The Use of Mobile Health Technology and Behavioral Economics to Encourage Adherence in Adolescents

Overview

Familial hypercholesterolemia (FH) affects over one million Americans and increases the risk of cardiovascular disease (CVD) by as much as 20-fold. Although the use of statins can substantially reduce this risk, adherence to statins in adults and adolescence is poor. In adults, lower rates of adherence are associated with an increased rate of CVD events and all-cause mortality, as well as an additional $44 billion annually in health care costs. Novel interventions are needed to improve medication adherence in patients with FH, starting in adolescents. An underused strategy to improve medication adherence incorporates the principles of behavioral economics. Traditional economic theory suggests that providing an incentive to perform a behavior will increase the frequency of that behavior. However, two prominent theories in behavioral economics, Present Bias and Loss Aversion, suggest that not all types of incentives are effective and that poorly structured incentives can actually be negative enforcers. With novel mobile health technologies (mHealth), interventions based on behavioral economics can now be studied on a larger scale. In this proposal, the investigators will test the use of monetary incentives ($30 per 30 days) to improve medication adherence in eligible subjects. The investigators will test the subject's adherence prior to the use of incentives (using the Morisky Medication Adherence Scale and the Wellth Mobile Application) and during the period of time the incentives are provided. Lastly, the investigators will test the subject's adherence (using the Morisky Medication Adherence Scale and Wellth App) during the 60 days following discontinuation of the incentives to determine if any effect of the incentive persists after the incentive is discontinued.

