[Veracity of smokers' reports of abstinence at smoking cessation clinics]

To assess the reliability of smokers' response as criteria for measuring abstinence and the necessity or not of confirming abstinence with carbon monoxide (CO) measurement.

A multicenter, prospective, longitudinal study was carried out on patients over 18 years of age from 5 smoking cessation clinics who underwent treatment with nicotine or bupropion. When the patient attended the clinic at 15, 30, 60, 90, and 180 days, abstinence was checked by self-reporting and expired-air CO levels. Sensitivity, specificity, and positive, negative, and overall predictive value of patient reporting, measured CO levels, and the 2 procedures in combination were calculated.

A total of 904 smokers (476 men and 428 women) with a mean (SD) age of 42.51 (10.09) years were enrolled in the study. Of the 904 patients that made up the study population, 820, 776, 687, 719, and 679, respectively, attended the scheduled visits to check abstinence. Self-reported point-prevalence abstinence at 15 days was 74.5% and at 180 days was 57.6% while abstinence determined by expired-air CO was 75.7% and 59.4% respectively. Results according to self-reporting, CO measurement, and the 2 methods in combination were not significantly different (P<.05) at any of the points in time. Neither sensitivity nor specificity showed significant differences in relation to patient variables.

The reliability of self-reported abstinence from smoking is high. Measurement of CO is therefore not essential, although it could be advisable for motivating patients rather than as a way of confirming abstinence.