Among 1,503 adolescents randomized, the 7-month follow-up rate was 70.8% (n = 1,064). Data on cannabis use was missing for 48 participants, who provided data only on 7-month nicotine vaping status. Thus, the adolescent analytic sample comprised n = 1,016 participants with follow-up data on both e-cigarette and cannabis use. There was no differential attrition by treatment assignment (p = 0.20), with 66.0% (501 of 759) of Intervention participants retained at 7 months versus 69.2% (515 of 744) of Control. Likewise, there was no differential attrition by baseline cannabis use (p = 0.74), with 68.4% (258 of 377) of Exclusive E-cigarette Users retained at 7 months versus 67.3% (758 of 1126) of Dual Users. At baseline, 74.6% (95% CI = 71.8, 77.3) of adolescents reported past 30-day cannabis use, which decreased to 50.1% (47.0, 53.2) at 7 months, a 24.5% point change (95% CI = 20.8, 28.0; McNemar’s test p < 0.001).
Among 2,588 YAs randomized, the 7-month follow-up rate was 76.0% (n = 1,967). Data on cannabis use was missing for 138 participants, who provided data only on 7-month nicotine vaping status. Thus, the YA analytic sample comprised n = 1,829 participants with follow-up data on both e-cigarette and cannabis use. There was no differential attrition by treatment assignment (p = 0.14), with 69.3% (904 of 1304) of Intervention participants retained at 7 months versus 72.0% (925 of 1284) of Control. Likewise, there was no differential attrition by baseline cannabis use (p = 0.86), with 70.9% (747 of 1053) of Exclusive E-cigarette Users retained at 7 months versus 70.5% (1,082 of 1534) of Dual Users. At baseline, 59.2% (95% CI = 56.9, 61.4) of YAs reported past 30-day cannabis use, which decreased to 55.0% (95% CI = 52.7, 57.3) at 7 months, a 4.2% point change (95% CI = 1.9, 6.4; McNemar’s test p < 0.001).
What were the overall patterns of abstinence from e-cigarettes and cannabis at 7-months?
As shown in Table 1, 31.7% (95% CI = 28.8, 34.6) of adolescents were Dual Abstinent, 18.2% (95% CI = 15.9, 20.7) were Exclusive E-cigarette Users, 15.1% (95% CI = 12.9, 17.4) were Exclusive Cannabis Users, and 35.0% (95% CI = 32.1, 38.1) were Dual Users.
As shown in Table 2, 15.6% (95% CI = 13.9, 17.3) of YAs were Dual Abstinent, 29.4% (95% CI = 27.3, 31.6) were Exclusive E-cigarette Users, 12.8% (95% CI = 11.3, 14.5) were Exclusive Cannabis Users, and 42.2% (95% CI = 39.9, 44.5) were Dual Users.
Was there a treatment effect in promoting dual abstinence at follow-up?
Yes. As shown in Table 1, among adolescents, the rate of Dual Abstinence was 13.5% points higher (95% CI = 7.8, 19.1; p < 0.0001) among those randomized to Intervention (38.5%; 95% CI = 34.4, 42.9) vs. Control (25.0%; 95% CI = 21.5, 29.0). As shown in Table 2, among YAs, the rate of Dual Abstinence was 4.6% points higher (95% CI = 1.3, 7.9; p = 0.007) among those randomized to Intervention (17.9%; 95% CI = 15.5, 20.6) vs. Control (13.3%; 95% CI = 11.2, 15.7).
Did treatment effects in promoting dual abstinence vary by baseline product use?
No. In the adolescent sample, the treatment advantage of Intervention over Control was comparable for Exclusive E-cigarette Users (12.4 points; 95% CI = 0.6, 23.8) and Dual Users (13.9 points; 95% CI = 7.4, 20.3), interaction p = 0.82 (Table 1). Among Exclusive E-cigarette Users, 44.0% of adolescents randomized to Intervention were Dual Abstinent (95% CI = 35.1, 53.1) compared to 31.6% of Control (95% CI = 23.8, 40.2). Among Dual Users, 36.7% of Intervention participants were Dual Abstinent (95% CI = 31.8, 41.8) compared to 22.8% of Control (95% CI = 18.7, 27.3).
Likewise, in the YA sample, the treatment advantage of Intervention over Control was comparable for Exclusive E-cigarette Users (7.4 points; 95% CI = 1.1, 13.7; p = 0.02) and Dual Users (3.7 points; 95% CI = 0.0, 7.1, p = 0.03), interaction p = 0.28 (Table 2). Among Exclusive E-cigarette Users, 29.7% of YAs randomized to Intervention were Dual Abstinent (95% CI = 25.0, 34.8) compared to 22.3% of Control (95% CI = 18.3, 26.8). Among Dual Users, 10.3% of Intervention participants were Dual Abstinent (95% CI = 7.9, 13.2) compared to 6.6% of Control (95% CI = 4.6, 9.0).
Was there an interaction effect between vaping status at 7 months and baseline tobacco product use on cannabis use outcomes?
Among adolescents, the difference in cannabis use at follow-up between continuing vapers and vaping abstainers was significantly weaker among baseline Exclusive E-cigarette Users than among baseline Dual Users (interaction p < 0.001). As shown in Supplemental Table 1, among 258 adolescent baseline Exclusive E-cigarette Users, cannabis use at 7 months was reported by 31.1% (95% CI = 23.4, 39.6) of those who were still nicotine vaping versus 21.1% (95% CI = 14.8, 29.2) of those who were vaping abstinent, a 10% point difference (95% CI = −0.8, 20.3). Among 758 baseline Dual Users, cannabis use at 7 months was reported by 77.3% (95% CI = 72.9, 81.3) of those who were still nicotine vaping versus 36.1% (95% CI = 31.1, 41.3) of those who were vaping abstinent, a 41.3% point difference (95% CI = 34.5, 47.4). In total, 97 out of 258 baseline Exclusive E-cigarette Users were dual abstinent (37.6%) compared to 225 out of 758 baseline Dual Users (29.7%), a significant difference at p = 0.019.
Among YAs, the difference in cannabis use at follow-up between continuing vapers and vaping abstainers was comparable (interaction p = 0.81) for baseline Exclusive E-cigarette Users and baseline Dual Users. As shown in Supplemental Table 2, among 747 YA baseline Exclusive E-cigarette Users, cannabis use at 7 months was reported by 27.2% (95% CI = 23.4, 31.2) of continuing nicotine vapers versus 16.8% (95% CI = 12.2, 22.3) of vaping abstainers, a 10.4% point difference (95% CI = 3.9, 16.2, p < 0.001). Among 1,082 baseline Dual Users, cannabis use at 7 months was reported by 79.5% (95% CI = 76.5, 82.2) of continuing nicotine vapers versus 68.1% (95% CI = 62.3, 73.4) of vaping abstainers, an 11.4% point difference (95% CI = 5.5, 17.6). In total, 193 out of 747 baseline Exclusive E-cigarette Users were dual abstinent (25.8%) compared to 92 out of 1082 baseline Dual Users (8.5%), a significant difference at p < 0.001.