A Randomized Trial of Educational Materials, Pillboxes, and Mailings to Improve Adherence with Antiretroviral Therapy in an Inner City HIV Clinic

• Objective: To determine whether interventions such as medication cards, pillboxes, and mailings of mo-­ tivational messages would impact antiretroviral adherence in an inner city HIV clinic. Factors influ-­ encing adherence were also assessed. • Methods: Patients were randomized to an adher-­ ence intervention protocol or to standard of care. Utilizing an adherence measurement tool, Medication Adherence Self-­Report Inventory (MASRI), we assessed the impact of our interventions on rates of adherence over a 24-­week period. • Results: 50 patients were enrolled: 27 in the inter-­ vention group and 23 in the control group. Only 10 patients (20.4%) completed all study visits. Seven patients (14%) did not return after randomization. Cumulative adherence was 96.7% in the interven-­ tion group and 97.4% in the control group (p > 0.05). Factors predicting adherence included base-­ line CD4 count (CD4 < 200 cells/mm3, 90% adher-­ ence versus CD4 > 200 cells/mm3, 99% adherence; p = 0.03) and active illicit drug use (83% versus 97.4%; p = 0.37). There was no difference based on intervention in CD4 increase or HIV viral load decline at the end of the study period. • Conclusions: Adherence interventions did not result in significant differences in rates of adherence or other outcomes in our study. Baseline lower CD4 counts and active drug use contributed to poorer adherence. High dropout rates in our study suggest that although focusing on adherence is a key aspect of HIV care, finding ways to increase patient reten-­ tion should also be a priority. AIDS morbidity and mortality has dramatically decreased with the advent of highly active antiretroviral therapy (HAART) [1]. These regimens may be difficult for patients to adhere to because of pill burden, dosing frequency, dosing schedule, food requirements, and side effects. Poor adherence significantly contributes to HAART failure [2]. Paterson et al [2] demonstrated that 22% of patients with adherence of 95% or greater had virologic failure as compared with 61% of patients with adherence of 80% to 95%. More recently, Sethi et al [3] demonstrated that both a cumulative adherence of 70% to 89% and missing a scheduled clinic visit in the previous month were independently associated with viral rebound and clinically significant drug resistance. In addition, several studies have shown improved survival in adherent patients compared with nonadherent patients with similar CD4 counts [4,5]. Methods available to measure adherence include the medication event monitoring system (MEMS), which uses medicine containers that mechanically count the number of times they are opened and closed, manual pill counts, patient self-reports, patient interviews, and monitoring of serum drug concentrations. One study showed that self-reported adherence and therapeutic drug monitoring correlated well with MEMS data [6]. However, some trials suggest that self-report, patient interview, and pill count overestimate adherence, and MEMS may underestimate adherence [7,8]. Walsh et al [9] developed a medication adherence self-report inventory (MASRI), a self-administered questionnaire that includes a nonjudgmental statement about medication adherence and asks specific questions about missed medications. It also includes a visual analogue scale that asks patients to mark the location on a line corresponding to the percentage of medications taken over the past month. They compared MASRI with MEMS and pill count and found that MASRI provides a valid method to measure adherence when compared with objective methods [9]. From the Sections of Infectious Diseases and Internal Medicine, Department of Medicine, Temple University School of Medicine, and the Department of Pharmacy Practice, Temple University School of Pharmacy, Philadelphia, PA.