Study objectives: To assess the optimal cutoff level of breath CO concentration to distinguish actual smokers from nonsmokers among patients with asthma and COPD.
Setting: Kyoto University Hospital outpatient clinic.
Subjects and methods: Three hundred thirty-one consecutive outpatients (161 with asthma and 170 with COPD) were examined cross-sectionally by self-reported smoking status, breath CO monitoring, and serum cotinine concentration. Actual smoking status was verified by serum cotinine concentration.
Results: Mean serum cotinine concentrations of never smokers, former smokers, and current smokers with asthma were 6.0 ± 5.2 ng/mL, 12.1 ± 25.0 ng/mL, and 198.3 ± 181.7 ng/mL, respectively (± SD). Mean serum cotinine concentrations of former smokers and current smokers with COPD were 23.2 ± 69.2 ng/mL and 191.1 ± 109.8 ng/mL, respectively. Mean breath CO levels of never smokers, former smokers, and current smokers with asthma were 6.1 ± 2.4 ppm, 7.7 ± 3.2 ppm, and 19.9 ± 17.3 ppm, respectively. Mean breath CO levels of former smokers and current smokers with COPD were 7.7 ± 4.3 ppm and 13.5 ± 6.5 ppm, respectively. The optimal cutoff level of breath CO to discriminate between actual smokers and nonsmokers was 10 ppm in patients with asthma and 11 ppm in patients with COPD, giving 85.0% and 73.1% sensitivity, and 85.8% and 84.7% specificity, respectively.
Conclusion: The optimal cutoff level of breath CO to assess actual smoking status was 10 ppm in patients with stable asthma and 11 ppm in patients with stable COPD. In patients with asthma and COPD, breath CO levels were potentially influenced by underlying airway inflammation, suggesting misclassification in the assessment of smoking status by breath CO.