METHODS: We collected anonymous demographic data, including smoking and tobacco-chewing status and history, as well as exposure to second-hand smoke, from 59 adults aged 21 to 67. Urine and saliva samples were tested using NicAlertTM strips following manufacturer's specifications.
RESULTS: 20 self-reported smokers (S), 20 non-smokers (NS), 14 non-smokers exposed to second-hand smoke (SHS), and 5 tobacco chewers (T) completed the protocol. NicAlertTM strips using urine confirmed smoking status in 17/20 S, 18/20 NS, all 14 SHS, and 4/5 T. NicAlertTM strips using saliva confirmed smoking status in 18/20 S, 17/20 NS, 12/14 SHS, and 4/5 T. Five subjects were misclassified by both urine and saliva tests, including two light daily smokers who smoked more in the evening, two non-smokers who had quit smoking within weeks of the study (one was still chewing Nicaret gum), and one chewer who consumed ½ to 1 tin per week. Three additional subjects were misclassified only by the saliva test, and one was misclassified only by the urine test. Assuming that the self-report data are correct in this sample of adults with no incentive to lie, the NicAlertTM results were accurate in 53/59 (90%) subjects using urine, and in 51/59 (86%) using saliva.
CONCLUSIONS: When they conflict with a subject's self-reported smoking status, NicAlertTM results should not be considered completely accurate in all circumstances. Urine was slightly more accurate than saliva, but not enough to be recommended over saliva.
Supported by NIH grant #R01-DE014889