Insufficient Evidence to End Medicare Health Support

Thomas Wilson, PhD, DrPH and Vince Kuraitis

Last Tuesday during the cocktail hour, CMS issued documents portending the end of the Medicare Health Support (MHS) project.   

We initially used the word “bizarre” to describe the announcement from CMS’ staff.  After further thought, “reckless” and/or “insubordinate” strike closer to home.

First, let’s have the head-honcho at CMS instruct us on how things should be done. In September 2007 Kerry Weems, the new CMS Administrator, declared that “cocktail hour press releases” from his agency must stop, and that a new era of “transparency” and “accountability” must begin.  

So here’s the list of decision criteria we draw from Mr. Weems statement:

1) No cocktail hour press releases

2) Transparency

3) Accountability

Now let’s see how CMS’ announcement ending the MHS program does on these criteria:

1) No cocktail hour press releases

Strike one.  CMS’ announcement to end MHS just “appeared” late in the day on Tuesday. It was buried deep in its website, never mind even the courtesy of a press release.  They apparently informed hardly anybody.

2) Transparency

Bear with us while we get a bit technical. The first (and only) report on the MHS program was issued last summer, but it was only based on 6 months of program data.   Yes, there were problems as we have pointed out

We were all eagerly awaiting more ‘evidence-based’ reports, but thus far none have been made public.  Instead, as part of its announcement to end MHS, CMS issued a list of Frequently Asked Questions.  In this FAQ list we find three statements that provoked our comments and questions:

a) There is “sufficient sample size to conduct the financial analyses for the evaluation.”

  • So, by what method was this determined?  Was sample size sufficient for clinical measures as well? Was sample size sufficient for sub-component analysis? 
  • Where is the transparency that Kerry Weems calls for?

b) “The financial reconciliation does not rely on statistical testing, so sample size is not an issue for that task”

  • Oh, so sample size is not important, (please see statement above where it was important)?  Does that mean that financial measures are immune from random fluctuations? 
  • Where is the transparency that Kerry Weems calls for?

c) Regarding the non-equivalence between the DM intervention population and the reference population: CMS has offered “an actuarial adjustment … [that] statistically adjusts the historical financial disparities between the intervention and comparison population, essentially making the adjusted costs of the intervention and comparison groups equal as of the program start date.”

  • So, the mandated randomization did not work, and statistical adjustments were made to make the two groups equivalent.  In the spirit of transparency, we need to see the method, in its entirety.  As Ronald Reagan famously said: “Trust, but verify!”  With this statistical adjustment do we still have a rigorously designed evaluation?
  • Where is the transparency that Kerry Weems calls for?

3) Accountability

CMS’ reckless announcement could have severe negative consequences for patients, the companies involved, and DM community as a whole.

For example, as pointed out in Wednesday’s blog posting, CMS’ actions resulted in vaporizing $389 million of stock value for Healthways investors.  Investors were surprised — something that should not happen in a program where a public, government funded agency does business with a publicly traded company.

So now what?  With the understanding that further examination might show some fault on both sides, wouldn’t a “we’re sorry” or at least an “oops” from CMS be in order?  or something more?  At least it might avoid an SEC investigation.


What can we learn from this very expensive experiment to improve the health of our seniors?  CMS needs to remove the clouds, follow the rules, and let the sun shine in.

The DM community is familiar with the 2004 Congressional Budget Office report noting: “there is insufficient evidence to conclude that disease management programs can generally reduce the overall cost of health care services.”

This time the tables are turned:  CMS did not show sufficient evidence to justify its actions in ending the Medicare Health Support program.