Abstract
The validity of self-reported tobacco use is often questioned given the potential for underestimation of use. This study used data from a double-blind, placebo-controlled clinical trial of varenicline for smokeless tobacco dependence in India to evaluate the accuracy of self-reported smokeless tobacco cessation using biochemical validation procedures and to evaluate correlates of reporting inaccuracy. Smokeless tobacco users attending a dental clinic at AIIMS were randomized to placebo or varenicline; all participants received counseling. Detailed smokeless tobacco use was recorded and abstinence was defined as cotinine-verified 7-day point prevalence cessation (cotinine < 50 ng/ml) and breath CO > 10 ppm at the end of 12 weeks of treatment. One-half of study completers (82/165) self-reported abstinence. Biochemical verification confirmed that (65.9%) subjects provided accurate self-reports while (34.1%) participants underreported tobacco use. These data indicate poor agreement between self-reported and biochemically confirmed abstinence (κ = -0.191). Underreporters of tobacco use had significantly higher baseline cotinine (p < 0.05), total craving (p < 0.012), and negative reinforcement craving (p < 0.001) vs. those whose self-reports were correctly verified. These findings provide evidence to support the need for biochemical validation of self-reported abstinence outcomes among smokeless tobacco users in cessation programs in India and identify high levels of pretreatment cotinine and craving levels as potential correlates of false reporting.