Another Disappointing Wellness Study

Wellness programs have become somewhat ubiquitous but a little passe in degree of hype, as “digital” health has stolen their thunder.  Research is still being conducted to ascertain impact on outcomes and the Journal of the American Medical Association reports on a recent study at a large employer.   (JAMA Article)  This is a well-designed study, utilizing one employer with over 200 sites in the eastern United States.  Twenty worksites were randomized to the wellness program, twenty as primary controls with no intervention, and the rest were status quo secondary controls.  Enrollment in the wellness program was voluntary, raising the possibility of self-selection.  Outcomes were tracked over an 18 month period.  The program was fairly typical, comprised of 8 modules on things like nutrition, exercise, and stress reduction.  There was a modest incentive to complete a module.   Self-reported health behaviors were one outcome, as were clinical measures.  Health care utilization and cost was gathered through claims data, but this only covered the subset of workers who enrolled in the insurance plan.  There were about 4000 workers in each of the treatment and primary control groups and attempts to adjust for various factors were made.

About 60% of the employees were full-time and they were generally not highly paid.  Program participation ranged from 12% in the first module to an average of 31% in the later ones.  Only 35% of workers participated in even one module.  25% completed the self-assessments.  Those employees in the wellness intervention group reported higher levels of exercise and weight management.  There may have been a modest improvement in smoking rates and alcohol use, depending on the statistical test used.  There was no apparent improvement in actual clinical measures such as high lipid levels, obesity or blood pressure in the intervention group versus the control one.  Total health care spending for the wellness participants was about $200 per year less, on around a $5000 per year total spend, a non-significant change in either utilization or cost.  While the study appears to be disappointing, it did result in some improved health behaviors for those who participated.  The absence of any real incentive to either participate or to change behavior undoubtedly affects the results, as does the short follow-up time.  I would really like to see results from a mandatory wellness program with real carrots and sticks for behavior change, with three to five-year results.

While research is important to evaluate the effect and value of various health improvement efforts, it can also help guide development of more effective programs by identifying which features and tactics work best.  Getting people to be engaged in their health and health care is supremely important.  “Healthy” behaviors need to be validated by research; and when they are, the best way to get individuals to act in accordance with those healthy behaviors can be elucidated by research.   And we need to think carefully about what we define as a successful outcome.  Employers and health plans tend to be most interested in relatively short-term cost reductions, although employers can think about total productivity impact in a way that plans don’t.  It may be, however, that it takes a number of years of healthy behaviors to cause significant burden of disease and cost declines.  Better health behaviors and better quality of life should be seen as ends in themselves that justify some investment.  But I think the most powerful incentive for people to change is making them bear the consequences of poor behavior.  If you give people support and opportunity to be healthier, and they won’t adopt good behaviors, they should have to pay the costs for the unhealthy behavior.