Accuracy of the Smoking Questionnaire

Abstract

Summary The smoking questionnaire (SQ), a multidimensional questionnaire covering the major dimensions of cigarette smoking, was developed to address the heterogeneity in the assessment of smoking exposure. It consists of eight questions and can be completed within a few minutes. Test-retest reliability of the SQ and concurrent validity with the Behavior Risk Factor Surveillance System (BRFSS) 2011 questionnaire were examined in a clinical study conducted in adult US current menthol cigarette smokers. The SQ and the BRFSS were self-administrated twice before and after randomization with a 6-day interval. The inter-temporal analyses included current smokers aged 22 to 66 years who completed the SQ at both timepoints. The percent agreement of items and 95% confidence intervals were calculated for the comparisons between the two timepoints and between the SQ and the BRFSS questionnaire. To evaluate the feasibility of the SQ and to capture subjects’ opinions about the SQ, a meta-questionnaire was administrated. High test-retest reliability levels (percent agreement of > 70 to 100% between the two timepoints) were found for SQ smoking behavior items, in particular for items related to current smoking status, 100-cigarettes lifetime, regular smoking, age of initiation and preferred brand. Moderate (55% agreement) to high test-retest reliability (84% agreement) was found for daily consumption of manufactured cigarettes. The comparison between the SQ and the BRFSS 2011 showed a high concurrent validity (98 to 100% agreement). The SQ was completed on average in 3 to 4 min and was assessed as easy to use. The findings demonstrate that the SQ is reliable in smokers and a practical tool to assess smoking exposure in clinical studies.