Workplace Wellness Programs

Wellness programs have been around long enough now that they don’t get all the attention they did in the early years.  As is almost always the case, they didn’t live up to early, probably unwarrantedly optimistic, expectations.  But they continue to be very popular and there is a growing body of research examining not only their impact on health, utilization and spending, but on what features of a program may or may not assist it in being more effective.  A paper at the National Bureau of Economic Research looks at one particular employer’s wellness program, which was designed and operated as a randomized trial.   (NBER Paper)  The study was carried out at the University of Illinois. Out of the thousands of employees invited to participate in the study, over 4800 agreed to do so.  3300 were assigned to the program and given paid time off to participate in program activities.   They also received an incentive, which varied from $50 to $350, randomly assigned, to complete program activities.  The control group consisted of about 1500 employees who were not permitted to participate.  The authors used claims, surveys and other data for their outcome analyses.  The primary questions they examined were the effect of incentives, was their employee selection into the program and what type of workers selected in or out, and the effects of participation on health spending, productivity and health behaviors.

The program included biometric screening, an online health risk assessment, and classes and coaching on the usual topics, such as smoking cessation and stress management.  There was a special incentive for the screening, which varied randomly across participants.  Raising the incentive from 0$ to $100 increased screening from 47% to 59%, but raising it another $200 only added four more percentage points.  So a $100 incentive will do a pretty good job of getting people to have a screening.  In general, however, the researchers found that the incentives tied to the wellness activities in general were more effective than those specific to screenings.  There was selection of workers into the program, with relatively healthier, lower spending workers most likely.  People at the highest and lowest extremes of spending were the least likely to participate.  There were no significant effects from participating in wellness activities, however, on total spending, productivity, health behaviors, or self-reported health outcomes in the first year of the program.  Disappointing but not inconsistent with prior research, which shows minimal impact from these programs, at least in the early years.

The researchers believe that the results reflect some beneficial evidence in regard to attracting and retaining workers through the use of wellness programs.  Participants rated the employer higher on caring about worker health and safety.  They also note that these programs, to the extent that they reward participation by incentives or premium reductions, and assuming the employer has a set total pool of compensation, have distribution of compensation impacts which tend to disadvantage lower-income workers, often in poorer health, who seem least likely to participate.  Greater efforts may need to be made to find a way to encourage participation by low-income, poorer health employees.  I do think that one year is too short a period for full evaluation of the effects of any health intervention of this nature.  A minimum of three years should be used, so hopefully the authors will keep the study going and report on additional analyses over future years.