social pressure to gamble, testing control over gambling, and conict
with others); and (3) Exploring Expectancies (comprising 18 activities to reduce positive
outcome expectancies organised into three groups: excitement, escape, and money). Most
1 3
728
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Journal of Gambling Studies (2024) 40:717–747
intervention activities take less than ve minutes to complete, consistent with the Gam-
blingLess: Curb Your Urge pilot JITAI (Hawker et al., 2021; Merkouris et al., 2020). The
relapse prevention model informed the development of intervention options (Larimer et
al., 1999; Marlatt & Gordon, 1985; Witkiewitz & Marlatt, 2004), as well as acceptability
feedback from the GamblingLess research program (Dowling et al., 2018, 2021; Hawker et
al., 2021; Humphrey et al., 2020; Humphrey et al., 2022; Merkouris et al., 2020; Merkouris
et al., 2017; Rodda et al., 2019). Hence, the strategies are primarily cognitive and behav-
ioural strategies that focus on the immediate determinants of relapse, but include third wave
approaches, including mindfulness-based and acceptance-based strategies (Larimer et al.,
1999; Marlatt & Gordon, 1985; Marlatt & Witkiewitz, 2005; Witkiewitz & Marlatt, 2004).
Cognitive-behavioural treatments are considered to be the gold standard intervention for
gambling problems (Cowlishaw et al., 2012; Gooding & Tarrier, 2009; Goslar et al., 2017),
with an emerging literature supporting the ecacy of mindfulness-based interventions (de
Lisle et al., 2012; Maynard et al., 2018).
Consistent with the HAPA model (Schwarzer & Luszczynska, 2008), the intervention
options for Gambling Habit Hacker were developed to target the cognitive and behavioural
processes which signal states of goal vulnerability (low strength of intention, low goal self-
ecacy, low urge self-ecacy, and high-risk situations; secondary proximal outcomes) for
spending more than intended (primary proximal outcome). Prior research has identied
multiple categories of self-enactable strategies gamblers use to adhere to their gambling
limits (Hing et al., 2019; Rodda, K. L. Bagot et al., 2018; Rodda et al., 2019a, b), but that
several factors, such as a failure to select t-for-purpose strategies, an inability to sustain
strategy use, shifting priorities, and using conicting strategies, can inuence strategy suc-
cess (Rodda et al., 2017). Goal setting, action planning, coping planning, and self-monitor-
ing were therefore selected as the intervention components for Gambling Habit Hacker to
bridge the gap between intention and behaviour. This JITAI comprises 120 individual strat-
egies (e.g., eat healthy) across 25 higher order strategy groups (e.g., support good health),
which were further organised into 10 higher order behaviour change categories (avoidance,
nancial management, maintaining momentum, managing emotions, rewards, substitution
activities, social support, staying in control while gambling, stress management, and urge
management) to facilitate comparison with the broader evidence base (Michie et al., 2013;
Rodda et al., 2018a, c, d).
In the action planning stage, individuals are prompted to select a tailored strategy group
based on their responses to the tailoring variables, followed by a relevant strategy accompa-
nied by implementation information drawn from lived experience research and prompts for
personalising each specic strategy. Individuals are then prompted to record a personally
tailored-action plan in an open text eld. In the coping planning component, individuals are
prompted to identify the main proximal barrier to the successful implementation of their
action plan (thoughts, emotions, motivation, situation, self-belief, nancial, and social),
describe the details of the barrier that was selected in an open-text box, and record a detailed
plan for this implementation barrier (Armitage, 2009). Finally, participants are encouraged
to participate in commitment and self-ecacy activities focused on strength of character
and mental rehearsal of the plan (Hamilton et al., 2019; Knäuper et al., 2009). We under-
took extensive work to adapt all behaviour change strategies for in-the-moment delivery.
For example, individuals selecting self-exclusion were prompted to engage in the next step
required to implement this strategy (e.g., download the application form). Similarly, coping
1 3
729
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Journal of Gambling Studies (2024) 40:717–747
planning, which is usually undertaken ahead of time (Sniehotta et al., 2005), is prompted
in real time by requesting individuals to consider immediate action to address the identied
barrier. The intervention activities across the action and coping planning components take
between 5 and 10 min to complete.
Importantly, both apps include a Get More Support feature, which enables click-to-call
and click-to-email functions to helpline and web-based specialist gambling services. These
direct linkages into other gambling treatment services allow individuals to escalate the type
of support they wish to receive, which includes immediate crisis support (Bakker et al.,
2016).
One of the biggest challenges when developing mHealth interventions is engagement
with content and client attrition. Although mHealth interventions increase accessibility, they
are characterised by high dropout levels and ‘non-usage attrition’ (unsustained engagement)
(Attwood et al., 2017; Milward et al., 2018; Yardley et al., 2016). Intervention engagement
and intervention fatigue, which uctuate over time, aect intervention adherence, retention,
and eectiveness (Carpenter et al., 2020; Kreyenbuhl et al., 2009; Milward et al., 2018;
Nahum-Shani et al., 2018). Receptivity is therefore emphasised in JITAI designs (Nahum-
Shani et al., 2015; Nahum-Shani et al., 2018). It is argued that the provision of support
when individuals are not receptive is unhelpful and may even be deleterious by exacerbating
intervention engagement and fatigue (Nahum-Shani et al., 2018).
To increase user engagement and minimise intervention fatigue, we aimed to create sim-
ple, aesthetically pleasing designs, and varied the way in which content was delivered in
terms of its presentation, form, and timing. For example, rather than repeatedly delivering
the same intervention content, we encouraged autonomy by incorporating users’ interven-
tion option preferences, whereby they drew from a menu of relevant intervention activities.
Moreover, intervention options are intuitive and easy to navigate, with optimal challenge
and interest levels. Text was written in non-judgemental, inclusive, simple, and hopeful
language and we considered the literacy of intended users in determining sentence and para-
graph length. In GamblingLess: In-The-Moment, we incorporated intervention options that
are interactive and gamied across multiple media platforms (video, audio, quizzes, person-
alised feedback, multiple-choice items, and open-ended items) and included a Pick For Me
feature on each module menu, whereby individuals could allow the app to randomly select
an intervention activity from the menu. We also repeatedly delivered brief static psychoedu-
cational messages via a Did You Know? feature to reduce text. In Gambling Habit Hacker,
we provided space to develop customised plans and included quotes representing the lived
experience of gamblers to enhance relatedness (Deci & Ryan, 2008; Ntoumanis et al., 2021;
Sheeran et al., 2020).
It was also important to consider the ethics of providing interventions in real-life settings
in terms of condentiality, privacy, safety and general welfare of the individual. For this
reason, both apps include a “provide nothing” option in the form of a “snooze” function,
for use in situations in which the individual does not require support or is unreceptive (e.g.,
ignores the EMA prompt), or when providing support may be inconvenient, unethical, or
unsafe (Klasnja et al., 2015; Nahum-Shani et al., 2015; Nahum-Shani et al., 2014; Nahum-
Shani et al., 2018). Individuals can complete an EMA within a two-hour window after the
initial notication to preserve the momentary nature of the treatment while accommodating
the potential for their possible unavailability at the initial notication time (Goldstein et al.,
2017; Klasnja et al., 2015). We will therefore estimate the inuence of the intervention on
1 3
730
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Journal of Gambling Studies (2024) 40:717–747
the proximal outcomes among people who are available for treatment at any given decision
point (Klasnja et al., 2015). Moreover, for GamblingLess: In-The-Moment, we included an
indication of the modality of each activity on each menu (e.g., text, video, interactive, audio,
or text and image) so individuals can make an informed decision regarding appropriate
intervention activities in their current situation. For Gambling Habit Hacker, intervention
content was presented for a range of contexts and situations, including preparing for a gam-
bling session through to gambling in a venue.
Formulating the Adaptation Strategy
Tailoring Variables
Tailoring variables are used to make decisions about when and how to intervene (Carpen-
ter et al., 2020; Collins et al., 2004; Nahum-Shani et al., 2015; Nahum-Shani et al., 2014;
Nahum-Shani et al., 2018). Tailoring variables can be obtained using passive assessments,
active assessments, or both (Kim et al., 2019; Nahum-Shani et al., 2014, 2018; Wang &
Miller, 2020). Passive assessments, which are low-burden because they require no or mini-
mal user engagement, use sensor-equipped smartphones and wearable devices to collect
automated data (e.g., physical activity, temperature, location, light, sound, sleep, blood pres-
sure, heart rate, respiration rate, social interactions, camera images) to make inferences
about internal states and contexts (Hekler et al., 2013; Klasnja & Pratt, 2012; Riley et al.,
2015). In JITAIs, active assessments, which are higher burden because they require user
engagement and compliance, are also known as EMAs (Nahum-Shani et al., 2018). An
EMA design is an event-level prospective methodology that involves the repeated mea-
surement of self-reported symptoms, emotions, behaviour, thoughts, and context in real-
time and in natural environments, usually via smartphones (Shiman, 2009; Shiman et
al., 2008; Stone & Shiman, 1994). Although recent evidence suggests that there is no
signicant dierence in JITAI outcomes when active and passive assessments are employed
(Wang & Miller, 2020), passive data collection does not comprehensively and accurately
evaluate internal states, such as mood, craving, and cognitions (Carpenter et al., 2020; Kim
et al., 2019; Stone & Shiman, 1994). EMA has the added advantage over passive assess-
ments of facilitating accurate self-monitoring, which in turn can lead to increases in emo-
tional self-awareness and regulation (Bakker et al., 2016; Heron and Smyth, 2010; Klasnja
et al., 2015; Walton et al., 2018).
