1. The law of non-usage attrition in a technology-based behavioral intervention for black adults with poor cardiovascular health
- Author
-
Muhammed Y. Idris, Mohamed Mubasher, Ernest Alema-Mensah, Christopher Awad, Kofi Vordzorgbe, Elizabeth Ofili, Arshed Ali Quyyumi, and Priscilla Pemu
- Subjects
Computer applications to medicine. Medical informatics ,R858-859.7 - Abstract
Digital health innovations, such as telehealth and remote monitoring, have shown promise in addressing patient barriers to accessing evidence-based programs and providing a scalable path for tailored behavioral interventions that support self-management skills, knowledge acquisition and promotion of relevant behavioral change. However, significant attrition continues to plague internet-based studies, a result we believe can be attributed to characteristics of the intervention, or individual user characteristics. In this paper, we provide the first analysis of determinants of non usage attrition in a randomized control trial of a technology-based intervention for improving self-management behaviors among Black adults who face increased cardiovascular risk factors. We introduce a different way to measure nonusage attrition that considers usage over a specific period of time and estimate a cox proportional hazards model of the impact of intervention factors and participant demographics on the risk of a nonusage event. Our results indicated that not having a coach (compared to having a coach) decreases the risk of becoming an inactive user by 36% (HR = .63, P = 0.04). We also found that several demographic factors can influence Non-usage attrition: The risk of nonusage attrition amongst those who completed some college or technical school (HR = 2.91, P = 0.04) or graduated college (HR = 2.98, P = 0.047) is significantly higher when compared to participants who did not graduate high school. Finally, we found that the risk of nonsage attrition among participants with poor cardiovascular from “at-risk” neighborhoods with higher morbidity and mortality rates related to CVD is significantly higher when compared to participants from “resilient” neighborhoods (HR = 1.99, P = 0.03). Our results underscore the importance of understanding challenges to the use of mhealth technologies for cardiovascular health in underserved communities. Addressing these unique barriers is essential, because a lack of diffusion of digital health innovations exacerbates health disparities. Author summary Mobile-health (mHealth) technologies continue to serve immensely valuable roles in current healthcare systems by bridging several accessibility gaps and facilitating patient-provider interactions. However, and ironically, they also present several accessibility gaps due to design features and/or individual characteristics. In this study, we explore the effects of having an adjunct health coach with a behavior-modification mHealth app, Health360x (H360x) in promoting user-interface interaction in a population of African American patients with significant cardiovascular (CV) risk factors. While we saw an improvement in CV risk factors, we also identified an interesting attrition in app usage in patients assigned a health coach; and observed similar effects in those with higher socioeconomic status. These findings may be due to a higher (baseline or acquired) self-efficacy in these patients, and suggests that longitudinal assessments of app engagement in this population should be approached with careful consideration of attrition.
- Published
- 2022