Related Experiment Video
Updated: Apr 4, 2026
Using Continuous Data Tracking Technology to Study Exercise Adherence in Pulmonary Rehabilitation
Published on: November 8, 2013
Effect of format on comprehension of adherence data in chronic disease: A cross-sectional study in HIV
Anita Lyons1, Warren B Bilker1, Janet Hines2
1Department of Biostatistics and Epidemiology, Perelman School of Medicine, University of Pennsylvania, PA, United States.
Objective:
To determine which formats communicate medication adherence effectively to patients.
Methods:
HIV-infected adults on antiretrovirals viewed examples of refill data in 5 formats: (1) percentage of doses, (2) number of days late to refill ("days late"), (3) calendar of days with/without medications, (4) pie chart of days with/without medications, and (5) letter grade. Five scenarios (>95%, 90-95%, <90%, <80% and <70% adherence) were presented in each format. Participants rated scenarios on adherence improvement needed. "Good understanding" was ≥ 4 of 5 scenarios correct. We calculated odds ratios for "good understanding" using logistic regression with percentage as the referent format.
Results:
124 participants were median age 48.5 years, 65% Black, 71% male. Understanding of all formats differed by education (all interaction p values<0.02). For ≤ 12 years education, odds ratios (OR) of understanding (95% CI) compared to percentage were: days late 3.3 (2.3-4.7), calendar 3.1 (2.2-4.3) pie chart 2.0 (1.4-2.7), and letter grade 1.8 (1.3-2.5). For >12 years education, ORs were: days late 1.3 (0.9-2.0), calendar 2.4 (1.5-3.8), pie chart 2.9 (1.8-4.6), and letter grade 1.7 (1.1-2.6). Calendar plot was most preferred.
Conclusions:
Adherence percentage was the least understood format regardless of education.
Practice Implications:
Calendars should be used to convey adherence information.
