Recommendation for the assessment of tobacco craving and withdrawal in smoking cessation trials

Abstract

This paper addresses methodological issues in the assessment of nicotine withdrawal and craving in clinical trials of smoking cessation therapies. We define withdrawal as a syndrome of behavioral, affective, cognitive, and physiological symptoms, typically transient, emerging upon cessation or reduction of tobacco use and causing distress or impairment of behavioral function. Offset effects (effects related to removal of a direct nicotine effect) are sustained effects of cessation or reduction of tobacco use that cause distress or impairment. Withdrawal and craving are important as potential predictors of relapse, as mediators and markers of treatment effects, and as clinical phenomena in their own right. Symptoms recommended for assessment include craving, irritability, depression, restlessness, sleep disturbance, difficulty concentrating, increased appetite, and weight gain; anxiety deserves further study. We recommend reporting of data on each of these individual symptoms, and use of multiple-item assessments. Although some standardized measures of withdrawal have promising psychometric properties, no measure has yet fully established its reliability, validity, and broad applicability and, therefore, we do not currently favor universal adoption of any one measure. Assessment of objective indices of withdrawal (e.g., hormonal changes) is currently technically challenging and of unknown value. Although weekly assessment may suffice in some large trials, more intensive measurement can provide better sensitivity. Analyses of withdrawal should include baseline measures and be sensitive to potential instability in baseline. Analytic approaches should take into account potential bias when only abstinent subjects are examined. Conversely, heterogeneity should be considered when smoking subjects are included in intent-to-treat analyses. Withdrawal data from clinical trials focused on assessing abstinence rates may be biased because of progressive subject loss to dropout and relapse; different designs and approaches are needed to investigate the process and natural history of craving and withdrawal.