For these reasons, the tailoring variables for GamblingLess: In-The-Moment and Gam-
bling Habit Hacker are assessed using an in-app EMA protocol employing time-based
sampling (i.e., using semi-random prompts for people to input their internal states and situ-
ational contexts) that incorporated event-based sampling (e.g., collecting gambling episode
and expenditure data). Because tailoring variables with low validity can produce high rates
of false positives, the tailoring variables employed in both apps were derived from validated
scales or previous EMA and EMI research (Collins et al., 2004; Goldstein et al., 2017;
Nahum-Shani et al., 2014; Nahum-Shani et al., 2018).
For GamblingLess: In-The-Moment, we selected three tailoring variables (craving inten-
sity, self-ecacy, and positive outcome expectancies) that signal an emerging cognitively
vulnerable state. Each EMA therefore included single items assessing each of these con-
1 3
731
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Journal of Gambling Studies (2024) 40:717–747
structs, which were scored on various 5-point response scales (from 0 to 4, where higher
scores indicate higher vulnerability) (Dowling et al., 2022). EMAs also included items
assessing other momentary internal states and situational contexts that will be employed
to explore the conditions under which the JITAI is more or less eective, including psy-
chological distress, readiness to change, subjective alcohol intoxication, impulsivity, social
context, nancial gambling availability, and location gambling availability (Dowling et al.,
2022).
Gambling Habit Hacker also employed tailoring variables based on the app’s proximal
outcomes that signal the emergence of goal vulnerability (i.e., low strength of intention,
low goal self-ecacy, low urge self-ecacy, and high-risk situations) for subsequent non-
adherence to gambling expenditure limits. These tailoring variables target both the moti-
vational (strength of intention for goal adherence) and volitional (ability to implement and
maintain actions that facilitate goal adherence) phases of the HAPA model (Schwarzer,
2008). The 18-item EMA protocol for this app comprised single items measuring strength
of intention, goal self-ecacy, and urge self-ecacy and 15 items measuring high-risk situ-
ations (including negative reinforcement, positive reinforcement, alcohol consumption, and
gambling proximity), each scored on various 5-point response scales (Rodda et al., 2022)
(from 1 to 5, whereby higher scores indicate lower goal vulnerability for strength of inten-
tion and goal self-ecacy; and higher goal vulnerability for urge self-ecacy and high-risk
situations).
The type, timing, and amount of support can be tailored to individual needs in a JITAI
(Heron & Smyth, 2010; Nahum-Shani et al., 2015; Nahum-Shani et al., 2014; Nahum-Shani
et al., 2018; Wang and Miller, 2020). In both apps, we individualise the type of treatment by
using the EMA tailoring variables to determine which intervention module (GamblingLess:
In-The-Moment) or strategy group (Gambling Habit Hacker) an individual will receive.
In both apps, we individualise the timing of treatment by delivering interventions at times
when individuals are particularly in need of support (Heron & Smyth, 2010) but not when
they do not require support or are not in a state of receptivity. Finally, in GamblingLess:
In-The-Moment, we tailor the amount or dosage of support to individual needs using an
intervention loop, which continues until the individual no longer requires support (based on
responses to specic post-intervention activity EMA items) or closes the app. Should data
allow, we may use these data explore the degree to which each index of cognitive vulner-
ability improves immediately after the delivery of an intervention activity.
Decision Rules
Pre-dened decision rules operationalise the adaptation in a JITAI by specifying which
intervention option is oered, to which people, and under which contexts (Carpenter et al.,
2020; Nahum-Shani et al., 2015; Nahum-Shani et al., 2014; Nahum-Shani et al., 2018).
Because the tailoring variables and intervention options are linked in a systematic way in
decision rules, each decision point is associated with a decision rule (Nahum-Shani et al.,
2014, 2018). Decision rules include the values (which can be levels, thresholds, or ranges)
of each tailoring variable that indicate which intervention option should be oered to each
individual (Nahum-Shani et al., 2014, 2018). Most available JITAIs employ decision rules
that are expressed as a series of conditional statements (e.g., if craving > [threshold], then
1 3
732
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Journal of Gambling Studies (2024) 40:717–747
JITAI recommends delivery of a craving management intervention) (Carpenter et al., 2020;
Nahum-Shani et al., 2015; Nahum-Shani et al., 2014; Nahum-Shani et al., 2018).
In GamblingLess: In-The-Moment, eligibility for an intervention is determined based
on the individual’s momentary level of craving intensity (tailoring variable 1), self-ecacy
(tailoring variable 2), and positive outcome expectancies (tailoring variable 3). At each deci-
sion point, individuals who fail to reach the cut-o point on any tailoring variable (i.e., score
zero) are not eligible for an intervention but are sent an encouraging message. In contrast,
individuals who exceed the cut-o point (i.e., score one or more) on one or more of these
tailoring variables are eligible to be delivered a tailored intervention (i.e., Curbing Crav-
ing, Tackling Triggers, or Exploring Expectancies). Given that the reformulated relapse
prevention model does not postulate that some factors have more inuence in determining
relapse than other factors (Witkiewitz & Marlatt, 2004), individuals who are eligible for
more than one intervention module will be randomly allocated to one of those modules.
After an intervention activity is completed, the intervention loop is potentially triggered
by an individual’s response to a post-intervention EMA item, which is subject to the same
decision rules. Individuals who exceed the cut-o point are presented with a personalised
feedback message and returned to the relevant intervention dashboard. Because there was
insucient empirical evidence to identify the cut-o point of each tailoring variable, we
do not know the level of each tailoring variable at which the delivery of each intervention
module is likely to be benecial versus unnecessary. Although the proposed trial design will
not allow for an evaluation of the causal eect of providing recommendations based on dif-
ferent levels of each tailoring variable (Nahum-Shani et al., 2015), we have deliberately set
low eligibility thresholds so we can potentially optimise the intervention by exploring the
inuence of dierent cut-points on treatment outcomes.
In Gambling Habit Hacker, intervention eligibility is determined based on the individu-
al’s momentary level of strength of intention (tailoring variable 1), goal self-ecacy (tailor-
ing variable 2), urge self-ecacy (tailoring variable 3), and high-risk situations (tailoring
variable 4 comprising 15 situations). Cut-points vary across these tailoring variables: score
of 3 or less for strength of intention and goal self-ecacy, score of 3 or more for urge self-
ecacy, and score of 2 or more on high-risk situations. At each decision point, individuals
who exceed the cut-o point on each tailoring variable are eligible for an intervention,
whereby they are delivered between 12 and 25 of the 25 available strategy groups. A pre-
determined hierarchy determines which strategies are delivered in the event that individu-
als exceeded the cut-points on multiple tailoring variables (see Rodda et al., 2022). In this
hierarchy, threats to adhering to gambling expenditure limits were ordered from most to
least: gambling proximity (engagement in planned/unplanned gambling or planned gam-
bling day), reduced urge self-ecacy, being in a high-risk situation, reduced strength of
intention, and reduced goal self-ecacy. In contrast, individuals who fail to reach the cut-o
point on any tailoring variable are not eligible for an intervention but are sent an encourag-
ing message.
1 3
733
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Journal of Gambling Studies (2024) 40:717–747
The Application of Micro-Randomised Trials
Collins and colleagues (Collins et al., 2009; Collins et al., 2005) have proposed that fac-
torial trial designs form part of the Multiphase Optimization Strategy (MOST), which is
a framework for engineering eective multi-component behavioural interventions. While
traditional factorial designs can be employed to explore the inuence of each intervention
component and important interaction eects, they are unable to determine the conditions
under which each intervention component is most eective (Klasnja et al., 2015). MRTs,
which are a type of sequential factorial design in which each person is randomly allocated
to intervention options at each decision point across a pre-specied period of time, over-
come these limitations (Collins et al., 2009; Collins et al., 2005). Moreover, MRTs are a
highly ecient trial design because the within-subject comparisons in which participants
act as their own control group require smaller sample sizes than traditional full factorial
designs (Klasnja et al., 2015). Although MRTs are an emerging experimental design, there
are several illustrations of their use in addiction science (Carpenter et al., 2020). MRTs
are specically designed to enable the optimisation of JITAIs, which involves deciding the
ways in which a JITAI should be adjusted to make it more eective, ecient, and scalable
(Collins & Kugler, 2018; Collins et al., 2005; Klasnja et al., 2015). Optimisation is particu-
larly important in JITAI design, given the potential burden and disengagement resulting
from these interventions (Collins & Kugler, 2018; Collins et al., 2005; Walton et al., 2018).
Optimisation involves investigating the eectiveness of each component (decision point,
intervention option, tailoring variable, and decision rules) and how well these components
operate together (Collins & Kugler, 2018; Collins et al., 2005). MRTs provide empirical
data for optimising JITIAIs by examining how and under what conditions we should deliver
intervention options to enhance their eectiveness (Carpenter et al., 2020).
