Skip to main content

Key Findings

This research had a large sample (n = 2000+), looked at multiple disease types (hyperlipidemia & type II diabetes), and used pharmacy refill data as the measure of adherence (v. self-report). The primary investigation compared the UW-MAP and the "best practice" Morisky et. al. self-report tool. Variants of the UW-MAP were then tested in an attempt to create an adherence predictive model with the greatest combination of face validity and simplicity. The resulting model served as the foundation for development of an adherence tool to be administered in provider practices.


The purpose of the CHESS medication adherence research was to test the effectiveness of a comprehensive medication adherence predictive model capable of both predicting adherence and guiding medication adherence interventions

The Need

Medication adherence is less than optimal, with rates ranging from 40-60%, and the ability to predict medication adherence has been limited. There has been a lack of intervention-based prediction instruments that can diagnose individual adherence barriers, and thus identify target areas for medication adherence. In a classic study, Morisky et. al.1 used self-reported adherence behaviors to predict long-term outcomes for blood pressure control. This tool, as well as other adherence predictive tools based in the Health Belief Model, has demonstrated limited success in intervention applications.


The main focus of the research was on the test of a Bayesian decision theoretic model (called the UW-MAP), but also on investigating ways to improve upon the model. There were two intended products:

  • A medication adherence prediction tool that health care providers can easily use to examine the likelihood of a patient's adherence, identify a patient's barriers to adherence, and prioritize those barriers to guide medication adherence interventions.
  • Integration of this model within an interactive health communication system that will efficiently help patients self-examine their likelihood of successfully adopting a new behavior and advise them on self-management strategies for increasing their chances of success.

Funding Period:
January - September, 1996
Principal Investigators:
David Gustafson Sr., Ph.D