Longitudinal Barriers to Thiopurine Adherence in Adolescents With Inflammatory Bowel Diseases

• Published in: Journal of pediatric psychology Vol. 44; no. 1; pp. 52 - 60
• Main Authors: Plevinsky, Jill M, Wojtowicz, Andrea A, Miller, Steven A, Greenley, Rachel N
• Format: Journal Article
• Language: English
• Published: United States Oxford University Press 01-01-2019
• Subjects: Adolescent; Child; Cross-Sectional Studies; Female; Humans; Immunosuppressive Agents - therapeutic use; Inflammatory Bowel Diseases - drug therapy; Male; Medication Adherence; Mercaptopurine - analogs & derivatives; Mercaptopurine - therapeutic use; Self Report; longitudinal research; adherence; inflammatory bowel disease
• DOI: 10.1093/jpepsy/jsy062
• EISSN: 1465-735X
• ISSN: 0146-8693

Abstract
Objectives
Cross-sectionally, more adherence barriers are associated with lower medication adherence. However, little is known about longitudinal associations between adherence barriers and adherence. Among adolescents with inflammatory bowel diseases (IBD), this study examined both (1) how time-varying self-reported adherence barriers affect daily thiopurine adherence and (2) how adherence barriers at baseline affect daily thiopurine adherence over a six-month period.
Methods
Eighty-one adolescents 11–18 years old prescribed a once-daily oral IBD maintenance medication participated in a six-month observational study. Adherence barriers were self-reported monthly via the Medication Adherence Measure (MAM): Medication Subscale. Daily adherence estimates were collected via Medication Event Monitoring System (MEMS) Track Caps.
Results
Generalized linear mixed modeling indicated that time alone did not significantly predict whether one was more likely to be adherent (p = .602). However, increasing adherence barriers lowered the likelihood that a participant would be adherent on a given day, and the interaction between time and barriers predicted likelihood of adherence on a given day (p < .01). Specifically, when participants reported no adherence barriers at baseline, adherence did not significantly change over time (p = .369). However, when barriers were endorsed, adherence decreased over time (p < .01).
Conclusions
Fewer adherence barriers over time predicted greater likelihood of adherence on a given day, which is consistent with previous cross-sectional research. Routine assessment of barriers to adherence over the course of adolescence is critical in addressing suboptimal adherence behavior in youth with IBD.