More on Penn State's Wellness Woes and The Evolving Science of Evaluating Health Promotion Program Outcomes: There Is No Gold St
Posted Aug 19 2013 9:56pm
Aside from keeping up to date with work buddies, the Disease Management Care Blog doesn't really use LinkedIn all that much. But when "Support our Penn State Colleagues in their Fight to Prevent Coercive Junk-Science Wellness Programs" postings began to appear in a Discussions board, the DMCB couldn't resist. It rose in support of its alma mater (College of Medicine, '77) faster than a med-mal attorney can calculate a contingency fee.
As noted in this prior DMCB posting , Penn State University launched a rather routine health promotion program that prompted some nasty and very public teaching faculty resentment. Calls for "civil disobedience" and sinister references to "eugenics" made the DMCB wonder how much of the reported push-back was mainstream employee opinion vs. mainstream media's biased reporting. That distinction didn't stop the LinkedIn board from running a mostly one-sided dialogue on the matter.
So, undeterred by the unfairness of so many vs. just one, the contrarian DMCB naturally jumped right in. Among the issues raised The RAND Study on wellness casts doubts on the merits of employer sponsored wellness programs:
Actually, RAND found employer-sponsored programs lead to statistically significant increases in exercise levels as well as reductions in tobacco abuse and body weight. To the disappointment of wellness vendors everywhere, however, these programs did not lead to statistically significant reductions in health insurance claims expense. The ever-optimistic DMCB points out that that means that these health improvements occurred without an increase in health care costs. While cost neutrality alone is good news, the DMCB also believes that an emerging generation of wellness programs will do a far better job of identifying persons with 1) actionable risk and 2) who are willing to take action. By husbanding wellness resources for subpopulations where it will have the greatest "bang," program costs will go down and claims savings will achieve statistical significance.
Actually, the original Health Affairs paper said that the finding of a $3.27 return on every dollar spent is subject to "(f)urther study.... to elucidate the time path of return on investment.... The assumption of a linear trend in savings from the beginning to the end of program evaluation may not reflect the reality of behavior change within organizations."
The point is that nuanced and calibrated conservatism is typical of excellent peer-reviewed research and, taken in context, the authors are being quite consistent in-print and on-air. Academics will always say more research is needed. Skeptics will over read that.
There are powerful arguments against the common wisdom that "wellness saves money," suggesting that the health promotion industry has been intentionally ripping employers off.
Actually, when it comes to wellness outcomes, there is no agreement on "the" measurement " gold standard ." Without any consensus on which assessment approach (for e.g., this vs. this ) is truly "better," only one thing is certain: much like the Betamax vs. VHS wars, the future owner of "the" standard stands to reap a consultant's bonanza. Until we declare a winner, assessing the truth will be a messy mix of triangulating on means, medians, confidence intervals, imperfect reference controls, suspect generalizability, human judgment, moving targets and evolving interventions.
What about [insert name of wellness program here] that is an obvious sham?
There have been women who have had mammograms with missed cancer, victims of car crashes who have died despite seat belts and times when the DMCB did something really dumb despite the advice of the DMCB spouse. That doesn't mean mammogram, seat belts or advice are worthless. The plural of anecdotes is not data .
Coda: By the way, the statistically significant " value of 0.05 " is more of a consensus than a gold standard. Why is a 5% chance that an observed result is not the result of randomness wiser than a 6% chance or a 4% chance?