We will employ this trial design to inform the optimisation of both of our JITAIs. We
made the decision to conduct MRTs for these JITAIs because there is insucient empirical
and theoretical evidence to fully construct the decision rules that precisely specify when
specic intervention components can be delivered to maximise their eects (i.e., to iden-
tify the decision points, tailoring variables, and intervention options that would form the
most eective intervention) (Carpenter et al., 2020; Klasnja et al., 2015). In both MRTs,
participants will be randomly allocated to a tailored intervention condition or a no inter-
vention control condition at each decision point across 28-days. With three decision points
per day, each participant can be randomised up to 84 times across each MRT. In the micro-
randomisation protocol for GamblingLess: In-The-Moment, eligible participants will have a
75% chance of being micro-randomised into the tailored intervention condition and a 25%
chance of being micro-randomised into the no intervention control condition. Participants
who are eligible for two intervention modules will have a 37.5% chance of receiving either
intervention module; and participants who are eligible for all three intervention modules
will have a 25% chance of receiving any of the intervention modules. In this trial, partici-
pants who are micro-randomised to the no intervention control condition will be delivered
a brief tailored message but no intervention activities. In the micro-randomisation protocol
for Gambling Habit Hacker, eligible participants will have a 50% chance of being micro-
randomised into the tailored intervention and no intervention control condition. In this trial,
participants who are micro-randomised to the no intervention control condition will be pre-
sented with the names of the 25 self-enactable strategies but no implementation guidance.
1 3
734
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Journal of Gambling Studies (2024) 40:717–747
We selected a higher ratio of intervention allocation for GamblingLess: In-The-Moment
because we wanted, should data allow, to explore which of the three intervention options
was most benecial.
MRTs are capable of answering four critical scientic questions that may be helpful
when attempting to optimise a JITAI (Carpenter et al., 2020; Klasnja et al., 2015; Walton
et al., 2018). Specically, they can be used to inform decisions about: (1) whether or not an
intervention option should be included by exploring causal proximal eects of randomised
specic intervention components; (2) which intervention options should be included by
comparing the proximal inuence of multiple randomised specic intervention options; (3)
under what conditions individuals should be interrupted to provide an intervention option,
or one type of intervention option over another, by examining how the proximal inuence of
intervention options vary depending on the timing of support, individual internal states, and
situational contexts; and (4) when an intervention option should be delivered, or how dier-
ent intervention options should be sequenced, by exploring how the proximal inuence of
intervention options change over the duration of the treatment (Carpenter et al., 2020; Klas-
nja et al., 2015; Nahum-Shani et al., 2014; Nahum-Shani et al., 2018; Walton et al., 2018).
The primary aim of both MRTs is to explore the degree to which it is worthwhile to
deliver a tailored intervention option at a time of cognitive vulnerability (GamblingLess: In-
The-Moment) and goal vulnerability (Gambling Habit Hacker). Compared with the delivery
of no intervention, the MRTs aim to explore whether the delivery of a tailored intevention
either reduces the probability of a subsequent gambling episode and improves craving inten-
sity, self-ecacy, and positive outcome expectancies (GamblingLess: In-The-Moment) or
increases adherence to subsequent gambling expenditure limits and improves strength of
intention, goal self-ecacy, and urge self-ecacy (Gambling Habit Hacker).
Should data allow, secondary exploratory research questions for the GamblingLess: In-
The-Moment MRT include: (1) Is the delivery of one intervention option (targeting crav-
ings, self-ecacy, or positive outcome expectancies) more likely to reduce the probability
of a subsequent gambling episode than the other intervention options?; (2) How do time-
variant (EMA) factors (time of day, time of week, craving intensity, self-ecacy, positive
outcome expectancies, psychological distress, impulsivity, subjective alcohol intoxication,
readiness to change, gambling availability (nancial), gambling availability (location), and
social context) and time-invariant (pre-intervention) factors (gambling symptom severity,
gambling frequency, gambling expenditure, gender, and age) inuence the intervention
eect on the probability of a subsequent gambling episode?; and (3) How does the eect
of a tailored intervention on the probability of a subsequent gambling episode change over
the course of the 28-day MRT? We may also explore the degree to which, compared to
the delivery of no intervention, the delivery of: (a) a craving intervention reduces subse-
quent craving intensity; (b) a self-ecacy intervention increases subsequent self-ecacy;
and (c) a positive outcome expectancy intervention decreases subsequent positive outcome
expectancies. Similarly, should data allow, secondary exploratory research questions for the
Gambling Habit Hacker MRT include: (1) How do time-variant (EMA) factors (strength of
intention, goal self-ecacy, urge self-ecacy or being in a positive or negative high-risk
situation, alcohol or drug consumption, and gambling proximity) and time-invariant (pre-
intervention) factors (age, gender, volitional phase, gambling symptom severity, gambling
expenditure, and planning propensity) inuence the intervention eect on subsequent adher-
ence to gambling expenditure limits?; and (2) How does the eect of the tailored interven-
1 3
735
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Journal of Gambling Studies (2024) 40:717–747
tion on subsequent adherence to gambling expenditure limits change over the course of the
28-day MRT?
Challenges and Future Directions
We encountered several logistical and methodological challenges when developing these
JITAIs, particularly in relation to applying an MRT design to their evaluation. Some of these
challenges have been identied by other researchers (Goldstein et al., 2017). A selection of
these issues, along with future research considerations, are discussed below.
Evaluation Considerations
Evaluations of JITAIs, particularly using MRTs, are characterised by a lack of longer-
term follow-up evaluations. This is particularly problematic for these interventions, given
that part of their rationale is that they encourage the use of skills in everyday life, thereby
increasing the probability of long-lasting behaviour change (Heron & Smyth, 2010; Krebs
et al., 2010; Wang & Miller, 2020). We therefore decided to supplement the MRTs with six-
month within group follow-up evaluations of both apps to examine within-group change
over a longer period of time, as well as to identify the factors that predict these longer-
term treatment outcomes. These evaluations will also conduct supplementary analyses of
the clinical signicance of the JITAIs in terms of meaningful changes in people’s lives by
reporting eect sizes and classifying the status of each participant as recovered, improved,
unchanged, or deteriorated on outcome variables according to reliable change indices
(Jacobson & Truax, 1991).
There is also widespread consensus that successful JITAI designs requires a user-centred
and iterative approach to development, whereby both mixed methods and in-depth qualita-
tive methods can be used to rene the intervention to meet the needs of the users (Heron &
Smyth, 2010; Yardley et al., 2016). Consistent with these recommendations, GamblingLess:
In-The-Moment was built on the user, acceptability, and pilot data provided by people with
lived experience for GamblingLess: Curb Your Urge (Hawker et al., 2021; Merkouris et al.,
2020). Similarly, the perspectives of people with lived experience of gambling problems are
represented in the development of Gambling Habit Hacker, whereby the implementation
support was sourced from lived experience accounts from more than 2000 gamblers across
counselling transcripts, online forums, in-venue surveys, and community-based qualitative
and quantitative surveys (Bagot et al., 2021; Rodda et al., 2017, 2018a, c, d, 2019b; Rodda,
Hing, Rodda et al., 2018a, c, d). Prior to each evaluation, we also subjected both apps to user
testing with gambling stakeholders, with feedback indicating that they are acceptable gam-
bling interventions (Dowling et al., 2022; Rodda et al., 2022). Moreover, we will explore the
acceptability of both JITAIs with trial participants using both quantitative and qualitative
methods, including surveys at post-intervention, indices of app use and engagement, and
semi-structured interviews.
1 3
736
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Journal of Gambling Studies (2024) 40:717–747
On-demand Intervention Content
In the organising framework provided by Nahum-Shani et al. (2018), JITAIs are dened as
“push” intervention approaches, in which it is assumed that individuals are often unaware of
the emergence of states of vulnerability and/or opportunity or are unmotivated to access the
support they require to manage these states (Nahum-Shani et al., 2015; Nahum-Shani et al.,
2018). This approach contrasts with “pull” approaches to intervention delivery that require
individuals to be able to recognise states of vulnerability and be suciently motivated to
initiate access to the support they need (Klasnja et al., 2015; Walton et al., 2018).
In designing these JITAIs, we were aware that some individuals may have higher emo-
tional self-awareness than others and that the EMA protocol could enhance the recognition
of these states over time (Bakker et al., 2016; Heron and Smyth, 2010; Klasnja et al., 2015;
Walton et al., 2018). From clinical and ethical perspectives, we wanted to encourage indi-
viduals to practice the coping skills delivered by the apps in their everyday lives when they
recognise states of vulnerability and/or opportunity and are motivated to access support,
with a view to enhancing the generalisation of learned skills to new settings and maintain
therapeutic gains (Bakker et al., 2016; Heron and Smyth, 2010; Klasnja et al., 2015; Walton
et al., 2018). Although they exclude interventions that solely rely on individuals initiating
and selecting from available support options, Nahum-Shani et al. (2018) acknowledge that
the addition of some “participant-determined features” to a JITAI may provide some advan-
tages. Adding such features may accommodate conditions in which individuals are in the
best position to know when support is required and what type of support would be helpful,
facilitate autonomy through agency and control, and reduce disruption as long as they do not
access support when they are unreceptive (Fukuoka et al., 2012; Nahum-Shani et al., 2018).
Nahum-Shani et al. (2018) suggest, however, that further research is required to evaluate
how to best add these features to a JITAI to ensure that planned and externally-initiated sup-
port is balanced with personal volition.
In designing both JITAIs, we considered allowing individuals to initiate decision points,
in addition to the three protocol-driven decision points each day. Participant-initiated deci-
sion points are those in which the user requests support or when the user accesses the
intervention content on-demand (Nahum-Shani et al., 2014). Participant-initiated decision
points have been employed in previous JITAIs (Ben-Zeev et al., 2014; Businelle et al., 2016;
Franklin et al., 2008; Free et al., 2011; Gustafson et al., 2014). In this model, however, the
micro-randomisation protocol would not be applied to participant-initiated decision points.
