Incentive-Based Interventions for Smoking Cessation: Findings from the SMILE Trial

Background: Smoking cessation services are not widely available in many low-resource settings. In this study, we tested several incentive-based interventions in a scalable workplace setting in a middle-income country.

Design/Methods: We undertook a 9-arm cluster randomized controlled trial of factories located in the metropolitan area of Bangkok, Thailand. We randomly assigned 101 worksites from 84 companies (containing 4,190 eligible smokers) to treatment arms that varied on two cross-randomized factors: presence of a deposit contract and type of bonus (none, individual, or team). The deposit contract required a minimum deposit of about $3 in a tamper-proof personal box, with voluntary contributions thereafter. A participant forfeited all deposits if he or she failed to abstain from smoking. The individual bonus involved a payment for abstaining of either $20 or $40, randomly varied by worksite. The team bonus involved a $40 payment if the person and a randomly assigned teammate each abstained. An additional arm included a deposit contract plus a teammate, but without any further incentives. All arms received smoking cessation counseling. All incentives depended on smoking status at 3 months, with biochemically verified abstinence collected at 3, 6, and 12 months.

Results: Intervention take-up was high across all arms, notably >50% in deposit arms. Controlling for socio-demographic, smoking, and trial-related factors, the deposits reduced take-up by 12 percentage points (p < 0.01), whereas the individual bonus of $20 and $40 increased take-up by 10 percentage points and 9 percentage points, respectively (p < 0.05 for each). We also find evidence of substantial heterogeneity in take-up by worksite. Biochemically verified 7-day smoking abstinence at 12 months was largest in the $40 individual bonus arm, a 43% (5 percentage point) increase relative to the control group (p < 0.01). Deposit contracts did not significantly increase abstinence. There were no strong interaction effects between deposits and the bonus, nor large team effects. Relapse was low following the end of the incentives.

Implications: This study advances the literature on the use of deposit contracts and social and monetary incentives to promote smoking cessation in low-resource settings. It provides some of the first rigorous evidence on the value of supplementing a basic deposit contract with individual and social incentives.