1954

Background As a major metabolite of tobacco-specific carcinogen, 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanol (NNAL) has been used as a biomarker for tobacco smoking exposure. Among current or former smokers, accurate self-reported smoking status is critical for determining the eligibility for lung cancer screening (LCS). This study used NNAL to verify self-reported smoking status among LCS-eligible population from a nationally representative sample in the National Health and Nutrition Examination Survey (NHANES). Methods We utilized NHANES data from 2007-2014 that included both urine total NNAL and self-reported smoking behavior to select the LCS-eligible population. We applied four cutpoints of NNAL (10pg/ml, 20pg/ml, 30pg/ml, 40pg/ml) as the gold standard to assess the performance of self-reported smoking status. To determine the validity and reliability of self-reported smoking status, we calculated the sensitivity (Se), specificity (Sp), positive predictive value (PPV), negative predictive value (NPV), accuracy, and reliability (Kappa). Results We found approximately 3.7% (n=985, equivalent to 8.5 million weighted population) of adults eligible for LCS according to the United States Preventive Services Task Force guideline. Validity and reliability of self-reported smoking status increased with increasing cutpoints of NNAL. Against NNAL-based smoking status with a cutpoint of 40pg/ml, self-reported smoking status showed the highest Se (88.8%), lowest Sp (99.4%); the corresponding PPV and NPV, respectively, was 99.5% and 85.5%; the accuracy was 93.0% and kappa was 0.86. The self-reported smoking status tended to be more valid and reliable among females than males across all NNAL cutpoints evaluated. Using the established 40pg/ml cutpoint as the gold standard, we found the accuracy was lowest among the non-Hispanic black (90.1%) and highest among the other race/ethnicity groups (93.9%). Conclusion Utilizing urine NNAL as the biochemical verification reference, we found that self-reported smoking status among people who are at risk of lung cancer is reasonably reliable.