Moreover, the on-demand intervention content would be untailored and easier to access than
that delivered via the more time-intensive EMAs, which would likely impact on the feasi-
bility of the MRTs. We also considered allowing participants to access tailored intervention
content via participant-initiated EMAs (Nahum-Shani et al., 2014), in addition to the three
protocol-driven EMAs each day, and subjecting both types of EMAs to the same MRT pro-
tocol. In this design, EMAs can be initiated by the participant in addition to the automated
system (Businelle et al., 2016; Free et al., 2011). This design, however, may also inuence
the feasibility of the MRTs as eligible participants who are randomly assigned to the no
intervention condition at any given decision point can immediately access the intervention
content via the on-demand feature.
Klasjna et al. (2015) argue that MRTs are only appropriate for evaluating push interven-
tions and are not appropriate for evaluating pull intervention components. In contrast, Wal-
1 3
737
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Journal of Gambling Studies (2024) 40:717–747
ton et al. (2018) argues that MRTs can be used to evaluate both push and pull interventions,
but focus their discussion on push interventions. On balance, we decided to maintain the
integrity of the MRT evaluations by excluding access to intervention content on-demand
during the MRT period, then allowing access to this content via participant-initiated EMAs
during the six-month follow-up periods when participants recognise they are in a state of
vulnerability and are motivated to access support. We will explore whether trial participants
prefer a more traditional pull approach or the addition of participant-determined features for
on-demand intervention content in the acceptability evaluations for both apps.
Intervention Optimisation
The optimisation of these JITAIs may involve us removing less eective components, and
determining when and in what contexts dierent treatments should be oered to maximise
eciency and minimise burden ((Klasnja et al., 2015; Walton et al., 2018). Moreover, for
both apps, we will be able to rene the decision rules by identifying appropriate timing of
the intervention and cut-points for each tailoring variable. An RCT to evaluate the interven-
tion compared to other interventions is appropriate only when there is sucient empirical
support for the optimal delivery of the intervention components (Carpenter et al., 2020;
Collins et al., 2009; Collins et al., 2005; Walton et al., 2018). Further, while conditional
statements are appropriate when there are relatively few statements, the complexity of the
model underpinning a JITAI expands exponentially as a result of adding additional contex-
tual considerations or intervention tailoring options (Goldstein et al., 2017). A more rig-
orous method of codifying behaviour is to develop mathematical models of the decision
process using machine learning methods (Goldstein et al., 2017). Machine learning, which
is a subeld of articial intelligence, can produce highly accurate predictive models from
large datasets and automatically adapt to new data in real time (Goldstein et al., 2017; Gus-
tafson et al., 2014; Nahum-Shani et al., 2014; Wang and Miller, 2020). Moreover, because
excessively broad conditional statements that apply to all users could result in inappropriate
intervention delivery to any specic individual (Goldstein et al., 2017), machine learning
approaches could be used to continually re-adapt decision rules for each individual over
time (Goldstein et al., 2017; Kim et al., 2019; Nahum-Shani et al., 2018; Riley et al., 2015).
Combining Active and Passive Assessments
Although EMA procedures are the gold-standard methodology for assessing dynamic inter-
nal states (Carpenter et al., 2020; Kim et al., 2019; Stone & Shiman, 1994), they require
high participant engagement and compliance, are associated with some degree of recall and
reporting bias, and possibly involve assessment reactivity (Goldstein et al., 2017; Kim et
al., 2019; Nahum-Shani et al., 2014; Nahum-Shani et al., 2018). In future versions of both
JITAIs, it may be possible to augment EMA data with some tailoring variables obtained
from passive assessments, which could reduce the burden on users, provide more contextual
information, and enhance user awareness of behaviour (Goldstein et al., 2017; Kim et al.,
2019). For example, data from sensors or other technologies to detect location (proximity
to a land-based gaming venue using GPS), social interactions (ambient noise detection),
increased heart rate (EEG), or states of unavailability or receptivity (e.g., driving, exercis-
ing, working, or sleeping), could be employed to mark states of heightened vulnerability to
1 3
738
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Journal of Gambling Studies (2024) 40:717–747
gambling episodes and expenditure (Goldstein et al., 2017; Gustafson et al., 2014; Kim et
al., 2019; Nahum-Shani et al., 2014; Wang and Miller, 2020).
Human Facilitation
Studies often provide JITAIs as only one intervention component in a treatment protocol
(Heron & Smyth, 2010), with evidence that guided m-Health interventions are generally
associated with superior treatment outcomes to unguided interventions (Baumeister et al.,
2014). Future iterations of these interventions could add personal coaches and assistants via
digital avatars (Fogg, 2007) or the involvement of coaches, guides or therapists to main-
tain engagement, motivation, and adherence to intervention requirements (Gustafson et al.,
2014; Klasnja & Pratt, 2012; Mohr et al., 2011; Yardley et al., 2016). For example, remote
coaching (in which healthcare providers review the data and work with individuals to sup-
port them in managing their conditions) and remote symptom monitoring (in which health-
care providers are alerted if concerning symptoms develop) can inform healthcare providers
about the individual’s condition and enhance the care interactions between providers and
patients (Klasnja & Pratt, 2012). We could also use these JITAIs to supplement face-to-face
or mobile psychological and behavioural therapies or deliver “booster” treatments following
these interventions to consolidate behaviour change (Heron & Smyth, 2010). Although the
evidence for social support in mHealth interventions remains unclear (Milward et al., 2018),
the addition of social support among individuals who share the same condition and goals
(peer-to-peer inuence) and peers who have succeeded in achieving similar goals (peer
modelling) could facilitate supportive social interactions and increase feelings of related-
ness and connectedness (Heron & Smyth, 2010; Klasnja & Pratt, 2012; Yardley et al., 2016).
The degree to which trial participants prefer the involvement of clinicians, guides, coaches,
peers, or digital avatars will be explored in the acceptability evaluations for both apps.
Given that unguided interventions can be eectively delivered at lower cost (Baumeister et
al., 2014; Yardley et al., 2016), however, future research is necessary to establish for whom
and when providing human support adds value.
Cost-eectiveness Evaluations
JITAIs are promoted as cost-eective solutions to health behaviour change due to their
potential to improve the ecacy of treatment and reduce overall treatment duration (Heron
& Smyth, 2010). There are, however, cost considerations such as the price of purchasing
treatment software and hardware, as well as the time needed to set up and implement them
(Heron & Smyth, 2010). Although there are some indications of the costs of JITAI treatment
(Przeworski & Newman, 2004), there are few cost-eectiveness studies designed to facili-
tate decisions about how healthcare expenditure is best spent (Agras et al., 1990). Future
research employing cost evaluation analyses that weigh up the relative costs and outcomes
of these two JITAIs with other interventions are therefore required to inform decisions about
resource allocation (Heron & Smyth, 2010).
1 3
739
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Journal of Gambling Studies (2024) 40:717–747
Transdiagnostic JITAIs
Finally, there is scope for redeveloping both apps to deliver transdiagnostic interventions
across the addictions. Transdiagnostic approaches represent a paradigm shift from the study
of discrete diagnostic categories to conceptualising psychopathology as comprising higher-
order underlying dimensions which are shared across multiple, apparently distinct, condi-
tions within individuals. The recent Component Model of Addiction Treatment suggests that
targeting enduring, but modiable, vulnerabilities that are common to all behavioural and
substance use addictions could serve to improve the ecacy, eciency, and cost-eective-
ness of treatment (Kim & Hodgins, 2018). The empirical support for the highly inuential
relapse prevention model across addictive behaviours indicates that GamblingLess: In-The-
Moment could be redeveloped to target the cognitive processes that precipitate episodes of
behaviour and use across all behavioural and substance use addictions. Similarly, Gambling
Habit Hacker forms part of a suite of implementation planning interventions across the
addictions, including gaming, alcohol and sugar (Brittain et al., 2021; Park et al., 2020;
Rodda et al., 2018c), suggesting that it could be redeveloped as a transdiagnostic interven-
tion to help people with a range of addictions set their behavioural intentions and support
adherence to their limits in real-time and real-world settings.
Conclusion
JITAIs are emerging “push” mHealth interventions that adapt the provision of the type,
amount, or timing of support to an individual’s dynamic needs. JITAIs, which have demon-
strated eectiveness across a range of health domains, are particularly suited to the treatment
of addictions. We attempted to redress our current gap in service provision by developing
two JITAIs, GamblingLess: In-The-Moment and Gambling Habit Hacker. We applied an
organising framework (Nahum-Shani et al., 2015; Nahum-Shani et al., 2014; Nahum-Shani
et al., 2018) to the development of these two JITAIs. These two interventions were designed
to meet the needs of both higher- and lower-risk gamblers to expand access to theoretically-
informed and evidence-based treatments to individuals across the continuum of gambling
risk who may be experiencing harm from their gambling.
Consistent with their respective theoretical underpinnings, we constructed dierent deci-
sion rules for each of these JITAIs. GamblingLess: In-The-Moment aimed to provide the
right type and amount of support required at times of cognitive vulnerability characterised
by high craving intensity, lowered self-ecacy, and positive outcome expectancies to reduce
the likelihood of a subsequent gambling episode, with a view to reducing gambling symp-
tom severity in the longer-term. In contrast, Gambling Habit Hacker aimed to provide the
type of support required at times of goal vulnerability characterised by low goal intention
strength, low goal self-ecacy, low urge self-ecacy, and high-risk situations to enhance
adherence to gambling expenditure limits, with a view to reducing gambling expenditure
in the longer-term. Given insucient theoretical and empirical evidence to fully construct
their decision rules, we plan to evaluate these JITAIs with MRTs so that we can obtain the
empirical data that is necessary for their optimisation. While JITAIs appear to be a promis-
ing intervention design in addiction science, we identied several key challenges and con-
siderations for future research. We therefore described the decisions, methods, and tools
1 3
740
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Journal of Gambling Studies (2024) 40:717–747
we employed in the development of GamblingLess: In-The-Moment and Gambling Habit
Hacker, as well as considerations of these challenges and future research directions, as they
are likely to apply to future JITAIs that target other addictive behaviours.
