More evidence is needed to determine if programmes delivered over the Internet can help people to stop smoking. This review found several trials reporting success rates for stopping smoking after six months or more. In combined results from three trials which were at risk of bias, Internet programmes that were interactive and tailored to individual responses led to higher quit rates than usual care or written self help at six months or longer. Some interventions appeared more effective than others within this group, but with no obvious reason. Two trials where the Internet programme was not tailored did not improve smoking outcomes but direct comparisons between interactive/tailored and non-interactive, non-tailored programmes did not show a difference between the two.
The Internet may have an additional benefit when used alongside other interventions, such as nicotine replacement therapy or other pharmacotherapy. Innovative smoking cessation interventions delivered via the Internet may be more attractive to young people and to women who smoke, and less attractive to smokers reporting depression.
Results suggest that some Internet-based interventions can assist smoking cessation at six months or longer, particularly those which are interactive and tailored to individuals. However, the trials that compared Internet interventions with usual care or self help did not show consistent effects and were at risk of bias. Further research is needed despite 28 studies on the subject. Future studies should carefully consider optimising the interventions which promise most effect such as tailoring and interactivity.
The Internet is now an indispensable part of daily life for the majority of people in many parts of the world. It offers an additional means of effecting changes to behaviour such as smoking.
To determine the effectiveness of Internet-based interventions for smoking cessation.
We searched the Cochrane Tobacco Addiction Group Specialized Register. There were no restrictions placed on language of publication or publication date. The most recent search was conducted in April 2013.
We included randomized and quasi-randomized trials. Participants were people who smoked, with no exclusions based on age, gender, ethnicity, language or health status. Any type of Internet intervention was eligible. The comparison condition could be a no-intervention control, a different Internet intervention, or a non-Internet intervention.
Two authors independently assessed and extracted data. Methodological and study quality details were extracted using a standardized form. We extracted smoking cessation outcomes of six months follow-up or more, reporting short-term outcomes where longer-term outcomes were not available. We reported study effects as a risk ratio (RR) with a 95% confidence interval (CI). Clinical and statistical heterogeneity limited our ability to pool studies.
This updated review includes a total of 28 studies with over 45,000 participants. Some Internet programmes were intensive and included multiple outreach contacts with participants, whilst others relied on participants to initiate and maintain use.
Fifteen trials compared an Internet intervention to a non-Internet-based smoking cessation intervention or to a no-intervention control. Ten of these recruited adults, one recruited young adult university students and two recruited adolescents. Seven of the trials in adults had follow-up at six months or longer and compared an Internet intervention to usual care or printed self help. In a post hoc subgroup analysis, pooled results from three trials that compared interactive and individually tailored interventions to usual care or written self help detected a statistically significant effect in favour of the intervention (RR 1.48, 95% CI 1.11 to 2.78). However all three trials were judged to be at high risk of bias in one domain and high statistical heterogeneity was detected (I² = 53%), with no obvious clinical explanation. Pooled results from two studies of an interactive, tailored intervention involving the Internet and automated phone contacts also detected a significant effect (RR 2.05, 95% CI 1.42 to 2.97, I² = 42%). Results from a sixth study comparing an interactive but non-tailored intervention to control did not detect a significant effect, nor did the seventh study, which compared a non-interactive, non-tailored intervention to control. Three trials comparing Internet interventions to face-to-face or phone counselling also did not detect evidence of an effect, nor did two trials evaluating Internet interventions as adjuncts to other behavioural interventions. A trial in college students increased point prevalence abstinence after 30 weeks but had no effect on sustained abstinence. Two small trials in adolescents did not detect an effect on cessation compared to control.
Fourteen trials, all in adult populations, compared different Internet sites or programmes. Pooled estimates from three trials that compared tailored and/or interactive Internet programmes with non-tailored, non-interactive Internet programmes did not detect evidence of an effect (RR 1.12, 95% CI 0.95 to 1.32, I² = 0%). One trial detected evidence of a benefit from a tailored email compared to a non-tailored one, whereas a second trial comparing tailored messages to a non-tailored message did not detect evidence of an effect. Trials failed to detect a benefit of including a mood management component (three trials), or an asynchronous bulletin board.