“…However, pill count may not be accurate in estimating adherence; it tends to overestimate medication adherence and yield an imprecise metric of the total dose taken (Farmer, 1999; Haynes, Ackloo, Sahota, McDonald, & Yao, 2008; McDonald, Garg, & Haynes, 2002; Pullar, Kumar, Tindall, & Feely, 1989). A biological measure, derived from blood, saliva, or urine concentrations of medication, is generally considered to be the most accurate measure of adherence and is commonly used in other research fields such as epilepsy (Dutta & Reed, 2006; Herkes & Eadie, 1990; Landmark, Rytter, & Johannessen, 2007; Malone, Eadie, Addison, Wright, & Dickinson, 2006; Mitchell, Scheier, & Baker, 2000; Specht, Elsner, May, Schimichowski, & Thorbecke, 2003; Vermeire, Hearnshaw, Van Royen, & Denekens, 2001; Williams et al, 2001). Given varenicline’s pharmacokinetic attributes such as minimal metabolism and long half-life (approximately 24 hours) (Faessel et al, 2010), assessment of varenicline is a valid method for biologically assessing varenicline adherence.…”