Monday, September 20, 2010

Healthways Disinters Medicare Health Support

Until now, the DMCB assumed that the hapless Medicare Health Support (MHS) pilot had not only died, but been embalmed, sealed in its casket and buried under six feet of lets-not-speak-of-it-again. But lo, Healthways has pulled out its shovel and has propped up the remains at our national health reform banquet. Come... see, say a host of elder wise-men Tom Daschle, Newt Gingrich and Ken Thorpe, it lives, it... LIVES!

As readers may recall, MHS was the unsuccessful, prospective, randomized and controlled trial of disease management vs. usual care for FFS Medicare beneficiaries. There were several vendors involved, including Healthways. The financial terms for the intervention group involved a monthly fee for disease management in exchange for the expectation of reduced claims expense with budget neutrality or savings in excess of fees. Unfortunately, the study was hobbled by a number of execution problems that included the unequal assignment of beneficiaries to the control and intervention groups, delays in notifying the vendors about hospitalizations and low patient contact rates. Combined with the disease management industry's irrational exuberance, the final negative did-not-save-money results were bitterly disappointing. You can read the disillusioned DMCB's past thoughts on the final results here and here.

What do you know, Healthways decided to take another look at its MHS data. Recall that the MHS report ultimately concluded that, compared to the control group, health care costs increased less (i.e., didn't go up as much) as a result of disease management, but that the savings, versus its average monthly fee of $94, were not budget neutral (go to page 67).

As the Disease Management Care Blog understands it, third party research through Emory University, Harvard University, The University of Milan and The University of Trieste applied a "Coarsened Exact Matching" to achieve a better statistical balance between the intervention and control groups. Based on this analysis, a recalculated mean savings per beneficiary totalled $2,848 per year, resulting in a net cost savings of $73 per beneficiary.

As noted in this news report, the Feds are unlikely to be impressed and are still likely to get a clawback of some of their fees. Given the lack of any movement in Healthway's stock price (HWAY), the market isn't impressed either.

Yet, should the DMCB and its growing readership be impressed? The DMCB isn't sure:

1) It's challenging to evaluate Healthways' claims based on a press release that provides only summary results. Given the stakes, the claim that savings were ultimately achieved warrants greater detail, more transparency and the input of peer review. Sure, it takes time and effort, but, in the long run, it's worth it.

2) The key feature of this statistical breakthrough is the use of "coarsened extract matching." The DMCB never heard of that before. That same problem is probably getting in the way of other interested stakeholders.

3) And then there's a credibility issue: the uphill part of any post-hoc analysis - even if third party and even if it is Harvard - is overcoming the perception that the methodology used was chosen precisely because of the likelihood of yielding a desired result.

No wonder why few seem impressed. Yet, the DMCB isn't sure reburial of the remains is necessary at this time. It'll look forward to seeing more data and analysis. Sort of like poking it with a fork.... to see if it... moves.

No comments: