Medicare Reform Gone Astray

by Douglas Holtz-Eakin

There are some good ideas that even government can’t botch. Unfortunately, the Medicare Advantage (MA) Stars program isn’t one of them. The original idea was simple: take the existing system of a 1- to 5-star rating system and transform it into a “pay-for-performance” system that would generate higher rewards for better health care. Reasonable idea. The actual implementation is anything but reasonable, resulting in a broken system where government is left picking winners and losers with seniors’ health care.

There is a straightforward logic to creating such a system: (a) set up a schedule of rewards as incentives, (b) treat the MA beneficiaries, (c) check outcomes for the MA plan beneficiaries, and (d) send bonuses to good performing MA plans.

What could go wrong? Well, as implemented by the Center for Medicare & Medicaid Services (CMS) the Star system actually: (1) treats patient, then (2) sets up incentives, (3) checks outcomes for patients that may or may not be MA plan beneficiaries, and (4) sends bonuses to MA plans. 

There is nothing wrong with steps (1) and (2) that could not be solved by time travel. But in the absence of that, it is impossible for health plans to adapt to achieve better performance and reach said incentives. For example, CMS published criteria in October 2011 to be applied to plan performance between January 2010 and June 2011. There is simply no way an MA plan can adjust its plan offerings, plan design, or performance in any way to increase its performance in the dimensions CMS chooses to evaluate. 

#more#Weirdly, the time period chosen is also not the same for every criterion. For example, the October 2011 criteria included cancer and cholesterol screenings for calendar year 2010, but flu vaccination for February through June 2011 (notice this excludes the peak flu vaccination season in the fall). In all, 36 different measures were computed using six separate, occasionally overlapping, time periods — all of which ended more than three months before the criteria were published.

The only thing a seniors’ health plan can do is engineer to fit the retroactive incentive criterion, thereby raising the suspicion that they are being led to fulfill some 2008 campaign promise.

Steps (3) and (4) are equally problematic. MA Star bonuses paid in 2013 will be based on the performance in 2010 and 2011, but the enrollment is for a plan in 2013. A plan that qualifies for a high star rating based on 2010 performance could adjust its bids, service areas, and marketing strategy for the 2013 plan year in order to increase enrollment in counties with the highest bonuses and rebates.

This is a recipe for conflict because there is an opportunity for CMS to use the Star system as a budget tool to starve Medicare Advantage. Indeed, there is some indication this is already happening. For example, the average rate of “all-cause readmission” within 30 days is 19.6 percent for traditional Medicare and a significantly better 15 percent for MA patients. Despite this, CMS issued criteria in October 2011 requiring this rate to be below 5 percent to get to the 5-star level. The only problem: No MA plan made this level and CMS knew this when it set the criteria. 

No plans meeting the criteria means no 5-star bonuses. No 5-star bonuses means one more way to starve Medicare Advantage.

So long as the Affordable Care Act remains the law of the land, the Stars program will affect Medicare patients. It is an important example of a broken product with potential that can and should be fixed.

Pay for performance requires letting performance dictate pay. Pay for performance means paying; not saving. Let’s design a system that stops government from picking political winners with seniors’ health care. 

The Corner

The one and only.