Review method and strategy
This project undertakes formative research to design and test
‘TraitMap’, a personalised mobile app designed to detect cravings and
intervene with self-monitoring and feedback as cravings develop. Thus
countering many of the harms associated with drug use. To achieve these
aims, a 3-stage app development project is proposed. First,
multi-stakeholder feedback is collated via online surveys. Second, a
mobile app ‘TraitMap’ is designed to respond to trait vulnerabilities in
people with M/SUDs. Finally, a pilot randomised controlled trial
examines the feasibility of TraitMap to reduce drug harms in people with
M/SUDs.
To design an app that is responsive to people with M/SUDs, the project
will need to overcome significant user-centric and technological design
challenges. To this end, the study will conduct several online surveys
to collect feedback from multi-stakeholders (e.g., healthcare providers,
medical researchers, app developers, end-users). These online surveys
will focus on three stages of the app design process: prototyping,
evaluation, and pilot study. This iterative process is scheduled to
evolve over a 12-month period (July 2016 – July 2017).
-
The Prototyping stage (July – September 2016) will develop the
TraitMap prototype based on aggregated findings from an online survey.
The survey will be compiled by Principle Investigator (Munro) using
Typeform.com and promoted via online support communities. Aggregated
multi-stakeholder feedback will inform the initial TraitMap prototype
which will be built using open-source development platform ResearchKit
(www.researchkit.org). Studies by
world-leading medical research units at Stanford, Johns Hopkins, and
Oxford show ResearchKit counters many of the methodological
limitations and data loss to follow-up that characterize Internet
trials (Jardine et al., 2015). Specifically built to support medical
research, ResearchKit enables highly automated data collection by
utilizing key features such as: (a) streamlined informed consent
processes, (b) interactive prompts that ‘nudge’ users to participate,
and (c) mobile sensing technologies such as GPS, touch screen, camera,
and accelerometer that collect infinitely richer data sets (Jardine et
al., 2015). This project will utilize these features of ResearchKit to
gather continuous data sets. First, self-report surveys will apply
subjective methods to map changes in personality traits, cravings,
drug use, and risk behaviours. Second, mobile sensing technologies
will apply objective methods to map changes in passively-collected
data. Computer algorithms will identify pattern irregularities in
location (from GPS) and movement (from accelerometer) that may trigger
craving and relapse events. Third, active tasks will be tailored to
disrupt and change users’ habitual behaviours under semi-controlled
conditions while phone sensors collect continuous data. Active tasks
are subjective and objective measures in that they require user input
which is passively collected by phone sensors. Three active task
reminders will be sent to participants daily, requesting they exercise
three personality traits outside their comfort zone: (i) a
social-extraversion task will ask participants to connect with a
stranger during a 20-minute stroll/walk (movement data collected by
accelerometer); (ii) a stress-reduction task will instruct
participants to sit in a quiet place for 10 minutes and be mindful,
then look directly into the camera (heart rate data collected via
camera), and (iii) a neural-plasticity task will instruct participants
to compete a simple cognitive exercise on the phone screen (cognitive
data collected by touch screen). Each active task will be tailored to
influence the relevant trait changes necessary to support resilience
to health risk: (a) positive activity interventions (PAIs) to promote
PEM/E (Layous, Chancellor, Lyubomirsky, Wang, & Doraiswamy, 2011),
(b) mindfulness-based stress reduction (MBSR) to promote NEM/N (Baer,
2003), and (c) life skills training (LST) to promote CON (Botvin &
Griffin, 2014).
-
The Evaluation stage (October – December 2016) will host a focus
group comprising of University of Auckland students (n = 5) and mental
health providers (n = 5). The initial focus group will begin with a
broad overview of TraitMap, its purpose, and general functionality.
Text content from the survey and informed consent features will be
evaluated to confirm their cultural validity and universality.
Questions will evaluate the ease at which users will engage with a
coded query for drug use in the form of a weather question. For
example, when asked “what was the weather like yesterday?” users
will be prompted to reply “clear” for no drug use, “cloudy” for
marijuana, “rainy” for alcohol, “snowy” for crack/cocaine, and
“other” for other illicit drugs. This textual analysis will be
followed by open-ended questions related to the suitability of the
active tasks to reduce drug cravings and risk behaviours. Questions
will specifically address whether people with M/SUDs may have issues
with (a) continuous phone monitoring, (b) frequency of probes sent to
their phone, and (c) adherence to a daily regimen of active tasks.
Finally, stringent measures for patient confidentiality such as
encryption methods and de-identified collected data will be evaluated.
Such concerns are of paramount importance if any smartphone is lost,
stolen, or sold during the trial. One month after design iterations
are made, the focus group (n = 10) will be asked to download and
pre-test TraitMap for 2 weeks. On completion of the pre-test, feedback
will be recorded via an online survey which will inform developers of
any design flaws that need correction prior to launching the pilot
trial.
-
The Pilot stage (January – June 2017) will conduct a pilot randomised
controlled trial to assess the viability of TraitMap to reduce drug
craving and risk behaviours in people with M/SUDs. The pilot trial
will aim to recruit a broad range of people with M/SUDs from online
support groups and the App Store. Consumer popularity of Apple
products combined with the benefits of being listed in the App Store
is likely to assist the project in gaining direct access to a
diversity of user groups. A series of screening questions will
correlate eligibility criteria with patient responses and thus
automate screening and enrollment. Interested participants will be
asked to complete a screening survey and admitted to the study only if
they are >18 years of age, own an iPhone and are familiar
with its use, and actively seek treatment for M/SUDs. A confirmation
message will be sent to patients who do not qualify. Respondents that
do qualify will complete an informed consent process via a
comprehension quiz, generated PDF and email, and request for
electronic signature. Participants meeting the inclusion criteria will
be randomised to the TraitMap group or non-personalised control arm.
Follow-up surveys at 4-, 8- and 12-week intervals will aim to identify
key differences and similarities in drug craving and risk behaviours
between groups.