E-cigarette support for smoking cessation: Identifying the effectiveness of intervention components in an on-line randomized optimization experiment

Abstract

Aims, design and setting: The aim of this study was to determine which combination(s) of five e-cigarette-orientated intervention components, delivered on-line, affect smoking cessation. An on-line (UK) balanced five-factor (2 × 2 × 2 × 2 × 2 = 32 intervention combinations) randomized factorial design guided by the multi-phase optimization strategy (MOST) was used.

Participants: A total of 1214 eligible participants (61% female; 97% white) were recruited via social media.

Interventions: The five on-line intervention components designed to help smokers switch to exclusive e-cigarette use were: (1) tailored device selection advice; (2) tailored e-liquid nicotine strength advice; (3): tailored e-liquid flavour advice; (4) brief information on relative harms; and (5) text message (SMS) support.

Measurements: The primary outcome was 4-week self-reported complete abstinence at 12 weeks post-randomization. Primary analyses were intention-to-treat (loss to follow-up recorded as smoking). Logistic regressions modelled the three- and two-way interactions and main effects, explored in that order.

Findings: In the adjusted model the only significant interaction was a two-way interaction, advice on flavour combined with text message support, which increased the odds of abstinence (odds ratio = 1.55, 95% confidence interval = 1.13-2.14, P = 0.007, Bayes factor = 7.25). There were no main effects of the intervention components.

Conclusions: Text-message support with tailored advice on flavour is a promising intervention combination for smokers using an e-cigarette in a quit attempt.

Keywords: Digital interventions; e-cigarettes; multi-phase optimization strategy (MOST); nicotine; smoking cessation; smoking reduction; tailored advice; tobacco; vaping.

FIGURE 1

Consolidated Standards of Reporting Trials (CONSORT) diagram for study participation. Note: n =1214 participants were randomised to one of the 32 computer‐generated block permutations, deriving from the 5 intervention components: 1‐tailored e‐liquid flavour advice, 2‐tailored device selection advice, 3‐tailored eliquid nicotine strength advice, 4‐text message support and 5‐brief information on relative harms, each intervention component ON or OFF (‘ON’ denotes intervention received). Thus, total number of participants in all intervention components combined exceed the overall sample size as participants could be randomised to more than one intervention component at a time (see supplementary file Table S1 for n per permutation). *Participants who failed to complete the 12‐week follow‐up survey were asked ‘Have you smoked at all in the last 4 weeks’ via email and text, n =107 responses were received.

FIGURE 2

Significant interactions for the primary outcome 4‐week point‐prevalence abstinence at 12 weeks post‐randomization; graph represents the flavour × text messages interaction.