Abstract Body (Do not enter title and authors here): Hypothesis and Purpose: Nonadherence to antihypertensive medications is common. Mobile health (mHealth) based behavioral economic approaches may improve adherence, but remain largely untested, especially in vulnerable populations. We designed BETTER-BP to test whether an mHealth-enabled incentive lottery would lower SBP and improve adherence.
Study Design and Methods: RCT at 3 safety-net clinics in NYC. Participants were followed for 6 months after randomization (2:1). Adherence in both groups was assessed with a wireless electronic monitoring device (EMD, Adheretech). For patients on multiple BP medications, one agent was chosen. SBP was measured in-person at baseline and 6 months. The trial enrolled from 7/14/20-10/9/24.
Sample Size: 400 participants (265 intervention:135 control).
Population: Inclusion: (1) age ≥18 years, (2) English- or Spanish-speaking, (3) hypertension diagnosis, (4) on BP medication, (5) SBP >140 mmHg in past year, (6) poor self-reported adherence. Exclusion: barrier to technology use (e.g. vision impairment).
Intervention: A regret lottery incentive program was administered via SMS text message. If medication was taken on a given day, the participant received a message coupled with a potential financial incentive (low chance of large payment, higher chance of small payment). If medication was not taken, participants received a message indicating that had they been adherent, they may have received payment.
Power Calculation: 400 participants with 20% attrition provided 85% power to detect a 10mmHg difference in SBP change between intervention and control arms.
Primary End Point: Change in SBP between baseline and 6 months.
Secondary End Point: Adequate adherence (≥80% days adherent) over 6 months.
Outcome(s): Mean baseline SBP was 139 (SD 19) mmHg, median age was 57 years, 60.5% were women, 61.5% were Hispanic, and 20.3% were non-Hispanic Black. Over 70% had safety-net insurance (Medicaid) or no insurance. The most common comorbidities were obesity (54.5%) and diabetes (46.5%). 6 month retention was 85%. Change in SBP did not differ significantly between arms (mean SBP change -6.0 mmHg vs. -6.9 mmHg, P=0.62) (Fig. 1). However, a higher proportion had adherence ≥80% with intervention (70% vs. 34%, P<0.001) (Fig. 2).
Conclusion: In a diverse safety-net population, the BETTER-BP intervention doubled the rate of adequate medication adherence but did not reduce SBP compared with control. Reasons for this discrepancy require further exploration.