Preventing cardiovascular disease (CVD) in children and adolescents depend partly on the efficacy of the patients' self-management of their disease between clinic visits. One example of a health behavior that is particularly important for patients with CVD is medication adherence. Although the development of statins provides an opportunity to substantially reduce this risk of CVD, adherence to statins in both adults and adolescents is frequently inadequate. In recent study of youth prescribed a statin using claims data, nearly 75% of adolescents have at least one 90-day period of nonadherence. In adults, statin nonadherence is often as low as 50% and is associated with an increased risk for of CVD events and all-cause mortality, as well as almost $44 billion annually in additional health care costs. Unfortunately, successful interventions to improve statin adherence in adults are rare and there are no formally tested interventions to improve statin adherence in adolescents. It is critical novel interventions are developed that address medication adherence in adolescents with chronic diseases that increase the risk of CVD. Studies in adolescents with other diseases demonstrate that they appear to be particularly vulnerable to nonadherence across a wide-variety of conditions, including in adolescents with solid organ transplants and diabetes mellitus. Unfortunately, previous studies using text messaging reminders to take a medication or computerized action plans have not consistently shown positive results. Therefore, the significance of this proposal is that I will pilot a novel mobile health (mHealth) intervention that incorporates two theories in behavioral economics, Present Bias and Loss Aversion, to develop a better understanding of barriers to medication adherence in adolescents with cardiovascular risk factors. Traditional economic theory suggests that providing individuals with incentives to complete an action will lead to increased frequency of the occurrence of the desired action. However, behavioral economics studies in adult populations suggest that not all incentive structures are effective and that some incentives may act like negative enforcers. Studies suggest that patients are biased towards the present and often discount future risks, which is termed Present Bias. For instance, studies of Present Bias suggest that even small rewards, if provided frequently and immediately after a participant completes a desired task can improve adherence in adults. This strategy can be improved by using the principles of Loss Aversion, which suggest that individuals place more value on money or objects that they have than on new objects of the same value. Therefore, incentives framed as a loss can be more powerful than if it is framed as a gain. Unfortunately, only a few studies in adolescents have incorporated behavioral economics into interventions aimed at behavior change. One promising study in youth with type 1 diabetes mellitus found behavioral economics approaches could influence adherence to insulin regimens and glucose monitoring. Whereas diabetes mellitus is symptomatic if not properly managed, statins typically are used as primary prevention in asymptomatic patients with familial hypercholesterolemia (FH), which may require unique incentive structures. Although monetary incentives have been used with varying degrees of success in adults, it is unclear if these findings translate to adolescents. Adolescents may discount future events more, have different attitudes towards health, and have less mature cognitive function and capacities than adults. In addition, interventions that improve medication adherence are often resource intensive, which decreases the likelihood of larger scale implementation. Using an mHealth application may reduce costs by automating processes previously requiring human time and resources. Further, an mHealth intervention may be more successful as it uses a medium that is ubiquitous in adolescence. For instance, mobile applications that use gamification have demonstrated that adolescents find these health applications engaging and that they are interested in using mHealth in the management of their disease across a diverse set of socioeconomic and cultural backgrounds. As many health behaviors track into adulthood, establishing healthy behaviors early is crucial, which requires using developmentally appropriate interventions to promote adherence and ultimately minimize the risk of CVD. APPROACH Determine the efficacy of a monetary incentive to improve medication adherence. Aim 1.1 The investigators will determine the baseline adherence level in adolescents with FH prescribed a statin using the Morisky Medication Adherence Scale (MMAS). Aim 1.2 The investigators will determine the efficacy of a monetary incentive on medication adherence. Patients who are found to have low adherence according to the MMAS in Aim 1.1 will be enrolled into an intervention using the Wellth® mobile health application (Wellth App). Aim 1.3 Next, the investigators will determine adherence 60 days after the incentives have been discontinued using the tracking features in the Wellth App and the MMAS. Description of Wellth Application (Wellth App) The Wellth App has the following three features/functions: 1) it reminds users to take the prescribed medication through alerts/notifications on the smart phone; 2) it records medication adherence by determining whether the user takes the prescribed medication in the allotted time frame through photographic evidence termed a "check in"; and 3) it determines whether the predetermined reward is delivered to the patient instantly. The first function reminds the patient with an audible or haptic alert (i.e., vibration) using the phone's alert notification function. These alerts are not delivered through an email or text messaging service, but rather are similar to news alerts, alarms, or low battery notifications. The second function of the Wellth app is to document whether a patient takes the prescribed medication within the correct period of time. The Wellth App determines adherence by having the patient to take a photograph of the correct dose of medication out of the bottle/container and in the patient's hand using the Wellth app and smart phone's built-in camera. This will provide a time-stamped and individually-linked document of adherence. For example, a valid photograph in the current study would be a photograph of a statin tablet in the patient's hand during a calendar day that starts at 00:00 and ends at 23:59. If no photograph is captured within this time period, the subject is considered non-adherent. Taking more than one dose of the medication another day does not "make up" for previous missed days, even if it is recommended by a health care provider. Participants will be able to download the app to their smartphone without any additional software or encryption needed. In order to prevent patients from being excluded based on the costs of the necessary hardware, the investigators will provide a smartphone and data plan free of charge to patients based on an existing partnership if needed. Patients use the App to receive reminders, make check-ins, check their reward balance, and view their adherence history. Description of the reward system: In this proposal, subjects will receive money to incentivize adherence. Subjects will receive a $30 credit at 30-day intervals, which can only be spent at the end of the 30-day period. For each payout, the full $30 will be earned only if the individual submits all the required photos of his/her medications each day through the Wellth app. For each missed check-in, the subject loses $2 from the amount to be paid out at the end of the 30-day period. In this study, there will be two consecutive 30-day periods in which the subject can earn rewards. After this 60-day period, the subject will continue to use the Wellth App to document adherence for 60 days, but will not receive any rewards during this time period. The final 60-day period is used to determine if the reward period has influence on medication adherence after the incentive is stopped. Adherence will be measured by two methods. The first will be through the Morisky Medication Adherence Scale (MMAS) and the second will be using the proportion of prescribed pills taken. The choice to measure adherence with two methods is because the investigators only want to include patients who have poor adherence in the study. As self-reported adherence may be inaccurate, the investigators will use the validated Morisky Medication Adherence Scale (MMAS) to assess baseline adherence. However, during the study period, adherence will be measured much more accurately through the use of the Wellth App. As the score of the MMAS cannot be converted to the proportion of prescribed pills taken, the investors will simply measure changes in MMAS score throughout the study as well as measure changes in the proportion of prescribed days taken. The investigators provide power calculations for both scenarios below. Participants will complete the MMAS at the screening visit, after the 60-day intervention, and 60 days after the intervention ends. The investigators will measure adherence using the Wellth app continuously once the subject has access to the Wellth app. The adherence values used for analysis will be the proportion of pills prescribed in which the subjects completes a "checked in" during the three following periods: a) "pre-intervention period" of 14 days prior to the start of incentive period, b) the entirety of the Incentive Period, and c) the final 30 days after the Post-Incentive period. Sensitivity analysis will also compare adherence in the last 14 days of each period.