Abstract
Tobacco use is a chronic relapsing disease, and remains the leading cause of preventable death in much of the world. Increasingly, tobacco use, chiefly cigarette smoking, is being framed as a chronic disease, with periods of use and periods of abstinence. An implicit component of this conceptualization is that treatment-both counseling and pharmacotherapy-may be needed at various intervals for extended periods of time, perhaps over an individual's lifetime. This would mirror the treatment of other chronic conditions, such as diabetes, hypertension, or hyperlipidemia. Yet, clinical trials of tobacco dependence treatment still generally model outcome measures in terms of cessation, abstinence, or quitting, measured at discrete time points. This reinforces the notion that smoking, or tobacco dependence, is a dichotomous condition, and that one is either "cured," or not. Although the goal of treating tobacco dependence is to ensure long-term abstinence (i.e. "quitting"), this model is discordant with clinical reality, in which of periods of tobacco use are interspersed with periods of abstinence. Hence, the goal of treatment is to lengthen the duration of the latter, while shortening the duration of the former. In the clinical arena, this dichotomous model of tobacco use is reflected in electronic health records, where smoking is generally categorized as current, former, or never. We propose that clinicians move away from the dichotomous categorization of tobacco use, and adopt methods used to categorize the status of other chronic conditions. Specifically, biomarkers such as carbon monoxide, cotinine, and anabasine, measured at regular intervals, can provide clinicians with much clearer, clinically relevant and actionable assessments of current tobacco use by their patients. This can be done without making reference to dichotomous states such as current or former use of tobacco. In psychological terms, one can frame tobacco use in terms of states, attributes in specific situations at discrete moments in time, rather than the more durable traits.