Acknowledgements The authors thank Chief Investigators George Youssef and Dan Lubman, as well as
research fellows, Kathleen Bagot, Chloe Hawker, Hannah Portogallo, Natalia Booth, Anna Thomas, Mingjun
Yang, Stephanie Dias, and George Loram for their support. The authors also thank 2and2 for their ongoing
support with the technical development of the app and their input into the technical aspects of the manuscript,
particularly Anna Goldfeder, Randy Olan, Colin Walker, Arun Ramalingam, Leon Young, and Toby Wong.
They thank Rick Loos for his assistance with the Get More Support functionality and the user-testing com-
ponent and the clinicians, researchers, and consumers who took part in the user testing. Finally, they thank
the New South Wales Oce of Responsible Gambling for its support, particularly John McInerney, Rhonda
Blackett, and Natalie Wright.
Funding Open Access funding enabled and organized by CAUL and its Member Institutions. The develop-
ment and evaluation of GamblingLess: In-The-Moment and Gambling Habit Hacker were funded by the
NSW Government’s Responsible Gambling Fund, with support from the NSW Oce of Responsible Gam-
bling. The funding body played no role in the study design or writing of the manuscript. There was no gam-
bling industry involvement in the development or evaluation of either app.
Data Availability No data was generated or analysed in this manuscript.
Declarations
Competing Interests The 3-year declaration of interest statement of this research team is as follows: All
authors have received funding from multiple sources, including government departments and the Victorian
Responsible Gambling Foundation (through hypothecated taxes from gambling revenue), and the Inter-
national Center for Responsible Gaming, a charitable organization that derives its funding from multiple
stakeholders (with funding decisions the responsibility of a scientic advisory board). SM is the recipient
of the New South Wales Oce of a Responsible Gambling Postdoctoral Fellowship. None of the authors
have received research funding from the gambling, tobacco, alcohol industries, or any industry-sponsored
organization.
Institutional Review Board Statement The evaluation of these interventions, which is currently underway,
is conducted in accordance with the Declaration of Helsinki, and approved by Deakin University Human
Research Ethics Committee (Ethics ID: 2020 − 304).
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License,
which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as
you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons
licence, and indicate if changes were made. The images or other third party material in this article are
included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material.
If material is not included in the article’s Creative Commons licence and your intended use is not permitted
by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the
copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
References
Agras, W. S., Taylor, C. B., Feldman, D. E., Losch, M., & Burnett, K. F. (1990). Developing computer-
assisted therapy for the treatment of obesity. Behavior Therapy, 21(1), 99–109.
Armitage, C. J. (2009). Eectiveness of experimenter-provided and self-generated implementation intentions
to reduce alcohol consumption in a sample of the general population: A randomized exploratory trial.
Health Psychology, 28(5), 545.
1 3
741
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Journal of Gambling Studies (2024) 40:717–747
Attwood, S., Parke, H., Larsen, J., & Morton, K. L. (2017). Using a mobile health application to reduce alco-
hol consumption: A mixed-methods evaluation of the drinkaware track & calculate units application.
Bmc Public Health, 17(1), 394.
Bagot, K. L., Cheetham, A., Lubman, D. I., & Rodda, S. N. (2021). Predictors of strategy engagement for the
prevention and reduction of gambling harm: A prospective application of the theory of planned behav-
iour. Int J Ment Health Addict, 19(5), 1812–1828.
Bakker, D., Kazantzis, N., Rickwood, D., & Rickard, N. (2016, Mar 1). Mental Health Smartphone Apps:
Review and Evidence-Based Recommendations for Future Developments. JMIR Ment Health, 3(1), e7.
https://doi.org/10.2196/mental.4984.
Baumeister, H., Reichler, L., Munzinger, M., & Lin, J. (2014). The impact of guidance on internet-based men-
tal health interventions—A systematic review. Internet Interv, 1(4), 205–215. https://doi.org/10.1016/j.
invent.2014.08.003.
Ben-Zeev, D., Brenner, C. J., Begale, M., Duecy, J., Mohr, D. C., & Mueser, K. T. (2014). Nov). Feasibility,
acceptability, and preliminary ecacy of a smartphone intervention for schizophrenia. Schizophrenia
Bulletin, 40(6), 1244–1253. https://doi.org/10.1093/schbul/sbu033.
Bijker, R., Booth, N., Merkouris, S. S., Dowling, N. A., & Rodda, S. N. (2022). Global prevalence of help‐
seeking for problem gambling: A systematic review and meta‐analysis. Addiction, 11 7(12), 2972–
2985. https://doi.org/10.1111/add.15952
Brendryen, H., & Kraft, P. (2008, Mar). Happy ending: A randomized controlled trial of a digital multi-
media smoking cessation intervention. Addiction, 103(3), 478–484. https://doi.org/10.1111/j.1360-
0443.2007.02119.x. discussion 485 – 476.
Brendryen, H., Drozd, F., & Kraft, P. (2008). A digital smoking cessation program delivered through internet
and cell phone without nicotine replacement (happy ending): Randomized controlled trial. Journal of
Medical Internet Research, 10(5), e51.
Brittain, M., Consedine, N., Bagot, K. L., Booth, N., & Rodda, S. N. (2021). Sugar habit Hacker: Initial evi-
dence that a planning intervention reduces sugar intake. J Behav Addict, 10(3), 471–481.
Browne, M., Greer, N., Rawat, V., & Rocklo, M. (2017). A population-level metric for gambling-related
harm. International Gambling Studies, 17(2), 163–175. https://doi.org/10.1080/14459795.2017.13049
73.
Businelle, M. S., Ma, P., Kendzor, D. E., Frank, S. G., Vidrine, D. J., & Wetter, D. W. (2016). Dec 12). An
ecological momentary intervention for Smoking Cessation: Evaluation of feasibility and eectiveness.
Journal of Medical Internet Research, 18(12), e321. https://doi.org/10.2196/jmir.6058.
Carpenter, S. M., Menictas, M., Nahum-Shani, I., Wetter, D. W., & Murphy, S. A. (2020). Sep). Develop-
ments in mobile health just-in-Time adaptive interventions for addiction science. Curr Addict Rep, 7(3),
280–290. https://doi.org/10.1007/s40429-020-00322-y.
Collins, L. M., & Kugler, K. C. (2018). Optimization of behavioral, biobehavioral, and biomedical interven-
tions (10 vol.). Springer International Publishing.
Collins, L. M., Murphy, S. A., & Bierman, K. L. (2004). Sep). A conceptual framework for adaptive preventive
interventions. Prevention Science, 5(3), 185–196. https://doi.org/10.1023/b:prev.0000037641.26017.00.
Collins, L. M., Murphy, S. A., Nair, V. N., & Strecher, V. J. (2005). Aug). A strategy for optimizing and evalu-
ating behavioral interventions. Annals of Behavioral Medicine, 30(1), 65–73. https://doi.org/10.1207/
s15324796abm3001_8.
Collins, L. M., Chakraborty, B., Murphy, S. A., & Strecher, V. (2009, Feb). Comparison of a phased experi-
mental approach and a single randomized clinical trial for developing multicomponent behavioral inter-
ventions. Clinical Trials (London, England), 6(1), 5–15. https://doi.org/10.1177/1740774508100973.
Coral, R., Esposito, F., & Weinstock, J. (2020). Don’t Go There: A Zero-Permission Geofencing App to
Alleviate Gambling Disorders. 2020 IEEE 17th Annual Consumer Communications & Networking
Conference (CCNC).
Cowlishaw, S., Merkouris, S., Dowling, N., Anderson, C., Jackson, A., & Thomas, S. (2012). Nov 14). Psy-
chological therapies for pathological and problem gambling. Cochrane Database Systematic Review,
11, CD008937. https://doi.org/10.1002/14651858.CD008937.pub2.
de Lisle, S. M., Dowling, N. A., & Allen, J. S. (2012, Dec). Mindfulness and problem gambling: A review of
the literature. Journal of Gambling Studies, 28(4), 719–739. https://doi.org/10.1007/s10899-011-9284-7.
Deci, E. L., & Ryan, R. M. (2008). Self-determination theory: A macrotheory of human motivation, develop-
ment, and health. Canadian Psychology/Psychologie Canadienne, 49(3), 182.
Dowling, N., & Smith, D. (2007, Sep). Treatment goal selection for female pathological gambling: A com-
parison of abstinence and controlled gambling. Journal of Gambling Studies, 23(3), 335–345. https://
doi.org/10.1007/s10899-007-9064-6.
Dowling, N., Smith, D., & Thomas, T. (2009). Jun). A preliminary investigation of abstinence and controlled
gambling as self-selected goals of treatment for female pathological gambling. Journal of Gambling
Studies, 25(2), 201–214. https://doi.org/10.1007/s10899-009-9116-1.
1 3
742
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Journal of Gambling Studies (2024) 40:717–747
Dowling, N., Merkouris, S., Rodda, S., Smith, D., Lavis, T., Lubman, D., … Battersby, M. (2018). Develop-
ment and evaluation of an online gambling self-directed program: eective integration into existing
services. Victorian Responsible Gambling Foundation.
Dowling, N. A., Merkouris, S. S., & Spence, K. (2020). Ecological momentary assessment of the relation-
ship between outcome expectancies and gambling behaviour. J Clin Med, 10(8), 1709. https://doi.
org/10.3390/jcm10081709.
