Health Policy$ense

Move It Or Lose It

Mitesh Patel Study Sheds Light on Employee Incentives

The latest study by LDI Senior Fellow Mitesh Patel and colleagues adds to our growing understanding of how best to frame financial incentives to encourage healthy behaviors, and employer wellness managers should take note. Patel and colleagues’ randomized control trial, published today in the Annals of Internal Medicine, shows that financial incentives framed as a loss (upfront rewards that are reduced when goals aren’t achieved) are most effective in increasing physical activity among overweight and obese adult employees of a major university (ok, it’s Penn).

That’s the good news if you’ve set up your wellness program that way. But two other incentive structures tested – a small daily reward and the opportunity to be entered into a daily lottery - had no effect relative to the control. And there was no lasting effect of increased daily physical activity for any of the intervention groups.

In Patel’s study, conducted with collaborators from LDI’s Center for Health Incentives and Behavioral Economics, the researchers apply behavioral economics concepts to design and test  incentive programs. The 26-week randomized control trial relied on a smartphone application to track whether nearly 300 adult employees (78% of whom were women) with a BMI of more than 27 walked more than 7,000 steps per day.

For the 13-week intervention period, participants were randomly assigned to a control group with daily feedback or 1 of 3 financial incentive programs with daily feedback plus: a gain incentive ($1.40 given each day the goal was achieved), lottery incentive (daily eligibility if goal was achieved), or loss incentive ($42 allocated monthly upfront and $1.40 removed each day the goal was not achieved). Participants were followed for another 13 weeks with daily performance feedback but no incentives. Only the loss-incentive group had a significantly greater average proportion of participant-days achieving the goal (45%) than the control group (30%). The loss-incentive group had greater mean daily steps (5,880) than the control group (5,031), but they were not significantly different. During follow-up, the loss-incentive group still had the highest average proportion of participant-days meeting the goal (30%) but it was not significantly different from the control group.

The Affordable Care Act (ACA) allowed group health plans to adopt wellness program incentives, or “health-contingent” wellness programs, that vary a person’s premiums or cost-sharing based on achieving certain health objectives. As a result employer wellness programs have skyrocketed in popularity. According to a Kaiser Family Foundation employer survey, 81% of large employers and 49% of small employers offer employees programs to help them stop smoking, lose weight, or make other lifestyle or behavioral changes. Of firms offering health benefits and a wellness program, 38% of large firms and 15% of small firms offer employees a financial incentive to participate in or complete a wellness program.

Detractors note that we have limited evidence that wellness programs work, or how to best design them. This study follows hot on the heels of the results from a different trial, also by Patel and some of the same co-authors, that showed premium-based incentives for weight loss have no effect on employees. Taken together, these studies cast doubt on the effectiveness of certain types of financial incentives, but that doesn’t mean that all types of financial incentives don’t work.  This nuance was lost in much of the media coverage of the findings of the premium-based incentives study.  

How incentive programs are designed clearly makes a big difference in how effective they are in changing behavior, and whether these changes persist. The new evidence gets us closer to an understanding of what works, and what doesn’t.