Hypnotherapy for smoking cessation: the evidence

Hypnotherapy for smoking cessation is Grade B: moderate evidence, not the Grade A that clinic marketing often implies. Randomised trials exist. They do not add up to a clear long-term advantage over other behavioural support. The honest benchmark is the 2019 Cochrane review of 14 trials (1926 people). Grades are explained on how we evaluate evidence.

Key evidence

Barnes, McRobbie, Dong, Walker, and Hartmann-Boyce searched to 18 July 2018. Their authors’ conclusions are the sentence this wiki uses: there is insufficient evidence to say whether hypnotherapy is more effective than other behavioural support or unassisted quitting. If a benefit is present, current evidence suggests it is small at most. The plain-language summary says the same.

The 14 trials were too different for one meta-analysis. Six trials (957 people) that matched hypnotherapy against time-matched behavioural support did not show a difference. Two trials (269 people) against longer counselling also did not. One small trial (40 people) against no treatment favoured hypnotherapy but was methodologically weak. Five trials of hypnotherapy added to existing treatment found an effect, but four of those five were at high risk of bias and the results were highly inconsistent.

Carmody et al. (2008) is one of the larger included trials: 286 smokers, hypnosis plus nicotine patches versus counselling plus nicotine patches. Point estimates favoured hypnosis; the confidence intervals included no difference. It is not “hypnosis versus patches.” Both arms received patches.

The 1992 Viswesvaran and Schmidt meta-analysis is widely quoted as showing hypnosis among the strongest cessation methods. It pooled 633 studies of many methods. Later trial-quality filters, including Cochrane, do not treat it as the benchmark: it included uncontrolled and unverified studies.

The marketing line “quit in one session with 90% success” is not supported by these trials.

Suggestion and expectancy (the methods words behind a smoking script) are defined on suggestion, and response expectancy. They are not outcome grades.

Limitations

“Hypnotherapy for smoking” is not one protocol. Trials range from a single session to multi-session packages. Many rely on self-report. Only some confirm abstinence with carbon monoxide or cotinine. Trials against nothing look better than trials against intensive counselling. Adverse-event reporting is thin.

Clinic session counts and clinic success percentages are not trial endpoints. This page does not equate a research protocol with a booking offer.

What we don't know yet

A large, independently run, biochemically verified RCT of one named hypnotherapy protocol versus intensive counselling is still the trial Cochrane asked for. Who responds, whether remote delivery is equivalent, and cost-effectiveness are open. See what we don’t know.

This page is informational only. It is not personal medical advice. Quitting smoking is a medical priority; first-line medicines and behavioural support are not replaced by this wiki.