Background: Exhaled carbon monoxide (CO(Exh)) measurement is used to confirm smoking status in smoking cessation programs, but the cut-off level is still a matter for discussion. The objective of this study was to compare CO(Exh) levels in smokers and non-smokers to validate the method in a Brazilian population and to estimate the probability of the patient still smoking according to different cut-off points.
Methods: In this cross-sectional study we studied non-atopic Caucasian volunteers with no respiratory infection or steroid therapy in the preceding four weeks. Exclusion criteria were: pregnancy; breast feeding; age<18 and >65 years old; and subjects not signing informed consent. Participants filled out a questionnaire and had their CO(Exh) levels measured. Bayes' theorem was used to calculate the posttest probabilities.
Results: We included 393 subjects of whom 239 (61%) were smokers. The mean CO(Exh) was 14.7 +/- 9.4 ppm and 4.3 +/- 2.5 ppm (p<0.001) in smokers and nonsmokers, respectively. Patients with CO(Exh) below 8 ppm had a likelihood ratio below 1 of still smoking. The levels 9 ppm and 10 ppm provided likelihood ratios of 1.50 and 1.93, respectively. Better discriminant power was obtained at >11 ppm, when the likelihood ratio became 63.80 (95%CI 16.1-253.1).
Conclusions: In smoking cessation practice, a likelihood ratio approach may be useful to determine the probability that an individual is still smoking according to various CO(Exh) cut-off points instead of using a fixed value for all patients.