A Field Experiment of the Effect of Cost Sharing on An Employee Weight Loss Program

An important unresolved question in the incentive literature is how to optimally set patient cost-sharing for services that require ongoing patient engagement. Standard economic theory would suggest that lowering prices to zero (or even below zero) would maximize program attendance; however, the goal of this program is not attendance per se but weight loss and it is possible that people who come to the program because it is lower cost or free will value it less and thereby exert themselves less and be less successful in losing weight. In other words, lowering the price by subsidizing the price of attendance might increase attendance but also result in decreased weight loss conditional on attendance, relative to a full price program. It is also unclear how much attendance would change with changes in price and whether the increased attendance would be worth the cost of the subsidies.

Utilization of cost-effective treatments or preventive services (e.g. weight loss support group meetings, as we test in this field experiment), could be increased by reducing their price. Such services could potentially save money to an employer by encouraging benign moral hazard and reducing the rate of adverse events from poorly controlled health and by increasing employer productivity. However, the effects of varying copayments for prescription drugs, for example, appear to be quite different depending on whether the copayments are increased to existing patients or decreased in the hope of attracting non-adherent patients. Increases in prescription copayments have been associated with decreases in medication adherence and increases in the rate of poor health outcomes in numerous studies, but reductions in copayments have not produced corresponding increases in adherence. Reasons for this are that it is much more difficult to change behavior in non-adherent patients (their demand curves are more inelastic) and the manner in which these programs have been implemented may have made the price reductions insufficiently salient to consumers.

In this paper, we report the results of a 12 month, 4-arm randomized controlled trial (N=1240) of different levels of employer subsidization of an at-work, group-support-based weight loss program run b by a major commercial weight loss program. Participants were randomized to one of four arms: a 100% subsidy (i.e., program is free); a 0% subsidy (i.e. program is full-price); a 50% subsidy (i.e. program is half price); versus a “hybrid” arm in which participants received a 50% subsidy upfront, with possibility for 100% subsidy based on attaining monthly participation goals.

Enrollment is now complete and by June, 2014 we will have results on the effect of the subsidization on three key outcomes of interest: 1. Uptake (the % of patients who sign up for the program, by arm); 2. Efficacy (program attendance and weight loss over a 12-month period); and 3. Effectiveness outcome among all patients to whom the program was offered.

This study will provide insight into how to set patient cost-sharing for services that require ongoing patient engagement.