Jean-François E T T E R Institute of Social and Preventive Medicine, University of Geneva, Switzerland Contact: J.-F. Etter, IMSP,CMU,case postale, CH-1211 Genève 4, Suisse, Tél Is "stage of change" a valid concept ?

Background  The concept of "Stage of change" is widely used by researchers and clinicians.  Is it valid ?  Several "staging" questionnaires are available, do they produce similar results ?

Methods u Setting and participants: - Survey on Internet in 1025 ever smokers, - Retest after 8 days in 318 people (31%), - Follow-up after 32 daysin 451 people (44%). u Three different questionnaires assessing "stages of change" based on the same definition of stages: Velicer et al. An expert system intervention for smoking cessation. Addict Behav 1993;18: Questionnaire developed by ourselves.

Results u % of participants were classified in the same stage by all questionnaires. u This result calls for a more precise definition of stages and for the use of a unique standardized questionnaire. u % participants were NOT classified in the same stage at 8-day retest: - reflects true movements across the stages ? - or unreliability of the questionnaires ? u 4 - 9% Precontemplators and Contemplators had quit after 32 days: did they skipped intermediary stages?

Results u "Stage of change" includes 4 distinct variables: - intention (stages 1 - 3) - past behavior (stage 3) - current behavior(stages 1 - 5) - duration of current behavior(stages 4 - 5) u None of these 4 variables is measured comprehensively. u Two of these variables (intention and duration) are continuous variables categorized by arbitrary cutpoints.

Intention u Assessed only in current smokers, but intention may also be relevant in ex-smokers. u Cutpoints are arbitrary (1 and 6 months), and may not be the best possible. u Heterogenous categories: % Precontemplators had "absolutely no intention" of quitting smoking, % Preparators had made a "firm decision" to quit. u Intention better measured as a continuous variable ? In this study, a 0-10 score of intention to quit was linearly associated with cessation at follow-up.

Current behavior u Dichotomy: smoker (stages 1-3) or ex-smoker (stages 4+5). u "Stage" ignores occasional smoking. u Action stage included 5-7% of occasional smokers, this is contrary to theory, which states that to be classified in the Action stage "only total abstinence counts". u The definition of stages is not based on the WHO conventional definition of ever smokers (>100 cigarettes in a lifetime), which limits the comparability of studies.

Past quit attempts u Used only to identify "Preparators". u Quit attempts also relevant in Precont. and Contempl. u % downgraded to Contemplation because they had not made a quit attempt in the previous year. u The "Contemplation" stage is a heterogenous category: - intention 6 months, with or without quit attempts, and - intention 30 days, without quit attempt. u Smokers can never be "prepared" for their 1 st quit attempt. u Test-retest : self-report of 24-hour quit attempts in previous 12 months were unreliable in 12% participants.

Past quit attempts u Quit attempts in previous 7 days or 30 days were better predictors of smoking cessation than quit attempts in the past 12 months (memory bias ?) u Quit attempt of 24 hours: = relatively insignificant events, not well remembered, = short, unsuccessful quit attempts u Once a quit attempt is made, cessation is predicted by variables not in this model: - level of dependence, withdrawal symptoms, - number and duration of past quit attempt, - use of pharmacotherapy.

Time since quitting u Continuous variable categorized by an arbitrary cutpoint. u Enter "Maintenance" stage even when no change occurs. u Also relevant: - number of years as a smoker, - duration of longest quit attempt, u This theory says nothing about: - how long smokers stay in each of the first 3 stages, and - when they progress to the next stage. u Thus stages are not "periods of time" and the TTM is not a model of when people change, contrary to what is often believed.

Conclusions u Measuring independently and comprehensively each of the four variables included in the stage concept may provide a more thorough understanding of the process of smoking cessation than using an incomplete mix of these variables.

Acknowledgements u This study was supported by grants from the Swiss National Science Foundation to J.-F. Etter ( and , PROSPER).