Multiphase Optimization Trial of Incentives for Veterans to Encourage Walking

Regular physical activity (PA) is essential to healthy aging. Unfortunately, only 5% of US adults meet guideline of 150 minutes of moderate exercise; Veterans and non-Veterans have similar levels of PA. A patient incentive program for PA may help. Behavioral economics suggests that the chronic inability to start and maintain a PA routine may be the result of "present bias," which is a tendency to value immediate rewards over rewards in the future. With present bias, it is always better to exercise tomorrow because the immediate gratification of watching television or surfing the internet is a more powerful motivator than the intangible and delayed benefit of future health. Patient incentives may overcome present bias by moving the rewards for exercise forward in time. Recent randomized trials suggest that incentives for PA can be effective, but substantial gaps in knowledge prevent the implementation of a PA incentive program in Veterans Affairs (VA). First, incentive designs vary considerably. They vary by the size of the incentive, the type of incentive (cash or non-financial), the probability of earning an incentive (an assured payment for effort or a lottery-based incentive), or whether the incentive is earned after the effort is given (a gain-framed incentive) or awarded up-front and lost if the effort is not given (a loss-framed incentive). The optimal combination of these components for a Veteran population is unknown. Second, the evidence about the effective components of incentives comes from studies conducted in populations that were overwhelmingly female; often employees at large companies, with high levels of education and income. VA users, in contrast, are mostly male and lower income, and most are not employed. This is important because the investigators have theoretical reasons to believe that the effects of components of incentives are likely to vary by income and gender. Finally, few studies have managed to design an incentive such that the physical activity was maintained after the incentive was removed. Indeed, a common theme in incentivizing health behavior change is the difficulty in sustaining behavior change once the incentives are removed.

The investigators propose to fill the research gaps through a Multiphase Optimization Strategy (MOST) trial of incentives for walking. A MOST trial is ideally suited for situations in which a proposed intervention has many potential intervention components. A MOST trial consists of three phases. A screening phase trial is used to efficiently identify-through a factorial designed randomized trial-the effective components of a complex intervention like incentives. A refining phase trial tests the optimal dose (size or duration) of the incentives. A confirmatory phase trial tests the optimal components and dose against a usual care control. The goal of the proposed study is to conduct the screening phase trial in 128 Veterans to identify the optimal components of incentives for increasing walking among physically inactive Veterans. All Veterans in this phase will be given various components of incentives for increasing average steps per day to 7,000 steps over a 12-week habit-building period, and then maintaining the increase through a 12-week habit maintenance period. The specific aims are: Aim 1: Conduct a 24 factorial designed screening-phase trial of incentives for increasing average steps per day to 7,000 steps over 12 weeks among physically inactive Veterans. Every patient in the trial will be given a Fitbit Inspire activity monitor and assigned to a group that receives different components of incentives. The investigators will test four different incentive factors: 1) lottery vs. loss framed incentives, 2) financial vs. non-financial incentives, 3) a pre-commitment postcard reminder of a Veteran's stated intrinsic reason for commitment to PA vs. no pre-commitment postcard, and 4) a request for PA advice from a Veteran on staying active vs. no request. The first factor has never been tested in a population like the VA. Factors 2-4 are designed specifically to sustain the effects of incentives after the incentive is removed. Factor 4 is a novel hypothesis that has never been tested outside of educational research: specifically, that asking for advice from a Veteran is more motivating than giving advice to them, even if that Veteran is struggling with low physical activity themselves. The primary outcome is change in steps per week from baseline to week 24. Aim 2. Conduct cost analyses and qualitative interviews. The cost of administering each component and qualitative assessments of the acceptability of each component to trial participants will inform the decision of which components to retain for the subsequent refining and confirmatory phase trials. Aim 3. Convene an expert panel to choose components for the next phases of the MOST trial. The panel will weigh each component in terms of its effect on step counts (Aim 1), administrative costs and participant-reported qualitative assessments (Aim 2), and the strength of the theoretical basis for the component's effect on physical activity.