Dowling, N. A., Merkouris, S. S., Rodda, S. N., Youssef, G., Bagot, K., Portogallo, H., Lubman, D., &
Hawker, C. (2022). GAMBLINGLESS IN-THE-MOMENT: Protocol for a micro-randomised trial of
gambling Just-In-Time Adaptive Intervention. JMIR research protocols. https://doi.org/10.2196/38958.
Dowling, N. A., Merkouris, S. S., Rodda, S. N., Smith, D., Aarsman, S., Lavis, T., Lubman, D. I., Austin, D.
W., Cunningham, J. A., & Battersby, M. W. (2021, May 21). & O, S. C. GamblingLess: A randomised
trial comparing guided and unguided internet-based gambling interventions. J Clin Med, 10(11), 2224.
https://doi.org/10.3390/jcm10112224.
Fogg, B. (2007). Mobile persuasion: 20 perspectives on the future of behavior change. Mobile Persuasion.
Franklin, V., Greene, A., Waller, A., Greene, S., & Pagliari, C. (2008). Patients’ engagement with Sweet
Talk–a text messaging support system for young people with diabetes. Journal of Medical Internet
Research, 10(2), e20.
Free, C., Knight, R., Robertson, S., Whittaker, R., Edwards, P., Zhou, W., Rodgers, A., Cairns, J., Kenward,
M. G., & Roberts, I. (2011). Jul 2). Smoking cessation support delivered via mobile phone text mes-
saging (txt2stop): A single-blind, randomised trial. Lancet, 378(9785), 49–55. https://doi.org/10.1016/
S0140-6736(11)60701-0.
Fukuoka, Y., Lindgren, T., & Jong, S. (2012). May-Jun). Qualitative exploration of the acceptability of a
mobile phone and pedometer-based physical activity program in a diverse sample of sedentary women.
Public Health Nursing, 29(3), 232–240. https://doi.org/10.1111/j.1525-1446.2011.00997.x.
Goldstein, S. P., Evans, B. C., Flack, D., Juarascio, A., Manasse, S., Zhang, F., & Forman, E. M. (2017, Oct).
Return of the JITAI: Applying a just-in-Time adaptive intervention Framework to the development
of m-Health solutions for addictive behaviors. International Journal of Behavioral Medicine, 24(5),
673–682. https://doi.org/10.1007/s12529-016-9627-y.
Gollwitzer, P. M. (1999). Implementation intentions: Strong eects of simple plans. American Psychologist,
54(7), 493.
Gonzalez, V. M., & Dulin, P. L. (2015). Comparison of a smartphone app for alcohol use disorders with an
internet-based intervention plus bibliotherapy: A pilot study. Journal of Consulting and Clinical Psy-
chology, 83(2), 335.
Gooding, P., & Tarrier, N. (2009). Jul). A systematic review and meta-analysis of cognitive-behavioural
interventions to reduce problem gambling: Hedging our bets? Behaviour Research and Therapy, 47(7),
592–607. https://doi.org/10.1016/j.brat.2009.04.002.
Goslar, M., Leibetseder, M., Muench, H. M., Hofmann, S. G., & Laireiter, A. R. (2017, Jun 1). Ecacy of
face-to-face versus self-guided treatments for disordered gambling: A meta-analysis. J Behav Addict,
6(2), 142–162. https://doi.org/10.1556/2006.6.2017.034.
Gustafson, D. H., McTavish, F. M., Chih, M. Y., Atwood, A. K., Johnson, R. A., Boyle, M. G., Levy, M.
S., Driscoll, H., Chisholm, S. M., Dillenburg, L., Isham, A., & Shah, D. (2014). May). A smartphone
application to support recovery from alcoholism: A randomized clinical trial. JAMA Psychiatry, 71(5),
566–572. https://doi.org/10.1001/jamapsychiatry.2013.4642.
Hamilton, K., Keech, J. J., Peden, A. E., & Hagger, M. S. (2019). Protocol for developing a mental imagery
intervention: A randomised controlled trial testing a novel implementation imagery e-health interven-
tion to change driver behaviour during oods. British Medical Journal Open, 9(2), e025565.
Hawker, C., Merkouris, S., Youssef, G., & Dowling, N. (2020). Exploring the associations between gambling
cravings, self-ecacy, and gambling episodes: An ecological momentary Assessment study. Addictive
Behaviors, 112 , 106574. https://doi.org/10.1016/j.addbeh.2020.106574.
Hawker, C. O., Merkouris, S. S., Youssef, G. J., & Dowling, N. A. (2021). Mar 26). A smartphone-delivered
ecological momentary intervention for problem gambling (GamblingLess: Curb your urge): Single-
arm acceptability and feasibility trial. Journal of Medical Internet Research, 23(3), e25786. https://doi.
org/10.2196/25786.
Hebert, E. T., Ra, C. K., Alexander, A. C., Helt, A., Moisiuc, R., Kendzor, D. E., Vidrine, D. J., Funk-Lawler,
R. K., & Businelle, M. S. (2020). Mar 9). A mobile just-in-time adaptive intervention for smoking ces-
sation: Pilot randomized controlled trial. Journal of Medical Internet Research, 22(3), e16907. https://
doi.org/10.2196/16907.
Hekler, E. B., Klasnja, P., Traver, V., & Hendriks, M. (2013, Sep). Realizing eective behavioral manage-
ment of health: The metamorphosis of behavioral science methods. IEEE Pulse, 4(5), 29–34. https://doi.
org/10.1109/MPUL.2013.2271681.
1 3
743
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Journal of Gambling Studies (2024) 40:717–747
Heron, K. E., & Smyth, J. M. (2010). Ecological momentary interventions: Incorporating mobile technology
into psychosocial and health behaviour treatments. The British Journal of Health Psychology, 15(1),
1–39. https://doi.org/10.1348%2F135910709X466063.
Hing, N., Browne, M., Russell, A. M., Rocklo, M., Rawat, V., Nicoll, F., & Smith, G. (2019). Avoiding
gambling harm: An evidence-based set of safe gambling practices for consumers. PloS One, 14(10),
e0224083.
Humphrey, G., Newcombe, D., Whittaker, R., Parag, V., & Bulleen, C. (2019). SPGeTTI: A smartphone-
based problem gambling evaluation and technology testing initiative nal report. National Institute
for Health Innovation &. Prepared for the Ministry of Health. https://www.health.govt.nz/publication/
spgetti-smartphone-based-problem-gambling-evaluation-and-technology-testing-initiative.
Humphrey, G., Chu, J., Dowling, N., Rodda, S., Merkouris, S., Parag, V., Newcombe, D., Ho, E., Nosa, V.,
& Ruwhui-Collins, R. (2020). Manaaki - A cognitive behavioral therapy mobile health app to sup-
port people experiencing gambling problems: A randomized control trial protocol. Bmc Public Health,
20(1), 1–11. https://doi.org/10.1186/s12889-020-8304-x.
Humphrey, G., Chu, J. T., Ruwhiu-Collins, R., Erick-Peleti, S., Dowling, N., Merkouris, S., Newcombe,
D., Rodda, S., Ho, E., & Nosa, V. (2022). Adapting an evidence-based e-Learning cognitive behav-
ioral therapy program into a Mobile app for people experiencing gambling-related problems: Formative
study. JMIR Formative Research, 6(3), e32940. https://doi.org/10.2196/32940.
Jacobson, N. S., & Truax, P. (1991, Feb). Clinical signicance: A statistical approach to dening meaning-
ful change in psychotherapy research. Journal of Consulting and Clinical Psychology, 59(1), 12–19.
https://doi.org/10.1037//0022-006x.59.1.12.
Kelly, J., Gooding, P., Pratt, D., Ainsworth, J., Welford, M., & Tarrier, N. (2012, Aug). Intelligent real-time
therapy: Harnessing the power of machine learning to optimise the delivery of momentary cognitive-
behavioural interventions. Journal of Mental Health (Abingdon, England), 21(4), 404–414. https://doi.
org/10.3109/09638237.2011.638001.
Khazaal, Y., Monney, G., Richter, F., & Achab, S. (2017). «Jeu-contrôle», rationnel d’une application de
soutien aux limites de jeux. Journal de thérapie Comportementale et Cognitive, 27(3), 129–137. https://
doi.org/10.1016/j.jtcc.2017.05.003.
Kim, H. S., & Hodgins, D. C. (2018). Component model of Addiction Treatment: A pragmatic transdiagnos-
tic treatment model of behavioral and substance addictions. Frontiers in Psychiatry, 9, 406. https://doi.
org/10.3389/fpsyt.2018.00406.
Kim, J., Marcusson-Clavertz, D., Yoshiuchi, K., & Smyth, J. M. (2019). Potential benets of integrating
ecological momentary assessment data into mHealth care systems. BioPsychoSocial Medicine, 13(1),
19. https://doi.org/10.1186/s13030-019-0160-5.
Klasnja, P., & Pratt, W. (2012, Feb). Healthcare in the pocket: Mapping the space of mobile-phone health inter-
ventions. Journal of Biomedical Informatics, 45(1), 184–198. https://doi.org/10.1016/j.jbi.2011.08.017.
Klasnja, P., Hekler, E. B., Shiman, S., Boruvka, A., Almirall, D., Tewari, A., & Murphy, S. A. (2015, Dec).
Microrandomized trials: An experimental design for developing just-in-time adaptive interventions.
Health Psychology, 34S(S), 1220–1228. https://doi.org/10.1037/hea0000305.
