| Publication Abstract Display | | Type: Published Abstract | | Title: Pilot randomized controlled trial of Care4Today Mobile Health Manager for improving antiretroviral adherence among persons living with HIV. | | Authors: Moore DJ, Rooney AS, Umlauf A, Gouaux B, Philips K, Vaida F | | Year: 2015 | | Publication: 5th Wireless Health Conference, Bethesda, MD | | Volume: Issue: Pages: | | Abstract:Background: Despite dosing formulations that simplify antiretroviral therapy (ART), ART adherence difficulties remain. Consequences of ART non-adherence include viral rebound and development of drug-resistant HIV strains. A mobile health (mHealth) application capable of optimizing adherence to multiple medications would be highly beneficial.
Purpose: To determine whether a randomized controlled trial using a smartphone application (Care4Today Mobile Health Manager) to alert individuals to take ART medication will improve medication adherence and medication dose timing among HIV+ persons as compared to those in a control condition.
Methods: Janssen Research & Development, LLC, sponsors of the present investigator-initiated project, developed Care4Today Mobile Health Manager (CARE), a mobile app that provides adherence reminders at participant-selected times. Participants were HIV-infected, on ART, and self-reported less than perfect ART adherence (i.e., detectable plasma viral load (VL), <100% adherence on a visual analogue scale, VAS). Sixty participants were randomized 1:1 to CARE or a no intervention control condition (CTRL). A single ART medication was tracked for a median of 34 days (Range: 29-55) using Medication Event Monitoring System Caps (MEMS). A strict definition of adherence was used: percent MEMS adherence (doses taken/doses prescribed) within ±2 hours of the dose time. Unadjusted differences in mean adherence and dosing time (minutes from indicated time) were compared using independent samples t-tests. Longitudinal data were analyzed using ordinal mixed effects model with subject specific random intercept, regressing ordinal variable doses taken (none/some/all) on treatment (CTRL/CARE), time (in days), and their interaction.
Results: Study retention was excellent; 57 participants (CARE n=28; CTRL n=29) completed follow-up. Dropouts were significantly older than completers (median age (IQR)=57[55, 67] vs. 49[44, 56], p=0.02), but did not differ on other baseline demographic or adherence variables. The proportion with baseline detectable VL did not differ by arm (CARE=42.9% vs. CTRL=51.7%, p=0.6). Mean MEMS adherence for CARE was 66.8% (SD=27.0%) as compared to 57.5% (SD=32.2%) for CTRL. The unadjusted differences between the two groups represents a small to medium effect size, but was not statistically significant (Cohen’s d=0.31, p=0.24). In multivariable analyses, a medium effect size was observed for tighter dose timing windows in the CARE group as compared to CTRL (Diff=35min, Cohen’s d=−0.52, p=0.06). A longitudinal analysis showed a non-significant interaction term (p=0.69) indicating no difference in adherence changes over time between arms. A subsequent model with baseline VAS adherence also showed no treatment effect (p=0.79), but better VAS scores were associated with greater probability of MEMS adherence (OR=1.04, p=0.02).
Conclusions: A medium effect size was observed for smaller dose-timing windows among HIV-infected persons using a mHealth adherence tool and suggests Care4Today may improve ART dose timing in a larger sample of HIV-infected persons; however, there were no significant overall adherence differences by group. With that said, better ART dose timing has been shown to lead to superior therapeutic coverage and better HIV disease outcomes. As ART adherence interventions for HIV-infected persons evolve, mobile health tools need to be considered as part of a comprehensive package to enhance long-term ART adherence. |
return to publications listing
|