Biological tests such as blood levels of medication metabolites are not subject to response bias or misreporting and, therefore, are selleck Imatinib generally deemed the most accurate measure of adherence (Vermeire, Hearnshaw, Van Royen, & Denekens, 2001). These tests reflect both compliance and rate of elimination of the drug and are more likely than other measures to represent medication actually taken by allowing researchers to directly determine the extent to which the medication is present in a participant��s blood stream. However, they are invasive and costly, may not be available for all medication types (Vermeire et al., 2001), may be influenced by individual differences in pharmacokinetics (Bosworth, 2006), and involve samples being collected within a relatively narrow window of time.

Self-report has been utilized as an indicator of adherence and is convenient and easy to administer (Bosworth, 2006). The self-report measures utilized in this study have been supported in HIV adherence research; higher medication adherence using 3-day recall and visual analogue scale (VAS) was associated with lower HIV viral load (Giordano, Guzman, Clark, Charlebois, & Blangsberg, 2004; Walsh, Madalia, & Gazzard, 2002). Several authors (Haynes, Ackloo, Sahota, McDonald, & Yao, 2008; McDonald, Garg, & Haynes, 2002) have recommended objective adherence measures such as pill count and MEMS because participants tend to overestimate their medication adherence via self-report. While these methods are more objective than self-report, they do not measure actual ingestion of the prescribed medications.

Thus, each assessment method offers relative advantages accompanied by its respective limitations. The paucity of research on adherence to smoking cessation pharmacotherapy may be partly attributed to a lack of consensus on the best methods for measuring adherence. Given the effectiveness of varenicline for smoking cessation and the importance of adherence to its effectiveness, this study compared the validity of three commonly used adherence measures, pill count, 3-day recall, and VAS, against plasma varenicline concentration for measuring medication adherence among African American smokers enrolled in a pilot trial of varenicline. Methods Participants Data are from a randomized pilot study of varenicline in combination with adherence support or standard care counseling for smoking cessation among African American smokers (Nollen et al., 2011). All participants in this study received varenicline. Of 72 enrolled participants, 11 were lost to follow-up at Day 12, 5 did not have pill count data for GSK-3 the 3 days prior to Day 12, and 1 took twice the prescribed dose of varenicline, leaving a final sample of 55 participants in the current study.