Quantifying smoking cessation outcomes: from The Women's Initiative for Nonsmoking Study (X): methodological implications.

Background: Classification of smoking status has a major impact on the conclusions drawn from smoking cessation intervention research, yet few studies have addressed this critical issue.

Objectives: The aim of this study was to compare three classifications (naïve, optimistic, and pessimistic) of smoking cessation outcomes or smoking status from the Women's Initiative for Nonsmoking Study (WINS).

Methods: This is a longitudinal prospective study nested within a randomized clinical trial (RCT) design of WINS, an RCT of 277 women over the age of 18 years who reported smoking cigarettes continuously for 1 month prior to a cardiovascular event requiring hospital admission. Women were randomized to either the usual care group (UC) or the intervention group (IG). Recruitment for WINS occurred between October 1996 and December 1998 in 10 hospitals in the San Francisco Bay area. Follow-up data on smoking status was obtained from the UC and the IG using a structured telephone interview at 6 and 12 months from baseline and was confirmed by family members and salivary cotinine levels.

Results: Seven-day point prevalence(self-report of not smoking in the past 7 days; "not even a puff") using the naïve (the most liberal) classification yields a greater number of nonsmokers than the pessimistic or most conservative classification (cotinine level verification of smoking status). The classification of smoking status also affects time to continuous smoking. The pessimistic classification results in the shortest time to continuous smoking, whereas the opposite is observed with the naïve classification.

Discussion: It is important to critically evaluate the underlying assumptions made by study investigators when measuring and reporting smoking status. The classification of smoking status and the selection of analysis, meaning point prevalence versus survival analysis, affect study results and contribute to the variability observed in the research findings of smoking cessation intervention trials and the challenges faced in making appropriate comparisons across studies.