Background: Enquiry into smoking status and recommendations for smoking cessation is an essential preventive service. However, there are few studies comparing self-reported (SR) and cotinine-verified (CV) smoking statuses, using medical check-up data. The rates of discrepancy and under-reporting are unknown.
Methods: We performed a cross-sectional study using health examination data from Healthcare System Gangnam Center, Seoul National University Hospital in 2013. We analyzed SR and CV smoking statuses and discrepancies between the two in relation to sociodemographic variables. We also attempted to ascertain the factors associated with a discrepant smoking status among current smokers.
Results: In the sample of 3,477 men, CV smoking rate was 11.1% higher than the SR rate. About 1 in 3 participants either omitted the smoking questionnaire or gave a false reply. The ratio of CV to SR smoking rates was 1.49 (95% confidence interval [CI], 1.38-1.61). After adjusting for confounding factors, older adults (≥60 years) showed an increased adjusted odds ratio (aOR) for discrepancy between SR and CV when compared to those in their twenties and thirties (aOR, 5.43; 95% CI, 2.69-10.96). Educational levels of high school graduation or lower (aOR, 2.33; 95% CI, 1.36-4.01), repeated health check-ups (aOR, 1.45; 95% CI, 1.03-2.06), and low cotinine levels of <500 ng/mL (aOR, 2.03; 95% CI, 1.33-3.09), were also associated with discordance between SR and CV smoking status.
Conclusion: Omissions and false responses impede the accurate assessment of smoking status in health check-up participants. In order to improve accuracy, it is suggested that researcher pay attention to participants with greater discrepancy between SR and CV smoking status, and formulate interventions to improve response rates.
Keywords: Cotinine; Mass Screening; Self Report; Smoking; Surveys and Questionnaires.