Demographic, Clinical, and Psychosocial Predictors of Change in Medication Adherence in the Support, Educate, Empower Program

Abstract

Purpose: To investigate whether demographic, clinical, or psychosocial factors act as moderators of change in medication adherence in the Support, Educate, Empower (SEE) program.

Design: Prospective, single-arm pilot study with a pre-post design.

Participants: Patients with glaucoma aged ≥ 40 years and taking ≥ 1 glaucoma medication were recruited from the University of Michigan Kellogg Eye Center. Those who had electronically measured adherence ≤ 80% in the 3-month eligibility monitoring period were enrolled in the SEE program.

Methods: Medication adherence was monitored electronically during the 7-month intervention and calculated as the percentage of doses taken correctly. Change in adherence at different points in the SEE program and cumulative change in adherence were modeled with linear regression, and baseline demographic, clinical, and psychosocial factors were investigated for significant associations.

Main outcome measures: Demographic, clinical, and psychosocial variables associated with change in medication adherence in the SEE program.

Results: Thirty-nine participants completed the SEE program. These participants were on average 63.9 years old (standard deviation [SD], 10.7 years), 56% (n = 22) were male, 44% (n = 17) were White, and 49% (n = 19) were Black. Medication adherence improved from an average of 59.9% (SD, 18.5%) at baseline to 83.6% (SD, 17.5%) after the final SEE session, for an increase of 23.7% (SD, 17.5%). Although participants with lower income (< $25 000 and $25 000-50 000 vs. >$50 000) had lower baseline adherence (48.4% and 64.1% vs. 70.4%), these individuals had greater increases in adherence during the first month of medication reminders (19.6% and 21.6% vs. 10.2%; P = 0.05 and P = 0.007, respectively). Participants taking fewer glaucoma medications also had significantly greater increases in adherence with medication reminders (P < 0.001). Those with higher levels of glaucoma-related distress (GD) had lower baseline adherence and greater increases in adherence with glaucoma coaching (P = 0.06).

Conclusions: Patient-level factors associated with relatively greater improvements in medication adherence through the SEE Program included lower income, fewer glaucoma medications, and increased GD. These findings demonstrate that the SEE program can improve glaucoma self-management even among participants with social and psychological barriers to medication adherence.

Keywords: Counseling; Glaucoma; Medication adherence; Medication reminders; Motivational interviewing; Personalized education; Support.

Figure 1.

Line plots displaying mean adherence at baseline and during the Support, Educate, Empower (SEE) intervention, stratified by demographic characteristics of patients, including A. age, B. sex, C. race, D. education, E. income, and F. race and income. Standard deviations are provided in Table 2. Triangles denote the mean adherence at baseline. Abbreviations: HS, high school.

Figure 2.

Line plots displaying mean adherence at baseline and during the Support, Educate, Empower (SEE) intervention, stratified by clinical features of patients, including A. better-eye mean deviation (MD), B. better-eye logMAR visual acuity (VA), C. number of glaucoma medications, D. total number of daily medication doses, E. number of comorbid conditions, and F. better-eye glaucoma severity. Standard deviations are provided in Table 2. Triangles denote the mean adherence at baseline.

Figure 3.

Line plots displaying mean adherence at baseline and during the Support, Educate, Empower (SEE) intervention, stratified by psychosocial measures self-reported by patients, including A. glaucoma-related distress, B. emotional burden, C. consideration of future consequences (CFC), D. national eye institute visual function questionnaire (NEIVFQ) composite score, E. functional health literacy (FHL), and F. glaucoma medication self-efficacy (GMSE). Standard deviations are provided in Table 2. Triangles denote the mean adherence at baseline.