The 13th World Conference on Tobacco OR Health

Objective: Little is known about the validity of self-reported smoking behaviour although this is of interest for the investigation of smoking patterns eg in studies of smoking cessation or of the influence of public policy interventions. This study investigated the precision of smokers' perception of their puffing behaviour and potential factors that may influence this.

Methods: In a cross-sectional study, a portable smoking topography device, the CReSSmicro, was used to assess the puffing behaviour of thirty smokers in the UK, who were also asked to provide information on their smoking behaviour and smoking dependence (heaviness of smoking, HSI). Self-reported puffing behaviour was compared to machine measured puffing behaviour on four variables (number of puffs, peak puff flow, puff volume and interpuff interval) and differences between self-reported and machine measured results were used to compute an overall index of accuracy.

Results: Results indicate that smokers were not aware of their own smoking behaviour as there were no significant correlations between self-reported and actual smoking behaviour on any of the four variables. Overall accuracy was however influenced by HSI. More dependent smokers were relatively better at assessing their own smoking behaviour than less dependent smokers (χ2=5, p=.025). This effect remained even after controlling for age, sex, low or regular tar yield and length of time of smoking (OR 2.5, CI 1.1-5.9). Implications: Self-reported smoking behaviour was not a valid tool especially for less dependent smokers. Results will be compared to Canadian and US samples to establish the reliability of these findings.