Validity of self-reported smoking status: Comparison of patients admitted to hospital with acute coronary syndrome and the general population

Abstract

Many studies rely on self-reported smoking status. We hypothesized that patients with acute coronary syndrome (ACS), a smoking-related condition, would be more prone to misclassify themselves as ex-smokers, because of pressure to quit. We compared patients admitted with ACS with a general population survey conducted in the same country at a similar time. We determined whether ACS patients who classified themselves as ex-smokers (n = 635) were more likely to have cotinine levels suggestive of smoking deception than self-reported ex-smokers in the general population (n = 289). On univariate analysis, the percentage of smoking deceivers was similar among ACS patients and the general population (11% vs. 12%, p = .530). Following adjustment for age, sex and exposure to environmental tobacco smoke, ACS patients were significantly more likely to misclassify themselves (adjusted OR = 14.06, 95% CI 2.13-93.01, p = .006). There was an interaction with age whereby the probability of misclassification fell significantly with increasing age in the ACS group (adjusted OR = 0.95, 95% CI 0.93-0.97, p<.001), but not in the general population. Overall, smoking deception was more common among ACS patients than the general population. Studies comparing patients with cardiovascular disease and healthy individuals risk introducing bias if they rely solely on self-reported smoking status. Biochemical confirmation should be undertaken in such studies.