Skip to content

Cessation Methods: The Evidence

How the evidence compares across the main ways people quit. This page is the evidence layer; the practical comparison lives on the main site methods section.

Evidence strength by method

Method Best available evidence Direction of findings
Varenicline Cochrane partial-agonists review (R-002); EAGLES (R-015) Most effective single medication in network analyses
NRT (patch + fast-acting) Cochrane NRT review (R-001) ~50–70% increase in quit rates vs placebo
Cytisine West 2011 (R-016); Walker 2014 (R-017) Superior to placebo; comparable to NRT at lower cost
Bupropion Cochrane antidepressant review (R-003) Roughly doubles quit rates vs placebo
Combination NRT Cochrane overview (R-008) Patch + fast-acting form more effective than one form
Counselling / behavioural support USPSTF 2021; WHO 2024 Effective alone; more effective combined with medication
Quitlines WHO 2024; US PHS 2008 Effective population-level support
Digital interventions WHO 2024 Recommended as complement; evidence base growing
Peer support Community-level; mixed formal evidence Real-world valued; formal trial evidence limited
Cold turkey (unaided) Observational data Most common in the community (see community statistics), lower per-attempt success than aided quitting in trial literature
Gradual reduction Cochrane R-004 Abrupt and gradual produce similar quit rates when supported; evidence on reduction alone is weaker

Three reading rules

  1. "Effective" is statistical, not personal. A method that helps more people on average may not be the right one for a given individual.
  2. Combination beats monotherapy. Medication + behavioural support is the single most consistent finding across guidelines.
  3. Community popularity ≠ evidence strength. Cold turkey dominates community reports but is not the most effective aided approach in trials. Both facts can be true — see Community Data.