Key Barriers to Medication Adherence in Survivors of Strokes and Transient Ischemic Attacks

Even though medications can greatly reduce the risk of recurrent stroke, medication adherence is suboptimal in stroke survivors. To identify key barriers to medication adherence in a predominantly low-income, minority group of stroke and transient ischemic attack (TIA) survivors. Cross-sectional study. Six hundred stroke or TIA survivors, age a parts per thousand yen 40 years old, recruited from underserved communities in New York City. Medication adherence was measured using the 8-item Morisky Medication Adherence Questionnaire. Potential barriers to adherence were assessed using validated instruments. Logistic regression was used to test which barriers were independently associated with adherence. Models were additionally controlled for age, race/ethnicity, income, and comorbidity. Forty percent of participants had poor self-reported medication adherence. In unadjusted analyses, compared to adherent participants, non-adherent participants had increased concerns about medications (26 % versus 7 %, p < 0.001), low trust in their personal doctor (42 % versus 29 %, p = 0.001), problems communicating with their doctor due to language (19 % versus 12 %, p = 0.02), perceived discrimination from the health system (42 % versus 22 %, p < 0.001), difficulty accessing health care (16 % versus 8 %, p = 0.002), and inadequate continuity of care (27 % versus 20 %, p = 0.05). In the fully adjusted model, only increased concerns about medications [OR 5.02 (95 % CI 2.76, 9.11); p < 0.001] and perceived discrimination [OR 1.85 (95 % CI 1.18, 2.90); p = 0.008] remained significant barriers. Increased concerns about medications (related to worry, disruption, long-term effects, and medication dependence) and perceived discrimination were the most important barriers to medication adherence in this group. Interventions that reduce medication concerns have the greatest potential to improve medication adherence in low-income stroke/TIA survivors.