In fiscal year 2010, the federal government spent about $113 billion on the Medicare Advantage program, a private plan alternative to the original Medicare program that covers about a quarter of Medicare beneficiaries. Medicare Advantage plans are paid a fixed monthly amount to provide beneficiaries with the same services as traditional Medicare. Most of these plans receive larger payments than would be required to provide traditional Medicare services. This allows them to provide additional services not covered by traditional Medicare.
The Centers for Medicare & Medicaid Services (CMS), the agency that administers Medicare, adjusts payments to Medicare Advantage plans based on the health status of each plans enrollees. This adjustment is intended to provide higher payments for sicker patients and lower payments for those who are less sick. CMS calculates a risk scorewhich is a relative measure of health statusfor every beneficiary. The risk score is based on a beneficiarys demographic characteristics, such as age and gender, and major medical conditions. To obtain information on the medical conditions of beneficiaries in traditional Medicare, CMS generally analyzes diagnosesnumerically coded by providers into Medicare defined categorieson the claims that providers submit for payment. For beneficiaries enrolled in Medicare Advantage plans, which do not submit claims, CMS requires plans to submit diagnostic codes for each beneficiary. Analysis has shown that risk scores are higher for Medicare Advantage beneficiaries than for beneficiaries in traditional Medicare with the same characteristics, and CMS has taken steps to reduce Medicare Advantage payments, saving $2.7 billion in 2010.
Risk scores for beneficiaries with the same demographic characteristics and health conditions should be identical, regardless of whether the beneficiaries are in a Medicare Advantage plan or traditional Medicare. This will be true if Medicare Advantage and traditional providers code medical diagnoses with the same level of reliability and completeness. However, Medicare Advantage plans and providers in traditional Medicare may code diagnoses differently. Medicare Advantage plans have a financial incentive to ensure that all relevant diagnoses are coded, as this can increase beneficiaries risk scores and, ultimately, payments to the plans. Many traditional Medicare providers are paid for services rendered, and providers have less incentive to code all relevant diagnoses. If Medicare Advantage risk scores are higher than traditional Medicare scores for beneficiaries with the same demographic characteristics and medical conditions simply because Medicare Advantage diagnostic coding is more comprehensive, then CMSs payment adjustment will not be accurate and Medicare Advantage payments will be too high.
Policymakers have expressed concern that risk scores for Medicare Advantage beneficiaries have grown at a faster rate than those for traditional Medicare beneficiaries, and some believe that systematic differences in coding diagnoses have contributed to this growth. The Deficit Reduction Act of 2005 directed CMS to conduct an analysis and adjust risk scores for differences in coding practices, to the extent that such differences could be identified in 2008, 2009, and 2010. The Health Care and Education Reconciliation Act of 2010 directed CMS to continue adjusting risk scores until the agency implements risk adjustment using Medicare Advantage data. In 2010, CMS estimated that 3.41 percent of Medicare Advantage risk scores were due to differences in diagnostic coding practices, and it reduced the scores by 3.41 percent, thereby saving $2.7 billion.
As GAO reported in January 2012, three major shortcomings exist in CMSs method for adjusting Medicare Advantage payments to reflect differences in diagnostic coding practices between Medicare Advantage and traditional Medicare. A revised methodology that addressed these shortcomings could have saved Medicare between $1.2 billion and $3.1 billion in 2010 in addition to the $2.7 billion in savings that CMSs 3.41 percent adjustment produceda total savings of between $3.9 billion and $5.8 billion. GAO expects savings in 2011 and future years will be greater. However, CMS has continued to use, or plans to use, its 2010 adjustment of 3.41 percent in 2011 and 2012.
First, CMS did not use the most recent data for its estimates. For 2010, the agency did not incorporate in its estimates 2008 data, the most recent data available. Similarly, the agency did not incorporate 2009 and 2010 data as it became available to update its estimates for 2011 and 2012. The most recent risk score data used by CMS in any of these estimates was 2007.
Second, CMS assumed that the annual impact of coding differences remained constant relative to coding differences from 2004 to 2007, despite evidence that the impact was increasing over time. Although CMSs 2010 estimate accounted for the cumulative impact of coding differences over the 3 prior years, CMS did not account for any additional years of accumulated impact in its 2011 or 2012 estimates.
Third, CMS only accounted for differences in age and mortality between the Medicare Advantage and traditional Medicare populations. GAO accounted for additional beneficiary characteristics, such as sex, diagnoses as a proxy for health status, Medicaid enrollment status, beneficiary residential location, and whether the original reason for Medicare entitlement was disability, thereby improving the accuracy of the estimate.
CMS could enhance its efforts to estimate effects of coding differences between Medicare Advantage and traditional Medicare and realize even greater cost savings than the $2.7 billion that it has identified. GAO demonstrated a methodology which incorporated additional data and identified additional savings$1.2 billion to as much as $3.1 billion in payment reductions to Medicare Advantage plans. In January, 2012, GAO made the following recommendations:
To help ensure appropriate payments to Medicare Advantage plans and improve the accuracy of the adjustment made for differences in coding practices over time, the Secretary of Health and Human Services should direct the Administrator of CMS
The information contained in this analysis is based on findings from the report listed in the related GAO product section. GAO estimated the impact of coding differences between Medicare Advantage and traditional Medicare on Medicare Advantage risk scores and payment to plans. GAO compared risk score growth for Medicare Advantage beneficiaries with an estimate of what risk score growth would have been if they had been in traditional Medicare.
GAO provided a draft of its January 2012 report to the Department of Health and Human Services. The Department of Health and Human Services did not comment on GAOs recommendations but provided general and technical comments, which were incorporated as appropriate. The Department of Health and Human Services characterized GAOs results as similar to those obtained by CMS, and found GAOs methodological approach and findings informative.
For additional information about this area, contact James C. Cosgrove at (202) 512-7114 or firstname.lastname@example.org.
GAO found that diagnostic coding differences exist between MA plans and Medicare FFS. Using data on beneficiary characteristics and regression analysis, GAO estimated that before CMSs adjustment, 2010 MA beneficiary risk scores were at least 4.8 percent, and perhaps as much as 7.1 percent, higher than they likely would have been if the same beneficiaries had been continuously enrolled in FFS...