Pilot Trial of Using SmokeBeat to Monitor and Increase Smoking Abstinence in Pregnant Women

Somatix, LGH, and the Center for Health Incentives and Behavioral Economics of the University
of Pennsylvania (CHIBE) will collaborate in using the unique smoking recognition technology
of SmokeBeat app to track and analyze the smoking habits of LGH target smoking population.
The target population for the SmokeBeat pilot is low-income pregnant women who smoke. This
will enable us to gain insights on this target population and reach higher effectiveness of
smoking cessation. Each participant will be enrolled in the trial for 12 weeks. Women
randomized to the intervention arm will be given an Asus Zenwatch 2 smartwatch and a
smartphone data plan for the duration of the study. Women assigned to the control group will
be given the same smartwatch to keep following their participation in the trial. The
tentative start date is June 2017.

After completing the consent participants will be randomized to one of two study arms. The
control arm will consist of 25 pregnant women. Each will be provided with usual care through
Healthy Beginnings or NFP but will not receive a smartwatch or the active intervention.
Smoking behavior will be measured by self-report and sample testing for the duration of the
trial. Self-reported smoking status will be collected weekly. Sample testing will occur
weekly and will be identical to the intervention group. Following their completion of this
trial they will receive a smartwatch as well. The intervention arm will consist of 25
pregnant women. Each will be enrolled and provided with a smartwatch. The SmokeBeat app will
be active. Smoking behavior will be measured by the smartwatch for the duration of the trial.
Self-reported smoking status will be collected weekly. Sample testing will occur weekly. All
participants assigned to the intervention arm will be allowed to keep the smartwatch for
their participation and be provided with a smartphone data plan for the duration of the
study.

Two types of tests will be conducted with each participant, regardless of self-reported quit
status. NicAlert strip tests (semi quantitative test) will be used to test saliva cotinine
during weeks 2, 3, 5, 6, 8, 9, 11, and 12 of the participant's participation. This test will
provide us with an accurate indicator of whether or not women have quit smoking while
reducing the cost of the pilot study. A quantitative liquid chromatography cotinine test
conducted by ARUP Labs will be used to test urine during weeks 1, 4, 7, and 10 for all women
regardless of self-reported and smartwatch smoking status. This test, while more expensive,
will provide us with the concentration of cotinine for each participant. This number can be
directly compared to the self-reported and smartwatch reported number of cigarettes smoked as
a way to validate self-report and smartwatch accuracy even when a participant has not quit
smoking completely.

The main research instruments will consist of questionnaires on a subjects smoking habits and
psychological tests that will help evaluate which types of patients are most able to quit
smoking. All participants will be asked to complete an intake questionnaire at baseline
(estimated length 15 minutes) and an exit questionnaire at the end of study participation 3
months after they start in the program (estimated length 15 minutes). Investigators will
examine whether baseline values and/or changes during 3 months in these variables confound,
causally mediate, or modify the effects of the intervention on quit rates. (a) Standard
questionnaires will be administered to collect the following information: demographics (age,
gender, ethnicity, marital status, and education), number of years of smoking, number of
prior quit attempts, duration of previous abstinence periods, and current smoking rate.
Nicotine dependence will be measured by the modified Fagerström Test for Nicotine Dependence
(FTND), a 6-item measure of nicotine dependence. The FTND has satisfactory internal
consistency (Cronbachs alpha=0.64) and high test-retest reliability (r=0.88).3 Heavy smokers
are at highest risk for smoking related illnesses. Some but not all studies suggest that
heavier smokers have lower quit rates. Investigators will evaluate whether FTND score either
confounds or modifies the effects of the intervention. (b) Investigators will measure
smoker's baseline readiness to quit using the Prochaska Stage of Change. (c) Investigators
will assess participant's types and locations of jobs and whether they are exposed to smoking
at home or in the workplace. (d) Participants use, durations, and doses of nicotine
replacement therapy, varenicline, and buproprion will be assessed throughout the study.
Participants will be asked to not use Nicotine patches or gum during the course of the study
so that it will not interfere with weekly testing for smoking abstinence. (e) Time (or delay)
discounting. Time discounting (known interchangeably as present-biased preferences or delay
discounting) is a measure of impulsivity that reflects people's tendencies to discount the
value of a reward as a function of how far in the future it would be received. People who
smoke and use other drugs have steeper delayed discounting functions, meaning they have
stronger present biased preferences. This is expected because smoking, like other unhealthy
behaviors, entails immediate gratifications in exchange for future health problems. This also
suggests that provision of early rewards, such as payments, might more effectively promote
smoking cessation than interventions focusing on future rewards such as long-term health
benefits, and that this effect might be strongest among those with the steepest discounting
functions. Consistent with prior work, investigators will use a 7-item monetary choice
questionnaire that defines ranges of delayed discounting by fitting a hyperbolic function to
participants expressed preferences for smaller immediate versus larger delayed rewards.
Information will also be collected weekly from participants on withdrawal and craving for
nicotine as well as self-reported number of cigarettes per week. Investigators will also
administer a 5 question weekly survey that addresses weekly levels of cravings, stress,
social support, and asks about current cigarette use. All questions will be rated on a Likert
scale, except for the number of cigarettes smoked, that will be an estimate of use.

The SmokeBeat app is a digital health app that allows for the user to monitor their smoking
behavior in real time, set smoking cessation goals, and allows investigators, doctors and
others using the app with participants to send messages directly related to smoking behavior.
The SmokeBeat app, developed by Somatix Inc., will collects time-stamped on hand movements
associated with smoking behavior. Participants will receive notifications on their phone and
smartwatch when a smoking bout has been detected. In addition this app will deliver daily
text messages aimed at improving quit rates among participants. Investigators will be able to
create the types of messaging they want to send and to modify when these messages will be
sent to all participants. http://somatixinc.com/smokebeat/.

A second pilot were begun in the summer of 2018. Intervention participants were offered a
monetary reward for each day they went without smoking. Control participants were offered a
financial incentive to wear the watch. In the fall of 2018 a third will start with an
increase in the financial incentive, and a new smart band model.

A third pilot was started in the winter of 2018/2019. New participants were recruited, who
were then randomized to one of two arms. Both arms were given smartbands that tracked their
smoking use. Participants in the control arm were offered financial incentives for every day
that they consistently wore the smartband. Participants in the experimental arm were offered
financial incentives for consistently wearing the smartband and abstaining from smoking. All
participants completed remote nicotine testing and survey questionnaires with study personnel
via video chat. Five weekly video chats are scheduled with participants.

Inclusion Criteria:

- Pregnant or within 3 weeks postpartum

- Self-reported smoking

- Part of the Nurse Family Partnership or Healthy Beginnings

- Uses an Android smart phone

Exclusion Criteria:

- Does not speak English

- Not interested in wearing a smart watch for 12 weeks