Testing Promising Behavioral Economic Interventions to Promote Enrollment Diversity in Cardiovascular Cohort Studies

Randomized clinical trials (RCTs) are the gold standard for determining the effectiveness and safety of drugs, medical devices, or healthcare delivery interventions. However, because RCTs require participants to consent to enrollment and randomization, typically only a minority of eligible patients enroll, raising concerns regarding the generalizability of trial results. Though Black and Hispanic individuals are disproportionately more likely to have uncontrolled risk factors for cardiovascular disease and to develop cardiovascular disease, they are substantially underrepresented in cardiovascular RCTs. In one review of 143 cardiovascular clinical trials conducted between 2008 and 2017, just 2.1% of enrollees were Black and 2.1% were Hispanic. Such low rates of enrollment have persisted despite mandates by funding and regulatory authorities that trial sponsors aim to enroll patient populations in RCTs that reflect the racial and ethnic diversity of real-world populations of patients with the disease under study. The failure of these recommendations to meaningfully increase the racial and ethnic diversity of RCT participants indicates a need for new and innovative strategies.

Behavioral economics is a scientific field of inquiry that incorporates principles from economics and psychology to better describe how individuals behave and to influence their behavior. While standard economics assumes that individuals are rational expected utility maximizers, behavioral economics recognizes that individuals' decision-making is limited by thinking capacity, available information, and time. As such, decision-making is governed by heuristics, or cognitive shortcuts. For instance, rather than trying to assess the myriad risks and benefits of participating in a research study, individuals may decide not to participate because the invitation to participate isn't from someone whose name they recognize or a perception that people like themselves do not participate. Knowledge of common heuristics can be leveraged to influence choices by modifying the context in which decisions are made, such as by changing default options or how choices are framed. Marketing studies have demonstrated the effectiveness of appeals to altruism, perceived scarcity ("This product is desirable because it is hard to get"), and social proof ("Are people like me using this product or service?") in increasing responsiveness to outreach, and these techniques can be re-purposed to increase representativeness of RCTs. Challenging social circumstances can be thought of as a "tax" on individuals' cognitive bandwidth, underscoring the opportunity to increase RCTs' racial and ethnic diversity by simplifying the enrollment process to make it easier to enroll, increasing the trustworthiness of communications, and reframing participation to make it more attractive.

The reasons for low enrollment of patients from underrepresented racial and ethnic groups in cardiovascular RCTs are multifactorial. Typically, RCTs that enroll vast majorities of patients from global regions with few Black or Hispanic patients will not enroll populations representative of U.S. demographics. Moreover, Black and Hispanic individuals are often less willing than White individuals to participate in RCTs; they are more likely to face barriers in accessing health services and to have mistrust of medical researchers, due to concerns about historic and ongoing scientific misconduct.

Newer means of communication, including patient portals, may increase or decrease representativeness depending on how they are deployed. However, few studies have systematically evaluated recruitment strategies in randomized fashion. As with any intervention, the effectiveness of different recruitment strategies is best assessed by testing different alternatives in RCTs. Accordingly, the overarching goal of this project is to conduct a series of RCTs to rigorously, systematically, and iteratively test strategies designed to increase enrollment of Black and Hispanic individuals in prospective cohort studies without diminishing trust. The investigators will embed RCTs of changes to the decision-making context, or "nudges," in the context of recruitment for prospective cohort studies because these studies' large sample size will facilitate accelerated rapid-cycle testing of multiple nudges, with results from each recruitment RCT incorporated in the next round of testing. Since the nudges are layered on top of existing cohort recruitment methods - changing only the method of outreach, message framing, or incentive structure - they are low cost, and the number of nudges that can be tested is bound primarily by the size of the cohort to be enrolled. The investigative group is uniquely positioned to conduct these studies. First, the investigators have conducted some of the only RCTs evaluating different RCT recruitment strategies. Second, they have established relationships with multiple cohort studies enrolling tens of thousands of patients with cardiovascular diseases. Third, they have extensive experience applying insights from behavioral economics to health behaviors.

The RCTs will sequentially test different strategies for participant enrollment in the following realms: method of outreach, message framing and default settings, and enrollment incentives. The primary outcome of each recruitment RCT will be the enrollment fraction of Black and Hispanic participants - the number of Black and Hispanic patients enrolled divided by the number attempted to be contacted. Because Black and Hispanic populations are heterogeneous, the investigators will also report the effect of each approach on enrollment fraction stratified according to socioeconomic status (SES), assessed using the Area Deprivation Index, a community-level SES metric. Other key secondary outcomes will include the overall enrollment fraction and the population-to-prevalence ratio, defined as the enrollment fraction of Black and Hispanic patients divided by the overall enrollment fraction. Across sequential RCTs of behaviorally-informed recruitment strategies, the investigators will accomplish the following specific aims:

Aim 1. To evaluate the effect of method of outreach on representativeness. First, patients will all receive a message via email (or traditional mail, if no email address is on file) with information about the cohort study and how to enroll and will be randomized to receive an additional message via text message or patient portal. In a subsequent RCT, potential participants will be randomized to receive this message from their personal clinic or from the research team.

Aim 2. To evaluate the effect of message framing on representativeness across two domains - appeal to altruism versus social proof versus neither. The investigators will evaluate the different framing choices in one RCT that will be embedded in an existing large cohort study.

Aim 3. To evaluate different incentive structures on representativeness. We will conduct a five-arm trial to test how incentive structure (no incentive vs. guarantee only vs. guarantee + small lottery vs. mid lottery only vs. large lottery only) affects enrollment fraction of Black and Hispanic patients.