Effect of Graphic Warning Labels on Cigarette Pack–Hiding Behavior Among Smokers: The CASA Randomized Clinical Trial

IMPORTANCE: The inclusion of graphic warning labels (GWLs) on cigarette packs is recommended for tobacco control but has not yet been implemented in the US. It is unknown whether and to what extent the inclusion of GWLs on cigarette packs affects smokers’ willingness to display the packs in public. OBJECTIVE: To determine whether the inclusion of GWLs on cigarette packs affects pack-hiding behavior among smokers in social settings. DESIGN, SETTING, AND PARTICIPANTS: This community-based randomized clinical trial assessed smokers’ real-world experience of using cigarettes repackaged to include GWLs (GWL packs) compared with standard US packs and blank packs over a 3-month intervention period with 12 months of follow-up between September 6, 2016, and December 3, 2019. The study included 357 participants aged 21 to 65 years from San Diego County, California, who smoked 5 or more cigarettes per day, were not actively planning to quit smoking, were not pregnant, and had no unstable medical conditions. Participants purchased and received cigarette packs through the study website. INTERVENTIONS: During the 1-month run-in period, participants received their usual US cigarette packs. During the 3-month intervention period, participants were randomized to receive GWL packs (study-manufactured packs with 3 rotating images under license from the Commonwealth of Australia; GWL pack group), blank packs (study-manufactured packs devoid of industry marketing imagery; blank pack group), or standard US packs (US pack group). MAIN OUTCOMES AND MEASURES: Pack hiding was queried daily (with participants reporting behavior within the last 4 hours) and weekly via interactive text messages during the 1-month run-in and intervention periods. Self-reported smoking behavior was biochemically validated. RESULTS: Among 357 enrolled participants, the mean (SD) age was 39.3 (11.8) years; 195 participants (54.6%) were female, 40 (11.2%) were Hispanic, 243 (68.1%) were non-Hispanic White, and 74 (20.7%) were of other non-Hispanic races (including American Indian or Alaska Native, Asian or Pacific Islander, Black or African American, or multiracial). A total of 18 987 cigarette packs were purchased and delivered during the run-in and intervention periods. Daily querying showed that the inclusion of GWLs on cigarette packs increased the percentage of smokers who hid their packs at least some of the time from 41.3% (95% CI, 39.6%-43.0%) during the run-in period to 57.1% (95% CI, 55.9%-58.1%) by the end of the intervention period. In the postintervention period, returning to standard US packs reduced pack-hiding behavior to the levels observed during the run-in period. Pack hiding remained at run-in levels for both the blank pack group (35.2%; 95% CI, 33.6%-36.8%) and the US pack group (41.4%; 95% CI, 39.7%-43.1%]) throughout the study. Although even participants in the GWL group with the lowest prestudy tendency to conceal score (ie, 1) had a mean (SE) probability of pack hiding during the intervention of 0.84 (0.02), this group’s probability of pack hiding decreased to a mean (SE) of 0.43 (0.03) after intervention. When social reactions to packs were queried at the end of the study, the modal response from participants in the GWL pack group was observers’ aversive reactions to the packs, whereas the modal response from participants in the blank pack group was observers’ positive interest in the study. Neither smoking prevalence nor consumption differed by group at any point in the study. CONCLUSIONS AND RELEVANCE: In this randomized clinical trial, receiving cigarettes in GWL packs vs blank packs increased pack-hiding behavior in social settings, which may be associated with aversive reactions from observers. However, 12-month smoking behavior did not change. TRIAL REGISTRATION: ClinicalTrials.gov Identifier: NCT02676193