Can We Trust National Smoking Prevalence Figures? Discrepancies Between Biochemically Assessed and Self-Reported Smoking Rates in Three Countries

BACKGROUND: National smoking prevalence estimates are the primary basis for assessing progress in tobacco control across the world. They are based on surveys of self-reported cigarette smoking. It has been assumed that this is sufficiently accurate for policy purposes, but this assumption has not been adequately tested. METHODS: We report data from the 2003 Health Survey for England, the U.S. National Health and Nutrition Examination Survey for 2001-2002, and the 2004 national smoking behaviors survey in Poland as examples of countries at different stages in the "tobacco epidemic." Self-reported cigarette and total tobacco smoking prevalence were assessed by means of the standard questions used in each country. In subsamples, specimens were collected for analysis of cotinine (saliva, N = 1,613 in England; serum, N = 4,687 in the United States; and saliva, N = 388 in Poland) providing an objective means of determining active smoking. A cut point of 15 ng/mL was used to discriminate active smoking from passive smoke exposure. RESULTS: Self-reported cigarette smoking prevalence using the standard methods underestimated true tobacco smoking prevalence by an estimated 2.8% in England, 0.6% in the United States, and 4.4% in Poland. Cotinine concentrations in those misclassified as nonsmokers were indicative of high levels of smoke intake. INTERPRETATION: Underestimation of smoking prevalence was minimal in the United States but significant in England and Poland. A review of methodologies for assessing tobacco smoking prevalence worldwide is urgently needed.