Related Experiment Video
Updated: Jun 2, 2026
A Machine Learning Approach to Design an Efficient Selective Screening of Mild Cognitive Impairment
Published on: January 11, 2020
Defining the minimal detectable change in scores on the eight-item Morisky Medication Adherence Scale
Paul Muntner1, Cara Joyce, Elizabeth Holt
1University of Alabama at Birmingham, Birmingham, AL, USA. pmuntner@uab.edu
Background:
Self-report scales are used to assess medication adherence. Data on how to discriminate change in self-reported adherence over time from random variability are limited.
Objective:
To determine the minimal detectable change for scores on the 8-item Morisky Medication Adherence Scale (MMAS-8).
Methods:
The MMAS-8 was administered twice, using a standard telephone script, with administration separated by 14-22 days, to 210 participants taking antihypertensive medication in the CoSMO (Cohort Study of Medication Adherence among Older Adults). MMAS-8 scores were calculated and participants were grouped into previously defined categories (<6, 6 to <8, and 8 for low, medium, and high adherence).
Results:
The mean (SD) age of participants was 78.1 (5.8) years, 43.8% were black, and 68.1% were women. Overall, 8.1% (17/210), 16.2% (34/210), and 51.0% (107/210) of participants had low, medium, and high MMAS-8 scores, respectively, at both survey administrations (overall agreement 75.2%; 158/210). The weighted κ statistic was 0.63 (95% CI 0.53 to 0.72). The intraclass correlation coefficient was 0.78. The within-person standard error of the mean for change in MMAS-8 scores was 0.81, which equated to a minimal detectable change of 1.98 points. Only 4.3% (9/210) of the participants had a change in MMAS-8 of 2 or more points between survey administrations.
Conclusions:
Within-person changes in MMAS-8 scores of 2 or more points over time may represent a real change in antihypertensive medication adherence.
