Workplace wellness - sense, cents, or nonsense?

Workplace wellness and chronic disease management programs (now encouraged by the ACA) are quite common, but being a $6 billion dollar industry doesn't make them useful. Recent personal experience with one such program has led me to look more closely at the topic.

A skeptic by nature, I try not to believe anything without checking the evidence and then thinking it through. Science, after all, operates based on disconfirmation; scientists do not prove a theory true, they only demonstrate that it explains what we see, and cannot be disproven. The less plausible something is, the more I want solid evidence before believing it. To my mind, workplace wellness programs are an area where plausibility is a big issue. Human behavior is very hard to change - ask any physician or mental health worker. Achieving improved employee health and substantial reductions in health care costs through either wellness programs or chronic disease management may be appealing on an intuitive level, but is without much supporting evidence.

When an institution pursues workplace wellness with a company like Edumedics, one wonders if the decision makers have training or experience in the fields of disease management, preventive care, biostatistics or epidemiology.  This is a fertile ground for the Dunning-Kruger effect ( discussed in this YANSS podcast) where people do not know enough to be able to recognize their incompetence.

The 2013 RAND Wellness Study, which Edumedics inaccurately quotes on its website, identifies and discusses two common types of workplace wellness initiatives:

  • Programs that focus on chronic disease management in patients with identified illnesses like DM, asthma, CAD, hypertension and COPD. These typically use billing data to identify high cost employees or employee groups. These programs are based on current costs as seen in billing data, not the risk of future costs. Properly done and narrowly targeted, these may save employers money, though the amounts are generally small. There is no evidence that they save money for employees.  In fact, a recent study suggests that the cost savings for employers are at least partly a result of shifting costs to the most vulnerable (most ill) employees (Wellness Incentives In The Workplace: Cost Savings Through Cost Shifting To Unhealthy Workers, by Jill R. Horwitz, Brenna D. Kelly, and John E. DiNardo)
  • Programs that focus on wellness and lifestyle, often with biometric screening to identify employees at risk. RAND notes that these programs are, by definition, broadly focused. They are also expensive, yield little if any cost savings, and result in insignificant health benefits. This is consistent with the consensus of other published studies, including a detailed review by the University of California and a RAND study of Pepsico. This is not surprising: only a small number of preventive services have been shown to improve health, and it takes a long time, on the order of a decade or more, for even the best preventive services to yield a clinical benefit in a population.

The 2013 RAND study and other analyses note that most of the published studies are done by the wellness industry and are of poor quality. They typically claim health or financial benefits based on lower costs and outcomes in participants, as compared to costs incurred by non-participants. This is what Edumedics claims for diabetes and hypertension on its website. This methodology is invalid because of serious selection bias. Unless employees are randomly assigned to participation or to a non-participating control, no conclusions can be drawn about differences in outcomes. That Edumedics would make unsupportable and misleading claims like this on its website does not inspire trust. If institutional leaders repeat these statistics when describing their wellness program to employees, it does not inspire confidence.

 

More bothersome to me, the Edumedics website also claims that: "A multi-year study of nearly 600,000 employees by the nonprofit research organization RAND Corporation found significant cost savings and health benefits from company wellness programs that target employees with chronic diseases." What the referenced RAND study actually said was:

“Our statistical analyses suggest that participation in a wellness program over five years is associated with a trend toward lower health care costs and decreasing health care use. We estimate the average annual difference to be $157, but the change is not statistically significant.”

Here is a summary of some other pertinent key findings from the 2013 RAND study:

  • Participation is low: fewer than half of employees elect to undergo screening or risk assessments, and only 20% of those found eligible for disease management participate in the program. (This, of course, results in tremendous selection bias when the costs and outcomes of participants is compared to that of non-participants, which is what Edumedics does on its site.)
  • Participants lost an average of only 1 pound a year for 3 years.
  • No significant reductions in total cholesterol levels.
  • Some evidence of successful smoking cessation – but only in the short term.
  • Statistically insignificant cost savings of $2.38/mo (year 1) and $3.46 (year 5).
  • Statistically insignificant reduction in cost or use of emergency department or hospital care.
  • Screening all employees for health risks and offering one-to-one counseling and coaching to those with such risks is expensive and not associated with significant benefits.

Not only does the Edumedics website misrepresent the findings of the 2013 RAND study, but the quote on the Edumedics site does not come from the referenced RAND study, and their link, phrased to appear that it links to the study, links instead to a press release by RAND that notes:

"The recently published RAND Wellness Programs Study, which included almost 600,000 employees at seven employers, showed that wellness programs are having little if any immediate effects on the amount employers spend on health care. This has been further confirmed by our new analysis of 10 years of data from a Fortune 100 employer."

This press release goes on to note that this separate multi-year analysis of a single Fortune 100 company (Pepsico with 60,000 participants) showed that the 13% of who participated in the chronic disease management program generated a $3.78 ROI on the money spent on that aspect of the program. The Pepsico program targeted ten (not three) chronic illnesses and included case management and a 24 hour nurse advice line. Healthcare costs tracked for the study were limited to hospitalizations and emergency visits, and did not include any potential higher costs from more intense management of the targeted conditions. The investment cost was limited to the vendor fees for operating the program. Most of the savings came from a 29% reduction in hospitalizations. The $360 annual cost reduction is small (3.6%) in the context of their annual health care cost of over $10,000 per participant. The $3.78 figure is cherry-picked  and misleadingly applied by Edumedics.

Perhaps most telling is a carefully done review published this past summer. It examined 51 studies in nine industry types in 12 nations with more than 260,000 participants. “We found that as methodological quality improved, return on investment decreased, and we found a negative ROI in randomized control trials.” In the randomized controlled trials, ROIs for the interventions studied had an overall mean value of -0.22. For every dollar spent, only 78 cents was gained back. (Siyan Baxter, Kristy Sanderson, Alison J. Venn, C. Leigh Blizzard, and Andrew J. Palmer (2014) The Relationship Between Return on Investment and Quality of Study Methodology in Workplace Health Promotion Programs. American Journal of Health Promotion: July/August 2014, Vol. 28, No. 6, pp. 347-363.)

(As an aside, it will be interesting to see if there is any impact on the popularity of workplace wellness programs as a result of the recent lawsuit  filed by the EEOC against Honeywell because their wellness program penalizes employees who do not undergo biometric screening. The EEOC claims that Honeywell’s incentives violate the ADA: employees are penalized in order to induce them to go through medical examinations that are not job-related or consistent with business necessity, and undergo exams that are not voluntary because Honeywell imposes a penalty on employees who decline to participate.)

My take:

  • Employee health and wellness are laudable goals.
  • Pursuing employee wellness is the right thing to do.
  • The evidence supporting broad based risk assessments and wellness programs based on lifestyle change is absent.
  • The use of penalties to coerce employees and their covered families to participate in programs that are ineffective  is inappropriate.
  • The evidence supporting disease management is conflicted at best. The available data suggests the best case scenario is achieving relatively small benefits in large corporations with heavily resourced and narrowly focused programs that are expertly done. It is not clear this can be generalized to most institutions.

 


 

Links to more on this topic::