Abstract 20109: Impact of Text Message Medication Reminders on Medication Adherence and Blood Pressure in a High Risk Urban Emergency Department Population

Background: Hypertension disproportionately affects urban African Americans, where medication adherence is often poor. Text message (TM) interventions have shown promise in improving medication adherence. Objectives: To assess the feasibility, acceptability, and impact of daily, automated, and customized TM medication reminders on medication adherence and blood pressure (BP). Methods: We conducted a randomized controlled trial of TM medication reminders with a convenience sample of cell phone owning African Americans with asymptomatic uncontrolled BP. Participants were recruited from an urban emergency department and received a 35-day supply of antihypertensive medications. Outcomes were measured at 30-day follow-up. Results: In total, 65 participants were randomized equally between control and intervention groups, and 57 (88%) completed the study. Our sample was largely male (54%), unemployed (47%), with a high school education (56%), and a total annual household income <$10,000/year (53%). Preliminary analyses revealed great support for the intervention. Out of 29 intervention participants, 100% agreed/strongly agreed that the program was easy to use, 97% would like to keep using the program, 97% were satisfied with the program, and 90% thought the program was helpful. Despite overwhelming acceptability among intervention participants, after 30 days, only 37% of control and 43% of intervention group participants were adherent to medications according to pill counts, and the difference between groups was not statistically significant. Using the Morisky Medication Adherence Scale, there were slight, yet non-significant improvements in medication adherence at 30 days (M = .63 and 1.0 points, respectively). Both groups exhibited improved BP at 30 days (M = -17.6 and -19.4 mm Hg in systolic BP and M = -4.0 and -5.3 mm Hg in diastolic BP for control and intervention groups, respectively), but differences between groups were not significant. Conclusions: The intervention was feasible and acceptable among participants, but no significant differences were found between intervention and control groups at 30-day follow-up. Further work with longer follow-up and a larger sample is warranted to more adequately determine intervention efficacy.