(A-001) Analyzing Medication Adherence as a Mediator in Disease Management

Background: Medication adherence plays a critical role in optimizing treatment outcomes, preventing disease progression, and reducing healthcare costs. However, its role as a mediator in pharmacoepidemiology remains understudied. Investigating this mediating effect can provide insights into improving therapeutic effectiveness and patient care outcomes.

Objectives: This study aims to assess medication adherence as a mediator in disease management, using schizophrenia as an example. Specifically, it examines the extent to which emergency department (ED) visits can be explained by adherence levels, providing insights into its impact on healthcare utilization and patient outcomes.

Methods: This retrospective cohort study analyzed data from the South Carolina Medicaid database, focusing on schizophrenia patients from 2004 to 2019. The primary exposure was race, with a one-year follow-up period after treatment initiation. Medication adherence, measured by the proportion of days covered (PDC), was assessed as a mediator during the first six months of follow-up. The primary outcome, ED visit, was measured during the second six months of follow-up. Descriptive statistics summarized patient characteristics, and causal mediation analysis using the ‘mediation’ R package evaluated the indirect effect of adherence on ED visits.

Results: Among 1,666 schizophrenia patients initiating antipsychotics, 517 (31%) had at least one ED visit, while 1,149 (69%) did not within six months post-adherence measurement. The cohort included 500 White (30%) and 1,166 Black (70%) patients. ED visits occurred in 27% of White patients and 33% of Black patients (p = 0.031). Medication adherence mediated 23% of the total effect on ED visits (p = 0.004). The Average Causal Mediation Effect (ACME) was significant (p < 0.001), while the Average Direct Effect (ADE) of race on ED visits was not (p = 0.096).

Conclusions: Medication adherence significantly mediates the relationship between race and ED visits, with 23% of the total effect attributed to medication adherence. These findings highlight the importance of managing medication adherence in reducing ED utilization among schizophrenia patients. Further research should explore additional mediators and interventions to improve adherence and optimize patient outcomes. This analytic approach in adherence research can be applied to different diseases to better understand and quantify the mediated effect of adherence on disease management.