Knäuper, B., Roseman, M., Johnson, P. J., & Krantz, L. H. (2009). Using mental imagery to enhance the
eectiveness of implementation intentions. Current Psychol, 28(3), 181–186.
Krebs, P., Prochaska, J. O., & Rossi, J. S. (2010). Sep-Oct). A meta-analysis of computer-tailored interven-
tions for health behavior change. Preventive Medicine, 51(3–4), 214–221. https://doi.org/10.1016/j.
ypmed.2010.06.004.
Kreyenbuhl, J., Nossel, I. R., & Dixon, L. B. (2009). Disengagement from mental health treatment among
individuals with schizophrenia and strategies for facilitating connections to care: A review of the litera-
ture. Schizophrenia Bulletin, 35(4), 696–703.
Ladouceur, R. (2005, Spring). Controlled gambling for pathological gamblers. Journal of Gambling Studies,
21(1), 49–59. https://doi.org/10.1007/s10899-004-1923-9.
Larimer, M. E., Palmer, R. S., & Marlatt, G. A. (1999). Relapse prevention. An overview of Mar-
latt’s cognitive-behavioral model. Alcohol Research & Health, 23(2), 151–160. https://doi.
org/10.4324/9780203503508-5.
Malaguti, A., Ciocanel, O., Sani, F., Dillon, J. F., Eriksen, A., & Power, K. (2020). Eectiveness of the use
of implementation intentions on reduction of substance use: A meta-analysis. Drug and Alcohol Depen-
dence, 214, 108120.
Marlatt, A. G., & Gordon, J. (1985). Relapse Prevention: Maintenance strategies in the treatment of addictive
behaviours. Guilford Press.
Marlatt, G. A., & Witkiewitz, K. (2005). Relapse prevention for alcohol and drug problems. Donovan (eds.)
Relapse prevention: Maintenance strategies in the treatment of addictive behaviors. Guilford Publica-
tions. G.A Marlatt & D.M
1 3
744
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Journal of Gambling Studies (2024) 40:717–747
Maynard, B. R., Wilson, A. N., Labuzienski, E., & Whiting, S. W. (2018). Mindfulness-based approaches
in the treatment of disordered gambling: A systematic review and meta-analysis. Res Soc Work Pract,
28(3), 348–362. https://doi.org/10.1177/1049731515606977.
McWilliams, L., Bellhouse, S., Yorke, J., Lloyd, K., & Armitage, C. J. (2019). Beyond planning: A meta-
analysis of implementation intentions to support smoking cessation. Health Psychology, 38(12), 1059.
Menon, J., & Kandasamy, A. (2018, Feb). Relapse prevention. Indian J Psychiatry, 60(Suppl 4), S473–S478.
https://doi.org/10.4103/psychiatry.IndianJPsychiatry_36_18.
Merkouris, S. S., Rodda, S. N., Austin, D., Lubman, D. I., Harvey, P., Battersby, M., Cunningham, J., Lavis,
T., Smith, D., & Dowling, N. A. (2017). Feb 23). GAMBLINGLESS FOR LIFE study protocol: A prag-
matic randomised trial of an online cognitive-behavioural programme for disordered gambling. British
Medical Journal Open, 7(2), e014226. https://doi.org/10.1136/bmjopen-2016-014226.
Merkouris, S. S., Hawker, C. O., Rodda, S. N., Youssef, G. J., & Dowling, N. A. (2020). GamblingLess:
Curb your urge: Development and usability testing of a smartphone-delivered ecological momentary
intervention for problem gambling. International Gambling Studies, 1–24. https://doi.org/10.1080/144
59795.2020.1749293.
Michie, S., Richardson, M., Johnston, M., Abraham, C., Francis, J., Hardeman, W., Eccles, M. P., Cane, J., &
Wood, C. E. (2013). Aug). The behavior change technique taxonomy (v1) of 93 hierarchically clustered
techniques: Building an international consensus for the reporting of behavior change interventions.
Annals of Behavioral Medicine, 46(1), 81–95. https://doi.org/10.1007/s12160-013-9486-6.
Milward, J., Drummond, C., Fincham-Campbell, S., & Deluca, P. (2018). What makes online sub-
stance-use interventions engaging? A systematic review and narrative synthesis. Digital Health, 4,
2055207617743354.
Mohr, D. C., Cuijpers, P., & Lehman, K. (2011, Mar 10). Supportive accountability: a model for providing
human support to enhance adherence to eHealth interventions. J Med Internet Res, 13(1), e30. https://
doi.org/10.2196/jmir.1602.
Moody, L. N., Tegge, A. N., Poe, L. M., Koarnus, M. N., & Bickel, W. K. (2018). To drink or to drink less?
Distinguishing between eects of implementation intentions on decisions to drink and how much to
drink in treatment-seeking individuals with alcohol use disorder. Addictive Behaviors, 83, 64–71.
Nahum-Shani, I., Smith, S., Tewari, A., Witkiewitz, K., Collins, L., Spring, B., & Murphy, S. (2014).
Just-in-Time Adaptive Interventions (JITAIs): an organizing framework for ongoing health behavior
support:(Technical Report No. 14–126). University Park, PA: The Methodology Center, Penn State.
https://doi.org/10.1007/s12160-016-9830-8.
Nahum-Shani, I., Smith, S. N., Spring, B. J., Collins, L. M., Witkiewitz, K., Tewari, A., & Murphy, S. A.
(2018). May 18). Just-in-Time adaptive interventions (JITAIs) in mobile health: Key components and
design principles for ongoing health behavior support. Annals of Behavioral Medicine, 52(6), 446–462.
https://doi.org/10.1007/s12160-016-9830-8.
Nahum-Shani, I., Hekler, E. B., & Spruijt-Metz, D. (2015, Dec). Building health behavior models to guide
the development of just-in-time adaptive interventions: A pragmatic framework. Health Psychology,
34S(S), 1209–1219. https://doi.org/10.1037/hea0000306.
Naughton, F., Hopewell, S., Lathia, N., Schalbroeck, R., Brown, C., Mascolo, C., McEwen, A., & Sutton,
S. (2016). Sep 16). A context-sensing mobile phone app (Q sense) for smoking cessation: A mixed-
methods study. JMIR mHealth uHealth, 4(3), e106. https://doi.org/10.2196/mhealth.5787.
Ntoumanis, N., Ng, J. Y., Prestwich, A., Quested, E., Hancox, J. E., Thøgersen-Ntoumani, C., Deci, E. L.,
Ryan, R. M., Lonsdale, C., & Williams, G. C. (2021). A meta-analysis of self-determination theory-
informed intervention studies in the health domain: Eects on motivation, health behavior, physical,
and psychological health. Health Psychol Rev, 15(2), 214–244.
Park, J. J., Booth, N., Bagot, K. L., & Rodda, S. N. (2020). A brief internet-delivered intervention for the
reduction of gaming-related harm: A feasibility study. Comput Hum Behav Reports, 2, 100027.
Przeworski, A., & Newman, M. G. (2004, Feb). Palmtop computer-assisted group therapy for social phobia.
Journal of Clinical Psychology, 60(2), 179–188. https://doi.org/10.1002/jclp.10246.
Riley, W., Obermayer, J., & Jean-Mary, J. (2008). Sep-Oct). Internet and mobile phone text messaging
intervention for college smokers. Journal of American College Health, 57(2), 245–248. https://doi.
org/10.3200/JACH.57.2.245-248.
Riley, W. T., Rivera, D. E., Atienza, A. A., Nilsen, W., Allison, S. M., & Mermelstein, R. (2011). Health
behavior models in the age of mobile interventions: Are our theories up to the task? Transl Behav Med,
1(1), 53–71.
Riley, W. T., Serrano, K. J., Nilsen, W., & Atienza, A. A. (2015, Oct 1). Mobile and Wireless Technologies in
Health Behavior and the Potential for Intensively Adaptive Interventions. Curr Opin Psychol, 5, 67–71.
https://doi.org/10.1016/j.copsyc.2015.03.024.
1 3
745
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Journal of Gambling Studies (2024) 40:717–747
Rodda, S. N., Hing, N., Hodgins, D. C., Cheetham, A., Dickins, M., & Lubman, D. I. (2017). Change strate-
gies and associated implementation challenges: An analysis of online counselling sessions. Journal of
Gambling Studies, 33(3), 955–973.
Rodda, S. N., Bagot, K. L., Cheetham, A., Hodgins, D. C., Hing, N., & Lubman, D. I. (2018). Types of change
strategies for limiting or reducing gambling behaviors and their perceived helpfulness: A factor analy-
sis. Psychology of Addictive Behaviors, 32(6), 679.
Rodda, S., Merkouris, S. S., Abraham, C., Hodgins, D. C., Cowlishaw, S., & Dowling, N. A. (2018a). Thera-
pist-delivered and self-help interventions for gambling problems: A review of contents. J Behav Addict,
7(2), 211–226.
Rodda, S. N., Booth, N., Landon, J., & Lubman, D. I. (2018c). Come up with a plan for booze free days:
Behaviour change strategies for limiting or reducing alcohol. Drug and Alcohol Review.
Rodda, S. N., Hing, N., Hodgins, D. C., Cheetham, A., Dickins, M., & Lubman, D. I. (2018d). Behaviour
change strategies for problem gambling: An analysis of online posts. International Gambling Studies,
18(3), 420–438.
Rodda, S. N., Bagot, K. L., Manning, V., & Lubman, D. I. (2019a). It was terrible. I didn’t set a limit:
Proximal and distal prevention strategies for reducing the risk of a bust in gambling venues. Journal of
Gambling Studies, 35(4), 1407–1421.
