Interventions for smokeless tobacco use cessation

Abstract

Rationale: While combustible tobacco has been the subject of a very large amount of research, smokeless tobacco products receive less attention. Most smokeless tobacco products are very harmful and cause global health inequality. It is therefore important to identify evidence-based cessation aids.

Objectives: To assess the effects of behavioural and pharmacological interventions for smokeless tobacco use cessation.

Search methods: We searched the following databases from inception to 16 February 2024: Cochrane Central Register of Controlled Trials (CENTRAL); MEDLINE; Embase; PsycINFO; ClinicalTrials.gov (through CENTRAL); World Health Organisation International Clinical Trials Registry Platform (through CENTRAL). We also searched references of eligible studies.

Eligibility criteria: We included randomised controlled trials (RCTs) recruiting people of any age using smokeless tobacco, regardless of tobacco smoking status. Eligible studies could test any intervention designed to support people to quit smokeless tobacco use, and had to measure abstinence from either all tobacco use or smokeless tobacco use at six months or longer.

Outcomes: The outcome of interest was abstinence from all tobacco use or from smokeless tobacco use at six months or longer.

Risk of bias: We used the Cochrane RoB 1 tool to assess bias in included studies.

Synthesis methods: We followed standard Cochrane methods for screening and data extraction. We grouped studies by comparisons of eligible interventions and comparators, reporting individual study and pooled effects as appropriate. We used a random-effects Mantel-Haenszel model for analyses of behavioural interventions and a fixed effect Mantel-Haenszel model for analyses of pharmacotherapies to calculate risk ratios (RR) with 95% confidence intervals (CI). We assessed the certainty of evidence using GRADE.

Included studies: We included 43 trials of 20,346 people. Thirty-three trials were conducted in North America, five in India, two in Scandinavia, one in Pakistan and one in Turkey. One study was conducted across multiple sites in Bangladesh, India and Pakistan. Studies tested behavioural interventions (e.g. cessation counselling and brief advice) and pharmacotherapies (e.g. nicotine replacement therapy (NRT), varenicline, and bupropion). We judged five studies to be at low risk of bias overall, 22 at high risk of bias, and the remaining 16 at unclear risk of bias.

Synthesis of results: We found moderate-certainty evidence of increased quit rates from counselling compared with minimal support (RR 1.76, 95% CI 1.44 to 2.16; I² = 69%; 21 studies, n = 7417; downgraded because of heterogeneity), brief advice compared with no support (RR 1.24, 95% CI 1.03 to 1.48; I² = 49%; 7 studies, n = 6271; downgraded because of imprecision), and varenicline compared with placebo (RR 1.35, 95% CI 1.08 to 1.68; I² = 0%; 2 studies, n = 508; downgraded because of imprecision). We found low-certainty evidence (downgraded because of imprecision and risk of bias) of increased quit rates from NRT compared with placebo or no medication (RR 1.18, 95% CI 1.05 to 1.33; I² = 39%; 11 studies, n = 2826). Low-certainty evidence (downgraded because of imprecision) did not show benefit from bupropion compared with placebo (RR 0.89, 95% CI 0.54 to 1.44; I² = 0%; 2 studies, n = 293). We planned subgroup analyses to explore whether smokeless tobacco type affects intervention efficacy, but found insufficient data.

Authors' conclusions: Cessation counselling, brief advice, and varenicline each probably help more people to quit smokeless tobacco use than minimal or no support, or placebo. NRT may help more people to quit smokeless tobacco use than placebo or no medication. Low-certainty evidence does not currently support bupropion as a smokeless tobacco cessation intervention. Despite the majority of smokeless tobacco users living in South and Southeast Asia, only a minority of trials are conducted in those regions. Future trials should address this imbalance.

Funding: None REGISTRATION: Protocol available via DOI: 10.1002/14651858.CD015314.