In health management (HM) whether applied to Medicare/Medicaid, commercially insured, or employee populations, there seem to be widely varying opinions as to how “successful” they are. Disease management (DM) seems to get consistently equivocal or unenthusiastic reviews when government examines its cost savings success. Meanwhile, employers are jumping on the employee health management (EHM) bandwagon in increasing numbers.

Part of the problem is necessarily the differences in both the definitions and measurement of success under different circumstances. CMS naturally looks for DM cost savings first, with quality measures added to be sure that such savings are not being achieved at the expense of patient care. Commercial insurers look for the same, and often focus on DM as well, since they may not retain members long enough to enjoy the payoff from health risk reductions that take years to return measurable savings.

It is employers who have the easiest time, in one sense, and the hardest in another. They have the greatest potential, by far, for cost savings, since they can save not merely through reduced healthcare costs, the only source of savings for CMS and private insurers, but through labor cost savings as well. Reductions in absenteeism, presenteeism, and turnover have usually been found in multiples of healthcare cost savings, when they are measured. Moreover, employers who invest in EHM tend to keep their employees long enough to gain significant savings from their investments.

GlaxoSmithKline, for example, which enjoys an employee turnover rate below 10% per year, followed the same cohort of over 6000 employees who participated in its “Wellness Contract” program for four years. While savings were modest in the first year, at $233 per participant, they rose significantly to $375 in the second, then “exploded” to $944 and $950 in the third and fourth years. [G. Stave, et al. “Quantifiable Impact of the Contract for Health and Wellness” JOEM 45:2 2003 109-117]

But success for any HM investor depends on the numbers of people in the population at risk who are:

• eligible for an HM intervention, and they must participate in a health risk assessment of some kind to determine how many and who among them are eligible for what
• willing to enroll and participate effectively and persistently in the HM intervention, whereas the majority drop out in most cases
• able to succeed in making the health behavior and health status changes needed to realize savings or perhaps revenue gains for their sponsor (if an employer)

While it is common practice for HM providers to report the success amounts they achieve, on a per member of the population, or perhaps per participant basis, it is rare for them to divulge how many participants actually succeed to the point of delivering economic benefits for their sponsor, and personal health/life quality benefits for themselves. And this “personal” success is key to both retaining those who are participating in HM interventions, and to attracting new participants thanks to peer communications.

I have found at least one referred journal reference on the subject of the “success rate” in HM programs, though it discusses DM only. It states that there are no scientific studies that reveal even the average success rate, but the common practice is to use a 50% rate, as is true for participation rates, apparently. It did report that roughly 50% of eligibles in a worksite clinic-based DM program were successfully contacted, and roughly 50% of them enrolled in the program, but gave no figures on how many succeeded, since the focus of the article was on showing how much more successful worksite clinics can be in recruiting participants, thanks to eligibles having a trust relationship with the clinicians practicing there. [S. Frazee, et al. “Leveraging the Trusted Clinician” Disease Management 10:1 2007 16-29]

If there is indeed a “convention” that success rates are roughly 50%, it may be based on no more than the Bayesian logic that if we know the upper and lower limits of a number, and a success rate can be no lower then 0%, and no higher than 100%, the “best guess” is always the midpoint of the two, assuming no other information. This sounds like the even older convention of “splitting the difference” when numbers are in contention, but it has the advantage of automatically reducing the possible error as much as it can be reduced in the circumstances.

There have been a few EHM providers that have described their success rates, at least. Marathon Health, in Colchester, Vermont, has reported that roughly 62% of employees at one client completed the biometric screening and health assessment needed to identify eligibles; 90% participating in an appropriate HM intervention, and 50% of those succeeding in making significant improvements. (personal e-mail from David Demers at Marathon on Oct 2, 2007).

Health Media, Inc. of Ann Arbor, Michigan has reported “efficacy rates” which are the equivalent of success rates for two of its EHM programs. The rate with smoking cessation, where “success” = quitting, has been 18% on average across the over 200,000 employees it has tracked. The success rate for stress management, meaning the percent who reported reduced stress when their health and productivity were assessed again, was 62%. [“HealthMedia, Inc. Hosts Webinar on Medication Adherence” HealthMedia.com Aug 13, 2007]

Success rates, in practice are bound to vary widely, even if they can only range between 0% and 100%. The characteristics of the population involved, the difficulty of the changes participants must make in order to succeed, the presence or absence and size of success incentives offered, the support of their peers, and the methodology used in the HM intervention all vary widely, so that success rates almost have to, as well.

But since success is both clearly measurable, in most cases for most economic benefit dimensions to sponsors, at least, and clearly valuable, it would seem wise to devote some added effort to measuring and reporting rates achieved by different HM providers and programs. If we are to have anything resembling “evidence-based medicine” in the HM arena, we need to have “transparency” for success rates, and success amounts as well.

If it became routine for HM providers to report their success rates, costs per success and economic benefit per success, along with the numbers eligible for and participating in their particular programs, we would at least have the foundation for the kind of “customer”-oriented responsibility and capability we are shooting for in “value-based purchasing” of sickness care.