“…Blood level assays of paroxetine would have provided a more trustworthy index of whether a given patient was taking any paroxetine at all, but paroxetine blood levels provide a poor index of the extent to which a patient has taken the prescribed amount of medicine because of very high variance in plasma concentrations in patients taking the same dose of paroxetine (e.g., Normann, Horn, Hummel, Grunze, & Walden, 2004). Furthermore, reliable relationships between plasma levels and clinical outcome have not been obtained in studies of paroxetine therapy (e.g., Kuhs, Schlake, Rolf, & Rudolf, 1992; Normann et al, 2004; Tasker, Kaye, Zussman, & Link, 1989). Judgments of compliance with CT were provided by subjective ratings that combined the therapist's estimates of the patient's use of CT techniques, homework completion, and session attendance.…”