Leveraging Behavioral Economics-based Interventions to Improve Medication Adherence

Abstract

Poor medication adherence to chronic medications can lead to added patient harm and increased costs. Concepts from behavioral economics have guided interventions to help overcome the tendency toward non-adherence with chronic medications. For example, commitment contracts, whereby people either put their reputation on the line (social incentive) or deposit money that they receive back only if they succeed (financial incentive), have substantial conceptual appeal as a method of changing health behavior. This dissertation assessed the relative effectiveness of behavioral economics-based interventions using financial or social incentives on enhancing medication adherence. A mixed-methods design was implemented. Data collection and analyses were conducted in 2 major phases. Phase I was quantitative and the research design was a longitudinal, randomized, controlled trial comparing the effectiveness of an intervention to improve adherence to antihypertensive or antihyperlipidemic medications that provides incentives in the form of a financially incentivized commitment contract vs. a socially incentivized commitment contract vs. usual care (no commitment contract). Participants were randomized to one of three groups: usual care (UC), financial incentives (FI), or social incentives (SI). Data collection: 1) electronically measured medication adherence via a Medication Event Monitoring System (MEMS) vial that electronically recorded a date and timestamp upon each vial opening; Daily adherence was measured over a 90-day. The FI group received $90 upfront, with $1 deducted each day a dose was missed. The SI group utilized a study website that displayed individual and group medication adherence for participants to see. The UC group were instructed to take their medications as prescribed. 2) Two self-reported questionnaires, baseline and 90-day follow-up, assessed demographics, self-reported adherence (via Medometer1), socioeconomic status, subjective social status (via MacArthur Scale of Subjective Social Status2), and perspectives of differing incentives. Phase II was qualitative and was implemented in two parts. Part 1 involved semi-structured interviews and Part 2 utilized focus group discussions to explore and understand to what extent social incentives may be applied to motivate medication adherence and healthful behaviors. Analysis of phase I, 15 participants were randomized to 1 of 3 groups (UC=6, FI=3, SI=6); 1 dropout and 3 lost to follow-up. The majority were female (57%), white (86%), currently married (57%), retired (50%), and had a combined household income of $50,000-$100,000 (57%). Age ranged from 40 to 82 years (mean=59±13). Mean percentage (SD) of MEMS-measured adherence, measured over 90 days, increased from UC group (77% 34%), to SI group (87% 20%), to FI group (95% 6%), but the differences were not statistically significant, F(2,10) = .492, p =.626. Participant perceptions of the incentives, indicated on baseline survey, suggested that financial incentives were moderately effective and the social incentives (i.e., wanting their family or pharmacist to see they are adherent) were moderately effective. Analysis of phase II, using thematic analysis, identified four themes among the participants of both focus group sessions: 1) Accountability, 2) Motivation, 3) Barriers and Solutions, and 4) Technology. Many use apps, smartphones, or wearable devices to help them stay on track with health behavior goals. Very few track medication adherence using an app however most disapproved of having social incentive features associated with a medication-taking app. Stating medication-taking behaviors are “private” and sharing this information is “too personal”. Others felt that medication-taking isn’t necessarily a goal or that social features attached to medication-taking aren’t trendy or appealing. Although the underpowered study limits statistical interpretation the results still provide meaningful insight to applying behavioral economic-interventions to medication adherence. Future research should seek to refine the methodology, namely adopt proactive recruiting strategies thus increasing sample size, prioritize low baseline adherence enrollees, and a diverse population so that the results are generalizable to a larger population. In consideration of the social incentive, a strategic approach should be taken to distinguish the effects of distinct social incentives on medication adherence and other health behaviors.