Smartphone-based Financial Incentives

Overview

Cigarette smoking during pregnancy increases risk for catastrophic pregnancy complications, growth retardation, other adverse fetal and infant health problems, and later-in-life chronic conditions among exposed offspring. The most effective intervention for reducing smoking during pregnancy is financial incentives whereby participants earn incentives (e.g., gift cards, cash) contingent on objective evidence of smoking abstinence. However, financial incentives-based interventions are typically delivered in relatively intense protocols requiring frequent clinic visits, which limits the geographical range over which services can be delivered and potentially denies treatment to those residing in remote or otherwise difficult to reach settings. The present study will examine the feasibility, efficacy, and cost-effectiveness of a smartphone-based financial incentives intervention whereby smoking monitoring and delivery of incentives are completed remotely using a mobile app (to be designed by DynamiCare Health, Inc.). Eligible participants who complete the informed consent process will be randomized to one of two conditions: an incentives condition wherein women will receive financial incentives contingent on the remote submission of breath and saliva specimens indicating abstinence from recent smoking (described below), or a best practices control condition in which women will receive usual care for smoking cessation that is provided at their obstetrical clinics, as well as three brief educational sessions and referral to the Vermont (or other state) pregnancy-specific quit line by our research staff. For inclusion in the study, women must meet the following criteria: (a) \> 18 years of age, (b) report being smokers at the time they learned of the current pregnancy, (c) report smoking in the 7 days prior to completing their phone eligibility screening, (c) \< 25 weeks pregnant, (d) speak English, (e) own a smartphone (Android or iOS; 81.8% of pregnant women in wave 1 \[2013-2014\] of the Population Assessment of Tobacco and Health \[PATH\] reported owning a smartphone). Exclusion criteria include: (a) current or prior mental or medical condition that may interfere with study participation (assessed via self-report during phone eligibility screening), (b) smoke marijuana more than once each week and not willing to quit (marijuana smoking can inflate breath CO), (c) exposed to unavoidable occupational sources of CO (e.g., car mechanic), and (d) self-report currently being maintained on opioid maintenance therapy (e.g., methadone, buprenorphine). Participants randomized to the incentives condition will select a quit date (either the first or second Monday following their enrollment), and will submit videos of themselves blowing into a breath CO monitor twice daily during week 1. They will receive incentives for every sample where expired breath CO is \< 6 ppm. Beginning in week 2 and extending through week 6, participants will submit videos twice per week (Monday/Thursday) for which they will receive incentives for providing videos of themselves completing saliva cotinine tests indicating smoking abstinence. From week 7 until delivery, participants will submit videos once per week and will continue to receive incentives for saliva cotinine tests indicating no smoking. During the postpartum period, women will submit videos twice weekly for the first 4 weeks and once weekly from weeks 5-12. Women will receiving incentives for negative breath and saliva samples, and the value of incentives will increase with each consecutive sample indicating smoking abstinence. Participants will not receive incentives for missed samples or samples that indicate smoking, and the incentive schedule will be reset at its starting value. However two consecutive negative samples following a missed or positive sample will restore the incentive to its prior value. Women in both conditions will complete seven formal assessments of their smoking status during their participation along with a treatment acceptability questionnaire and semi-structured interview on barriers and facilitators of treatment engagement. We conducted a power analysis to estimate the number of participants required to detect treatment effects assuming late-pregnancy abstinence rates of approximately 40% vs. \< 10% (incentives vs. best practices, respectively), and 24-week postpartum abstinence rates of approximately 20% vs. \< 5%. The proposed sample size of 76 per treatment condition will result in at least 80% power to detect a difference between the two treatment conditions in abstinence rates of 40% vs. 10 % at late-pregnancy or 20% vs. 5% at 24-weeks postpartum assessments using a chi-square test and significance level of 0.05.