Medication Samples and Smoking Cessation Among Adults: A Randomized Clinical Trial

Smoking cessation interventions could ultimately offer greater impact to the extent that they are brief, concrete, and face valid (to the individuals presenting and receiving the intervention), which would then make these interventions scalable across a broad spectrum of adults who smoke (AWS). Medication sampling is one potential strategy to meet that need.

To determine outcomes of varenicline sampling in a fully powered randomized clinical trial (RCT), conducted from February 2021 to April 2025.

This decentralized RCT included non-treatment seeking AWS with varying levels of motivation to quit who were recruited throughout South Carolina.

Participants were randomized to receive a 4-week sample of varenicline, nicotine replacement therapy (NRT; active control), or quitline referral (inactive control) in a 2:1:1 ratio. Accompanying 4-week medication supply, intervention messaging in both sampling groups emphasized naturalistic use, ie, a participant-driven experience with self-determined uptake, use, and goals for use.

The primary outcome was self-reported 7-day point prevalence abstinence (PPA) at 6-month follow-up. Secondary outcomes included carbon monoxide (CO)-verified abstinence, floating abstinence (any 7-day period of not smoking throughout follow-up), quit attempts, and smoking reduction.

The study sample included 651 AWS (mean [SD] age, 52 [11] years; 431 (66%) female), with 161 randomized to the no-sampling control, 172 randomized to NRT, and 318 randomized to varenicline. Compared with the no-sampling control, AWS receiving varenicline samples had higher rates of self-reported PPA at month 6 (16 of 161 [10%] vs 53 of 318 [17%]; P = .048), floating abstinence throughout follow-up (33 [20%] vs 108 [34%]; P = .003); and greater incidence of 50% or greater reduction in cigarettes per day (CPD) by 6 months (31 [19%] vs 106 [33%]; P = .002); however, rates of CO-verified abstinence at 6 months were not significantly different. Varenicline sampling was superior to NRT sampling for 6-month self-reported abstinence (53 [17%] vs 14 of 172 [8%]; P = .01); floating abstinence through 6 months (108 [34%] vs 43 [25%]; P = .04); and incidence of 50% or greater reduction in CPD at week 8 (111 [35%] vs 38 [22%]; P = .005) but not at 6 months.

In this RCT of 651 AWS, varenicline sampling was efficacious, with potentially superior outcomes compared with NRT. Results provide additional evidence in support of medication sampling as a pragmatic option to engage AWS in cessation, worth further evaluation within an applied settings.

ClinicalTrials.gov Identifier: NCT04525755.