Financial Incentives for Homeless Smokers: A Community-based RCT

People experiencing homelessness have a 3.5-fold higher prevalence of cigarette smoking in comparison to non-homeless people, contributing to 2-fold higher rates of lung cancer and 3- to 5-fold higher rates of tobacco-attributable death. Homeless smokers want to quit, but studies have not yet uncovered the optimal approach to help them do so. In an 8-week pilot randomized controlled trial (RCT) at Boston Health Care for the Homeless Program (BHCHP), we found that financial incentives for smoking abstinence were associated with 7-fold higher odds of brief smoking abstinence in comparison to a non-incentivized control condition. These results suggest that financial incentives are a promising approach for reducing smoking in this vulnerable population, but further investigation in a larger sample is needed to improve the duration of on-treatment abstinence, assess post-treatment effects, and better understand mechanisms of action and contextual factors that may influence treatment response. To address the above gaps in evidence, we will conduct a community-based RCT of financial incentives for smoking abstinence among adult smokers at BHCHP. Recognized as a leader in homeless health care, BHCHP serves 12,000 currently and formerly homeless patients annually throughout greater Boston. We will randomize 180 participants recruited from 3 BHCHP sites: a shelter clinic, a day center clinic, and a medical center clinic. All participants will be offered 12 weeks of varenicline, 5 sessions of tobacco coaching, and 10 cotinine monitoring visits over a 12-week period. Participants randomized to the financial incentives arm (n=90) will receive escalating debit card payments (range $25-$70) at each monitoring visit for saliva cotinine levels \<30 ng/ml. Control arm participants (n=90) will receive a fixed payment ($10) at each monitoring visit regardless of their saliva cotinine level. We will use an embedded-experiment mixed methods design, where qualitative ('qual') data collection is embedded within a larger quantitative ('QUAN') RCT with the following specific aims: Aim 1. (QUAN) To determine the effect of the financial incentives intervention on cotinine-verified 7-day smoking abstinence at A) the end of treatment (12 weeks) and B) 12 weeks after treatment (24 weeks). Hypotheses: Incentive arm participants will have significantly greater cotinine-verified 7-day smoking abstinence than control arm participants at A) 12 weeks and B) 24 weeks. Aim 2. (qual) To assess why, how, and under what circumstances homeless smokers A) achieve abstinence in response to financial incentives and B) maintain abstinence after incentives are stopped. Interviews with participants at A) 12 weeks (N=30) and B) 24 weeks (N=20) will examine cognitive ('why?'), procedural ('how?'), and contextual ('under what circumstances?') dimensions of their response to financial incentives to generate hypotheses about potential mechanisms for on-treatment and post-treatment effects and to inform modifications of the intervention for future use.