Cut-off levels for breath carbon monoxide as a marker for cigarette smoking.

摘要: Aims  Current clinical studies often use a breath carbon monoxide (BCO) cut-off level of 8 parts per million (p.p.m.) or higher to identify smoking. In this study, the BCO as an indicator smoking over past 24 hours was re-examined. Design  and self-reported were obtained each weekday for up 14 weeks in 213 subjects paid deliver reduced values. Analysis 12 386 paired values reported analyzed. Findings  The 25% quartile, median 75% quartile 1, 1 2 p.p.m. on non-smoking days 2, 5 12 p.p.m. days, respectively. Receiver-operating characteristic (ROC) analysis indicated that provided high diagnostic accuracy distinguish between [area under curve (AUC) = 0.853, P < 0.0001]. highest combined sensitivity specificity observed at 3 p.p.m. (sensitivity = 71.5%; specificity = 84.8%). At 8 p.p.m. 40.6% 98.2%, respectively, indicating many smokers would be falsely classified abstinent. Finally, percentage true tests (positive negative) (80.2%). Conclusions  levels well below 8 p.p.m low 2–3 p.p.m. may more useful when it is important maximize identification abstinence with degree certainty.