Abstract
INTRODUCTION: There is a paucity of studies evaluating passive smoking (PS) by comparing self-report (SR) and biomarkers. Our study aimed to confirm whether SR could accurately reflect PS compared to biomarkers, a golden standard for assessing the exposure of non-smokers. METHODS: We used the 2014-2020 Korea National Health and Nutrition Examination Survey data and selected 29622 non-smokers aged >19 years as the study participants. The PS rate by SR was assessed during the last 7 days, and participants were interviewed to investigate their exposure at home, work, indoors, and in public places. In addition, participants having a limit of detection ≥0.5 ng/mL in urine cotinine (UC) was defined as the exposure group. All analyses reflected the weights of complex sampling. We first compared the rates of PS using biomarkers and SR, and then the sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) were calculated based on biomarkers. RESULTS: PS exposure by UC was the highest (44.4%), and the exposure by SR was significantly lower (5.1-29.5%). Kappa and sensitivity in PS in the indoor home (HPS) were lower than those in indoor workplaces (WPS) and indoor public places (PPS). Moreover, overall sensitivity and PPV were lower, and specificity and NPV were relatively higher in accuracy. Lastly, the sensitivity was poor, and the specificity was relatively good, which means that measurement by SR would identify people who were actually exposed to PS as non-exposed. CONCLUSIONS: Despite exposure to PS, the use of the SR method is more likely to classify participants in the non-exposed group. Hence, to overcome measurement error in SR and reflect exposure in any place and setting, biomonitoring and SR should be performed.