Rodda, S. N., Bagot, K. L., Manning, V., & Lubman, D. I. (2019b). Only take the money you want to lose’
strategies for sticking to limits in electronic gaming machine venues. International Gambling Studies,
19(3), 489–507.
Rodda, S. N., Bagot, K. L., Manning, V., & Lubman, D. I. (2020). An exploratory RCT to support gamblers’
intentions to stick to monetary limits: A brief intervention using action and coping planning. Journal of
Gambling Studies, 36(1), 387–404.
Rodda, S. N., Bagot, K. L., Merkouris, S. S., Youssef, G., Lubman, D. I., Thomas, A. C., & Dowling, N. A.
(2022). 2022/7/26). Smartphone app delivery of a Just-In-Time adaptive intervention for adult gam-
blers (gambling habit Hacker): Protocol for a microrandomized trial. JMIR Research Protocols, 11(7),
e38919. https://doi.org/10.2196/38919.
Rodda, S. N., Merkouris, S., Lavis, T., Smith, D., Lubman, D. I., Austin, D., Harvey, P., Battersby, M., &
Dowling, N. A. (2019, Dec). The therapist experience of internet delivered CBT for problem gam-
bling: Service integration considerations. Internet Interv, 18, 100264. https://doi.org/10.1016/j.
invent.2019.100264.
Rodgers, A., Corbett, T., Bramley, D., Riddell, T., Wills, M., Lin, R. B., & Jones, M. (2005). Do u smoke after
txt? Results of a randomised trial of smoking cessation using mobile phone text messaging. Tobacco
Control, 14(4), 255–261.
Schwarzer, R. (2008). Modeling health behavior change: How to predict and modify the adoption and main-
tenance of health behaviors. Applied Psychology, 57(1), 1–29.
Schwarzer, R., & Luszczynska, A. (2008). How to overcome health-compromising behaviors: The health
action process approach. European Psychologist, 13(2), 141–151.
Sheeran, P., Wright, C. E., Avishai, A., Villegas, M. E., Lindemans, J. W., Klein, W. M., Rothman, A. J.,
Miles, E., & Ntoumanis, N. (2020). Self-determination theory interventions for health behavior change:
Meta-analysis and meta-analytic structural equation modeling of randomized controlled trials. Journal
of Consulting and Clinical Psychology, 88(8), 726.
Shiman, S. (2009). Ecological momentary assessment (EMA) in studies of substance use. Psychological
Assessment, 21(4), 486–497. https://doi.org/10.1037/a0017074.
Shiman, S., Stone, A. A., & Huord, M. R. (2008). Ecological momentary assessment. Annu Rev Clin Psy-
chol, 4, 1–32. https://doi.org/10.1146/annurev.clinpsy.3.022806.091415.
Sniehotta, F. F., Schwarzer, R., Scholz, U., & Schüz, B. (2005). Action planning and coping planning for
long-term lifestyle change: Theory and assessment. Eur J Soc Psychol, 35(4), 565–576.
Spruijt-Metz, D., & Nilsen, W. (2014). Dynamic models of behavior for just-in-time adaptive interventions.
IEEE Pervasive Computing, 13(3), 13–17.
Stone, A. A., & Shiman, S. (1994). Ecological momentary assessment (EMA) in behavorial medicine.
Annals of Behavioral Medicine.
Suoletto, B., Callaway, C., Kristan, J., Kraemer, K., & Clark, D. B. (2012, Mar). Text-message-
based drinking assessments and brief interventions for young adults discharged from the emer-
gency department. Alcoholism, Clinical and Experimental Research, 36(3), 552–560. https://doi.
org/10.1111/j.1530-0277.2011.01646.x.
Sutton, S. (2008). How does the health action process approach (HAPA) bridge the intention–behavior gap?
An examination of the model’s causal structure. Applied Psychology, 57(1), 66–74.
Vidrine, D. J., Arduino, R. C., Lazev, A. B., & Gritz, E. R. (2006). Jan 9). A randomized trial of a proactive
cellular telephone intervention for smokers living with HIV/AIDS. Aids (London, England), 20(2),
253–260. https://doi.org/10.1097/01.aids.0000198094.23691.58.
1 3
746
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Journal of Gambling Studies (2024) 40:717–747
Walton, A., Nahum-Shani, I., Crosby, L., Klasnja, P., & Murphy, S. (2018). Jul). Optimizing Digital Inte-
grated Care via Micro-Randomized trials. Clinical Pharmacology and Therapeutics, 104(1), 53–58.
https://doi.org/10.1002/cpt.1079.
Wang, L., & Miller, L. C. (2020). Nov). Just-in-the-moment adaptive interventions (JITAI): A Meta-Analyt-
ical Review. Health Communication, 35(12), 1531–1544. https://doi.org/10.1080/10410236.2019.165
2388.
Weinstock, J., Whelan, J. P., & Meyers, A. W. (2004, Mar). Behavioral assessment of gambling: An appli-
cation of the timeline followback method. Psychological Assessment, 16(1), 72–80. https://doi.
org/10.1037/1040-3590.16.1.72.
Weitzel, J. A., Bernhardt, J. M., Usdan, S., Mays, D., & Glanz, K. (2007). Using wireless handheld computers
and tailored text messaging to reduce negative consequences of drinking alcohol. J Stud Alcohol Drugs,
68(4), 534–537.
Witkiewitz, K., & Marlatt, G. A. (2004). May-Jun). Relapse prevention for alcohol and drug problems: That was
Zen, this is Tao. American Psychologist, 59(4), 224–235. https://doi.org/10.1037/0003-066X.59.4.224.
Witkiewitz, K., Desai, S. A., Bowen, S., Leigh, B. C., Kirouac, M., & Larimer, M. E. (2014). Development
and evaluation of a mobile intervention for heavy drinking and smoking among college students. Psy-
chology of Addictive Behaviors, 28(3), 639.
World Health Organization [WHO] (2017). mHealth: Use of appropriate digital technologies for public
health: report by the Director-General. https://apps.who.int/iris/bitstream/handle/10665/274134/
B142_20-en.pdf?sequence=1&isAllowed=y
Yardley, L., Spring, B. J., Riper, H., Morrison, L. G., Crane, D. H., Curtis, K., Merchant, G. C., Naughton,
F., & Blandford, A. (2016). Nov). Understanding and promoting eective Engagement with Digital
Behavior Change Interventions. American Journal of Preventive Medicine, 51(5), 833–842. https://doi.
org/10.1016/j.amepre.2016.06.015.
Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and
institutional aliations.
1 3
747
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
1.
2.
3.
4.
5.
6.
Terms and Conditions
Springer Nature journal content, brought to you courtesy of Springer Nature Customer Service Center
GmbH (“Springer Nature”).
Springer Nature supports a reasonable amount of sharing of research papers by authors, subscribers
and authorised users (“Users”), for small-scale personal, non-commercial use provided that all
copyright, trade and service marks and other proprietary notices are maintained. By accessing,
sharing, receiving or otherwise using the Springer Nature journal content you agree to these terms of
use (“Terms”). For these purposes, Springer Nature considers academic use (by researchers and
students) to be non-commercial.
These Terms are supplementary and will apply in addition to any applicable website terms and
conditions, a relevant site licence or a personal subscription. These Terms will prevail over any
conflict or ambiguity with regards to the relevant terms, a site licence or a personal subscription (to
the extent of the conflict or ambiguity only). For Creative Commons-licensed articles, the terms of
the Creative Commons license used will apply.
We collect and use personal data to provide access to the Springer Nature journal content. We may
also use these personal data internally within ResearchGate and Springer Nature and as agreed share
it, in an anonymised way, for purposes of tracking, analysis and reporting. We will not otherwise
disclose your personal data outside the ResearchGate or the Springer Nature group of companies
unless we have your permission as detailed in the Privacy Policy.
While Users may use the Springer Nature journal content for small scale, personal non-commercial
use, it is important to note that Users may not:
use such content for the purpose of providing other users with access on a regular or large scale
basis or as a means to circumvent access control;
use such content where to do so would be considered a criminal or statutory offence in any
jurisdiction, or gives rise to civil liability, or is otherwise unlawful;
falsely or misleadingly imply or suggest endorsement, approval , sponsorship, or association
unless explicitly agreed to by Springer Nature in writing;
use bots or other automated methods to access the content or redirect messages
override any security feature or exclusionary protocol; or
share the content in order to create substitute for Springer Nature products or services or a
systematic database of Springer Nature journal content.
In line with the restriction against commercial use, Springer Nature does not permit the creation of a
product or service that creates revenue, royalties, rent or income from our content or its inclusion as
part of a paid for service or for other commercial gain. Springer Nature journal content cannot be
used for inter-library loans and librarians may not upload Springer Nature journal content on a large
scale into their, or any other, institutional repository.
These terms of use are reviewed regularly and may be amended at any time. Springer Nature is not
obligated to publish any information or content on this website and may remove it or features or
functionality at our sole discretion, at any time with or without notice. Springer Nature may revoke
this licence to you at any time and remove access to any copies of the Springer Nature journal content
which have been saved.
To the fullest extent permitted by law, Springer Nature makes no warranties, representations or
guarantees to Users, either express or implied with respect to the Springer nature journal content and
all parties disclaim and waive any implied warranties or warranties imposed by law, including
merchantability or fitness for any particular purpose.
Please note that these rights do not automatically extend to content, data or other material published
by Springer Nature that may be licensed from third parties.
If you would like to use or distribute our Springer Nature journal content to a wider audience or on a
regular basis or in any other manner not expressly permitted by these Terms, please contact Springer
Nature at
onlineservice@springernature.com
