Institutional members access full text with Ovid®

Share this article on:

Self-Report Adherence Measures in Chronic Illness: Retest Reliability and Predictive Validity

Jerant, Anthony MD*; DiMatteo, Robin PhD; Arnsten, Julia MD, MPH; Moore-Hill, Monique MA*; Franks, Peter MD

doi: 10.1097/MLR.0b013e31817924e4
Original Article

Background: Patient self-report is a practical method for measuring adherence, but little is known about its optimal use.

Objectives: To examine the retest reliability and predictive validity of 3 different types of self-report adherence measures among patients with common chronic illnesses.

Research Design: Correlation and regression analyses of data from an ongoing randomized controlled trial.

Subjects: Patients (N = 415) aged ≥40 years recruited from a primary care network with arthritis, asthma, chronic lung disease, congestive heart failure, depression, and/or diabetes mellitus, plus impairment in ≥1 basic activity and/or a score of ≥4 on the 10-item Center for Epidemiologic Studies Depression Scale.

Measures: Self-report adherence (administered variously at baseline, 2, 4, and 6 weeks, and 6 months): number of pills taken/number of pills prescribed (PT/PP), using 1–7 days recall, and global reports of medication adherence and overall adherence tendencies. Six-month functional outcomes: Health Assessment Questionnaire (HAQ) and Short Form-36 (SF-36).

Results: Correlation coefficients among contemporaneously administered 1–7 days PT/PP measures were ≥0.78. Correlations among PT/PP measures and global adherence measures, and among PT/PP measures at 2 and 4 weeks, ranged from 0.11 to 0.54. PT/PP measures using ≥3–4 days recall significantly predicted adjusted 6-month HAQ but not SF-36 score.

Conclusions: Self-report PT/PP and general medication adherence measures tap different behavioral constructs. Self-reported PT/PP at a given point in time is not necessarily representative of medication adherence over time. Among chronically ill patients, 3–4 days recall of PT/PP yield adherence estimates, which are practically as reliable and valid as longer intervals and which predict functional outcomes.

From the *Department of Family and Community Medicine, University of California Davis School of Medicine, Sacramento, California; †Department of Psychology, University of California, Riverside, California; ‡Division of General Internal Medicine, Albert Einstein College of Medicine, Montefiore Medical Center, Bronx, New York; and §Center for Healthcare Policy and Research, University of California Davis School of Medicine, Sacramento, California.

Supported in part by the Agency for Healthcare Research and Quality Grant R01HS013603.

Reprints: Anthony Jerant, MD, Department of Family and Community Medicine, University of California Davis School of Medicine, 4860 Y Street, Suite 2300, Sacramento, CA. E-mail: afjerant@ucdavis.edu.

© 2008 Lippincott Williams & Wilkins, Inc.