The extent of non-adherence in a large AIDS clinical trial using plasma dideoxynucleoside concentrations as a marker.

Objectives: To assess adherence to study medications in an AIDS clinical trial, to evaluate whether study participants adhered to only one component of a multidrug regimen ('differential adherence'), and to determine whether there was evidence of non-uniform adherence to study medications among treatment groups. Setting: This was a substudy of AIDS Clinical Trials Group protocol 175, a large, double-blind, randomized study of monotherapy versus combination dideoxynucleoside therapy. Participants were required to adhere to a complex regimen of zidovudine, zalcitabine and didanosine, or their matching placebos. Design: Between October 1992 and January 1994, study sites were selected at random, and a 1-week period was designated during which study participants attending routine clinic visits provided a blood sample and dosing history. Participants were not informed of the purpose of the substudy. Measurements: Adherence was assessed using plasma drug concentrations and defined by the presence of detectable drug in a plasma sample obtained within a specified analysis window. Results: Of 722 plasma samples analyzed, approximately 75% contained detectable concentrations of the assigned drugs and 5-14.5% contained no detectable drugs. Approximately 7 and 13% of samples from participants assigned to monotherapy arms contained non-prescribed dideoxynucleosides, and 14 and 19% assigned to combination therapies contained, only one drug. Conclusions: Various non-adherence behaviors were observed, including patterns of underdosing and taking non-prescribed drugs. Non-adherence was moderate but uniform amongst the treatment groups and may have contributed to a marginal reduction in the power of the primary intent-to-treat analysis to detect differences in efficacy amongst the assigned treatments.