Background: Smoking is the main cause of lung cancer, so data linked to smoking behaviour are important in lung cancer screening trials. However, self-reporting data concerning smoking behaviour are mainly used. The aim of this study was to biochemically determine the validity and reliability of self-reported smoking status among smokers at high risk for developing lung cancer participating in the Dutch-Belgian lung cancer screening (NELSON) trial. Method: For this sub study, a random sample of 475 men was selected who were scheduled for the fourth screening round in the NELSON trial. They were asked to complete a short questionnaire to verify the smoking behaviour for the previous seven days and a blood sample was collected to measure the cotinine level. The validity (sensitivity (Se), specificity (Sp), positive predictive value (PPV), and negative predictive value (NPV)) and reliability (Kappa) of the self-reported smoking status compared to the cotinine level (as golden standard) were determined. Results: Both a completed questionnaire as well as a cotinine level were available for 199 (41.9%) participants. Based on these data, Se and Sp were respectively 98% (95%-Confidence Interval (CI): 91-99) and 98% (95%-CI: 93-100). PPV and NPV were 98% and 96% and Kappa was 0.96. Conclusion: In conclusion, the validity of the self-reported smoking status turned out to be reliable amongst men at high risk for developing lung cancer who participate in the NELSON lung cancer screening trial.

, , , ,,
Lung Cancer
Department of Public Health