$10,000 Reward to Anyone Who Can Prove that their Pre-Post Methodology Is Even Slightly Valid
I am offering a $10,000 Reward to the first industry trade or professional association, outcomes committee, benefits consulting firm, actuarial firm that "validates" outcomes, US citizen or for that matter undocumented alien with a fake Social Security number who proves that, for population-based pre-post analysis, the DMPC methodology in Why Nobody Believes the Numbers is mathematically invalid and that their methodology is valid. This specifically includes the Care Continuum Alliance’s pre-post methodology, which they themselves have declined to defend even though there is $10,000 in it for them if they do, blaming their own outcomes committee instead for the mistakes in their own publication. (See p. 5 of Why Nobody Believes.)
Switching to the second person here (that’s from “their” to “your”, for those of you keeping score at home), all you need to do is find/invent one set of numbers involving up to 10 real or hypothetical people and their real or hypothetical claims, with up to three periods — a baseline period, a "lookback" period at your option, and a study period — in which the DMPC methodology gives an incorrect answer for the change in claims cost from baseline year to study year, and your methodology yields a correct answer. Example, if claims for everyone who has the condition fall 20% in the hypothetical example between the baseline and study periods, your methodology would have to show a 20% decline and the DMPC methodology would have to show something different.
The people in your hypothetical need not have the condition in every year. They may have it in the baseline year, the lookback period, or the study year, all three, or two of the three.
Let me start you off by giving you the first two strikes against me. First, you can assume that your (provably invalid, of course) non-chronic trend adjustment is perfectly valid, so that you've teased out all the inflation in the three years, which are now comparable to one another.
Second, you don't need to assume that someone who presents with the condition in the study year also had it in the baseline year. Just be clear whether they do or they don't.
So with these two strikes already called against me, all you need to do is throw one more strike to win the $10,000.
You'd actually be doing me a favor since if there is a mistake, it's worth $10,000 to find it.