Comparison of three carbon monoxide monitors for determination of smoking status in smokers and nonsmokers with and without COPD

In this (CAMOXI) study, three carbon monoxide (CO) monitors and salivary cotinine are assessed regarding their ability to distinguish smokers from nonsmokers, both in chronic obstructive pulmonary disease (COPD) and healthy people. Twenty-six healthy smokers, 25 healthy nonsmokers, 25 smoking, and 25 former smoking stable COPD patients (age 40-72 years) were included based on self-report (N = 101). All volunteers were measured following a 12-h abstinence period. Sensitivity, specificity, and predictive values of a positive and negative test result were assessed for a range of cutoff points for both CO and salivary cotinine. The prescribed 9-ppm cutoff point of the Breath CO (R) generates a sensitivity of 68% and 42% for COPD patients and healthy people, respectively. Using the prescribed cutoff point (10 ppm) the Smokelyzer (R) produces 56% sensitivity for COPD patients and 23% for healthy people. Both monitors generate 100% specificity in both groups. The cutoff point for the Micro CO meter (R) (5 ppm) generates 88% sensitivity and 92% specificity for COPD patients, and for healthy people 92% and 88%, respectively. The optimal cutoff points depend upon the goal of the test. Salivary cotinine has a 100% sensitivity, specificity, positive predictive value, and negative predictive value over the range of 15 ng/mL through 40 ng/mL for healthy participants and at 10 ng/mL for COPD patients. The prescribed cutoff points for all three CO monitors generate misleading results concerning the determination of the smoking status in both populations. Salivary cotinine measurement outperforms CO measurements and a combination of the two tools is recommended.