Wellesley, MA (December 18) The Disease Management Purchasing Consortium (DMPC) announced today the winners of the Disease Management (DM) Intelligent Design Awards, given annually to those contributions which most set back evolution of the disease management and wellness fields. Just as engineers say that more is learned from a single bridge which collapses than from 100 which stay up, there are serious lessons to be learned from these often-humorous failures.

First place is awarded to Trestletree, for the claim on its website that it “reduced absenteeism by 300%.” The winning screenshot (it is fifth in succession on the rotating www.trestletree.com home page as of this writing) is also available here from DMPC.

UPDATE: Trestletree has now changed its website to reflect use of wellness industry-standard measurement metrics. If this award were bestowed today, they would not qualify for it. We wish them the best of luck in their efforts now that their outcomes are not tainted by math error. Fortunately for DMPC, though, there is no shortage of candidates to take their place at the top of this list. Watch this space.

The lesson: This is perhaps the most glaring example of the DM/wellness field’s rampant innumeracy – its willingness to accept results which are either mathematically or epidemiologically impossible or which, within the report, contain self-contradictions. DMPC is responding to this by offering Certification for Outcomes Reporting/Analysis. Details will be following very early in 2007.

Second place goes to the well-publicized failure of the Mississippi Medicaid RFP, an RFP that the state was urged unsuccessfully by multiple experts to withdraw and rewrite early on. Most of any Medicaid RFP is boilerplate, so the analysis will begin with Section 3.3.7, which states: “[There is a] guaranteed...ROI of 1.05, or 100% of the program cost...plus at least a 5% net savings through reduced medical costs.”

The problem, as any disease management professional can quickly ascertain, is that this clause is not just ambiguous but rather self-contradictory, as when Casey Stengel told his Mets players to “line up alphabetically by height.”

Is the ROI 1.05, as the first part of the sentence says (and is a very easy target to hit) or is it 5% (500 basis points) above the cost of the program, which is equivalent to roughly a 2x ROI, which the second half of the sentence implies?  The difference between a 1.05x and a 2x guaranteed ROI in a $10,000,000/year program is roughly $9,500,000.

Obviously these two ROI goals are very different. One can be easily achieved while the other is much harder. In the bidders’ conference, the state declined to answer the question, instead referring bidders back to the sentence itself.

While also another example of innumeracy, the specific lesson here is that states ought to read other state Medicaid RFPs before sending out their own. Georgia, Illinois and Wyoming are the three which have won awards from the Health Industries Research Co. All state Medicaid RFPs are in the public domain. A little research into what works and what doesn’t hugely facilitates the RFP process and prevents the embarrassment of a failed RFP. The DMPC report “Lessons Learned from Failed Medicaid DM RFPs” is available free to any state which requests it as well as any DMPC member.

The honorable mention goes to Indiana Medicaid’s internal disease management program. They produced an alleged 12:1 ROI in heart failure, projecting to $39,000,000 in annual savings. Unfortunately, this number would be difficult to achieve because this is more than the state spends in heart failure hospitalization for the disabled population and more than Georgia, a somewhat larger state which procured savings very effectively, is guaranteed to save in all conditions combined. There is also a simple but critical basic arithmetic mistake in their CHF outcomes. The slides are available from DMPC, and also (as of this writing) from Indiana.

The reason it is only an honorable mention is that they have now withdrawn these slides claiming any outcomes. Instead in December 2006 they presented their 180-person pilot ending in September 2005 which they say they “may have had some positive effect” and they will be studying for “several years,” while most other states, which are outsourcing, count their savings and monitor their improved outcomes.

Besides being another example of innumeracy, there are two other lessons. First, apply a “plausibility check” before presenting numbers. A very high savings estimate should be checked against total relevant spending, for example. Second, states should outsource their disease management. A state is not a business and, as Indiana’s experience illustrates, is not set up for large-scale (or even small-scale) DM programs implementations and sophisticated (or even simple) program measurement. Also, it is also harder to evaluate a 180-person study than it is to evaluate the entire disabled population, where a bend in the trend can be readily ascertained.

About the Disease Management Purchasing Consortium

The Disease Management Purchasing Consortium Int'l, Inc. (DMPC) provides strategy and procurement services to support the disease management efforts of six dozen health plans and eleven states. The DMPC is led by Al Lewis, widely accepted as the founder and most influential leader in the disease management industry. Associate members include national health care accreditation organizations such as JCAHO and URAC, the Harvard School of Public Health, Bain & Co., Boston Consulting Group and McKinsey, and eight investment banks. The DMPC specializes in calculating savings guarantees from DM programs using the field's only valid counting metrics and obtaining "most favored nations pricing" for DM purchasers. Programs and services are available at www.dismgmt.com.

Next Intelligent Design Award