Concerns about overpayment as a result of favorable risk selection of healthier-than-average beneficiaries have confronted the Medicare program throughout its history of contracting with Medicare Advantage (MA) plans. The goal of the MA risk adjustment system is to use risk scores to assure that MA plans that enroll sicker-than-average beneficiaries are paid appropriately, but not to increase payment for an average beneficiary. A Medicare and Medicaid Research Review (MMRR) paper, titled “Measuring Coding Intensity in the Medicare Advantage Program,” used recent data and improved methodology to estimate the effects of coding intensity on MA risk scores.
MA Risk Adjustment System
The MA payment system creates incentives for MA plans to find and report as many diagnoses as can be supported by the medical record. For example, the MA payment system uses diagnostic information to assign a risk score to each beneficiary, where the average beneficiary in Medicare fee-for-service (FFS) has a risk score of 1.0. MA plans are paid the product of their plan bid multiplied by the enrollee’s risk score. Therefore, if an MA plan bids $1,000/month for an enrollee with a risk score of 1.0, and then enrolls a beneficiary with a risk score of 1.2, the plan gets paid $1,200/month for that enrollee.
The implicit assumption in the design of the MA payment system is that a beneficiary will have the same risk score of 1.0 whether they are enrolled in FFS or MA. As such, if a FFS beneficiary had a risk score of 1.1 if enrolled in MA, then the MA plan would be overpaid. The term “coding intensity” is the difference between the scores that a group of beneficiaries would have if enrolled in MA and their scores in FFS.
The Government Accountability Office (2012) estimated that in 2010 MA beneficiary risk scores were at least 4.8 percent and perhaps as much as 7.1 percent higher than they would have been if the same beneficiaries had been enrolled in FFS.
- According to the MMRR paper, the mean risk scores in MA increased faster than in FFS from 2004 to 2013—by 0.305 in MA, compared to 0.106 in FFS. As measured using the 2004 risk model, MA risk scores grew from 90.2 percent of FFS risk scores in 2004 to 109.1 percent in 2013, an increase of 18.9 percentage points. From 2004–2013, the mean MA risk score increased 2.2 percentage points per year more quickly than the mean FFS score using the 2004 risk model and 1.6 percentage points more quickly using the 2014 risk model. According to the paper, the main reason that MA risk scores increased more quickly than FFS scores appeared to be due to increases in relative coding intensity.
- CMS and the Congress have responded to the increase in risk scores over time in several ways.
- Starting in 2010, CMS lowered payment by 3.41 percent by applying an across-the-board coding adjustment. This coding intensity adjustment will increase to 4.91 percent in 2014 and to at least 5.91 percent in 2018.
- Starting in 2013, CMS set the four most severe diabetes Hierarchical Condition Categories (HCCs) (HCC15-HCC18) to have the same payment coefficient. As a result, recording diagnoses that move enrollees from HCC18 (diabetes with ophthalmologic or unspecified manifestation) into HCC15 (diabetes with renal or peripheral circulatory manifestation) will no longer increase revenue for MA plans.
- CMS made further changes to the model in 2014, removing some of the HCCs that were the subject of MA efforts at increasing coding intensity.
- The legislated increase in coding intensity adjustment is 0.25 percent per year from 2014 through 2018.
- The President’s Budget for 2015 proposes increasing the minimum adjustment to 0.67 percent per year through 2020, topping off at 8.51 percent in 2020, which is much closer to the expected increase in relative risk scores.
The paper concludes that it is challenging to accurately measure the effects of coding intensity on MA risk scores and even more challenging to devise optimal policy responses. It notes that HHS is continuing its analysis of the relative risk of MA and FFS enrollees in order to improve the ability of the Medicare program to accurately pay for MA beneficiaries.