Predictors of medication adherence: fact or artifact

Area of Science:

• Health Services Research
• Behavioral Science
• Clinical Pharmacy

Background:

• Numerous studies have explored factors influencing medication adherence, yet findings on predictors remain inconsistent.
• Variability in research outcomes suggests potential confounding factors, possibly including adherence measurement methods.

Purpose of the Study:

• To investigate whether different methods of measuring medication adherence identify different sets of predictor variables.
• To analyze adherence predictors in relation to socio-demographic, psychosocial, and clinical factors across distinct measurement techniques.

Main Methods:

• Utilized data from two randomized controlled trials involving adherence interventions.
• Examined 21 days of baseline data from 698 subjects using the Aardex Medication Event Monitor and the Morisky Self-Report Scale.
• Employed Spearman rho, Pearson r, and multiple logistic regression analyses to identify predictors.

Main Results:

• Significant differences were observed in the identification of adherence predictors based on the measurement method used.
• Socio-demographic, psychosocial, and clinical factors showed varying associations with adherence depending on whether electronic monitoring or self-report was employed.

Conclusions:

• The findings support the hypothesis that adherence measurement methodology influences the identification of predictor variables.
• Researchers and clinicians should be aware that different adherence assessment tools may reveal different insights into patient behavior and influencing factors.