Subscribe if you want to be notified of new blog posts. You will receive an email confirming your subscription.
CMS: “Rumors of Medicare Health Support’s Death Have Not Been Greatly Exaggerated”
Vince Kuraitis and Thomas Wilson, PhD, DrPH
Today’s POO (persistent obfuscatory orations) Award goes to the Centers for Medicare Services (CMS) for their posting of two bizarre documents updating progress on the Medicare Health Support (MHS) program. The documents “appeared” (no press release, no announcement, a reference to “today” but no date) on the web page for the MHS demonstration project.
While a casual reader would probably review these documents and scratch their head, we interpret them as a death knell for the Medicare Health Support chronic disease management (DM) project and that there will be no MHS Phase II.
As we wrote in our previous essay — Disease Management and the Medicare Health Support (MHS) Project: “Houston, we have a problem” — we caution readers NOT to generalize this as “DM doesn’t work in Medicare”. MHS is only one narrow, outdated business model option of how Medicare could choose to address the challenges of chronic disease; the problems are now understood better than ever, and there a number of other alternatives that can and should be tested.
What Did Medicare Say?
The two documents that appeared on Medicare’s web site are entitled Fact Sheet – Completion of Phase I of Medicare Health Support Program and Completion of Phase I of Medicare Health Support Program FAQs. They make for some confusing reading.
Here’s some key language from the Fact Sheet:
The experience of the MHS program indicates that Phase I of the program is not meeting the statutory requirements of improved clinical quality outcomes, improved beneficiary satisfaction, and the achievement of financial savings targets. Based on the results of the independent evaluation of all of the Phase I programs, CMS will consider whether to expand to Phase II.
One of the goals of the MHS program is to see a reduction in Medicare claims costs for the intervention group. To date there has been nominal impact on Medicare claims costs as a result of the MHS program. At the level of individual organizations, some MHS programs show increases while others show modest decreases in Medicare claims costs. However, to achieve budget neutrality, the five MHSOs in current operation need to reduce Medicare claims costs by between $300 and $800 per participant per month for the remaining months of the pilot program. This represents a 20 to 40 percent reduction in claims costs from the current levels that are being billed.
…and from the FAQ document:
The statute limits Phase I to three years and makes expansion to Phase II contingent on a program (or components of such program) improving the clinical quality of care, improving beneficiary satisfaction, and achieving targets for savings. Unless all three of those conditions are met, there is no authority to begin Phase II.
What Does Medicare Mean?
Here’s our interpretation of what CMS is saying:
MHS is not meeting targets for financial savings. While it is theoretically possible that the MHS program could climb out of the hole financially during the remaining months of the program, we are doubtful that this will happen — so much so that we are scheduling the patient’s (MHS’) funeral even though technically we are not allowed to pronounce the patient dead yet. In the event that hell freezes over and the program revives, we will then schedule Phase II, but don’t hold your breath.
In a new twist on an old saying by Mark Twain: The rumors of MHS’ death have NOT been greatly exaggerated.
This work is licensed under a Creative Commons Attribution-Share Alike 3.0 Unported License. Feel free to republish this post with attribution.
I have heard that CMS is basing these decisions on new data that they have on the programs. Have you heard of new data being released?
Also, what are your thoughts on why they would grant budget neutrality and then decide to quietly kill the program?
I believe they are protecting the industry from significant financial harm by granting budget neutrality. The contracts are still valid from Phase I so even with no Phase II, the original fee guarantees would strangle some of these companies if they were held to them….
This is sad news.
Is it possible that these pilots were adversely affected by not adequately taking into account the problem of comorbid mental disorders, which have been shown to increase health costs and are associated with poorer adherence to treatment and poorer clinical outcomes in several chronic conditions?
I am very relieved to see that vendor lobbying will not defeat scientific evidence.
I am very relieved to see that vendor lobbying will not defeat scientific evidence. In response to Dr. Locke – the failure to demonstrate cost savings is not related to co-morbid conditions – mental or any other. It’s a randomized controlled trial, and the intervention and control groups are comparable.
Ah, but what if the intervention had included screening for depression and supported self-management and/or behavioral health referral for depression in the 20-25% of cases where it may have likely been comorbid (integrated into the intervention). That is my point.
A commentator says with authority and sincerity that this was a “randomized clinical trial” — thus we should accept the results as flawless and irrefutable. However, the original clinical trials and resulting “evidence” upon which the MSP measures are based, were never tested with at risk subpopulations (women, elderly, racial and ethnic groups, disabled, those with multiple co-morbid conditions and co-occuring disorders, including mental health issues). Nor are most trials followed by even quasi-experimental studies in actual care setting treatments involving broadly representative patients — to see what worked and what didn’t, for whom, why and when. Therefore, why would this “demonstration” based upon weak science yield results that are meaningful for all populations since teh metrics were never intended for use with all populations — yet they are being applied to all. Going into a major demonstration like this knowing little about unique patient characterstics and unique treatment responses — the results are more likely no more than junk science. Just because there are so-called intervention and control groups doesn’t mean the criterion are sufficiently sensitive or reliable or the least bit related to the actual experiences of individual patients or that it is even a well-structured clinical trial. Quality experts worked hard to make MS work — including CMS. But a problem today is that “quality” has become a dirty word. It’s the new code word for “how do we justify measures that save money and eliminate/ration care options” (sorry — it’s called reducing waste and variation — assumed to be bad). Waste to whom? Enormous effort has gone into making the MS demonstration work but it was handicapped at the start by numerous ill-conceived assumptions about what we know about quality “patient” care. It turns out patients are actually all quite different and don’t often fit easily into rigid protocols/measures.
In response to CJ:
A randomised study is the gold standard for producing a fair comparison of the intervention with the control.
I think, though, that CJ is raising a couple of distinct points not directly related to the main study design–(1) the measures used to study program effects may not have been the best, (2) the results may not be representative of people who didn’t participate in the pilot.
(1) It could well be the measures used in the pilot were insensitive to important changes in participants. Not being familiar with the measures used in the pilot, I’d like to know what measures would have been better? Also CJ seems to be saying that there isn’t good evidence for treating “women, elderly, racial and ethnic groups, disabled, those with multiple co-morbid conditions and co-occuring disorders, including mental health issues” because previous clinical trials haven’t included such people. So how should we care for these populations? Is the quality of their care fine as is? If not, how do we know that, and is there a way to measure the deficiencies?
(2) It’s true that results from randomised studies may not be easily generalisable to broader populations, but my impression of MHS was that it was population based–they included most people living in the pilot communities who had the target diagnoses.
I agree with CJ that patients are different. I also agree that standard protocol measures do not come close to capturing or meaningfully quanitifying these differences. But, if something works EVIDENCE of it will show up somewhere. Do you really think that if there was a $300 or $500 savings per patient it would not be detected? The obvious reason for why the CMS Phase 1 did not show an impact is because there wasn’t any. The Emperor has no clothes.
While we’re on the subject of transparency, I, for one, am confounded with the current practice of CMS and HHS to release major announcements on a late Friday before a long weekend. This strategy may satisfy the call for transparency in theory, but in practice appears to really be a subterfuge tactic that delays dissemination of results. What do we call this type of transparency? Obfuscated transparency?