Thirteen months after over 500 comments were submitted in response to a CMS proposal, Medicare Part C (“Medicare Advantage”) plan sponsors and other stakeholders now know the methodology CMS will use in calculating payment errors through extrapolated estimates in audits based on risk adjustment data validation (“RADV”). On February 24, 2012, CMS published the Final Payment Error Calculation Methodology for Medicare Advantage RADV Audits (the “Methodology”). CMS states that the purpose of RADV audits is to determine whether diagnosis codes submitted by Medicare Advantage organizations can be validated by medical record documentation. Those instances in which diagnoses cannot be validated through medical records will form the basis of the extrapolation. CMS plans to audit approximately 30 Medicare Advantage contracts each year, which it estimates could amount to overpayment recoveries of approximately $370 million in the first audited year. The Methodology states that CMS “expects that these contract-level audits will have a sentinel effect on the quality of risk adjustment data submitted for payment by MA organizations.”
Notwithstanding some prior audits and the position taken by CMS in its proposed methodology, audits for payment year 2011 will be the first time that CMS will conduct RADV audits using extrapolated estimates. After contracts have been selected for RADV audits, enrollees will be sampled from each contract, and the results of medical record review of those sampled enrollees will be extrapolated to the universe of RADV-eligible enrollees.
To be considered “RADV-eligible,” Medicare Advantage enrollees must:
- be enrolled in a Medicare Advantage contract in January of the payment year based on CMS enrollment files;
- be continuously enrolled in the same Medicare Advantage contract from January of the data collection year through January of the payment year;
- neither have end stage renal disease (ESRD) nor hospice health statuses from January of the data collection year through January of the payment year;
- be enrolled in Medicare Part B for all 12 months during the data collection year;
- have at least one risk adjustment diagnosis submitted during the data collection year that led to at least one CMS-Hierarchical Condition Category (HCC) assignment for the payment year.
To determine the enrollees subject to medical record review, RADV-eligible enrollees will be divided into three groups, based on risk score, and an equal number of enrollees from each risk score category will be selected. CMS will select a total of up to 201 enrollees for medical record review (this number may be less for contracts with smaller numbers of RADV-eligible enrollees). Then, a “sample weight” is assigned to each sampled enrollee that is projected to the universe of RADV-eligible enrollees in the contract’s risk score categories.
The Methodology offers the following example:
If a contract has 3,000 RADV-eligible enrollees, the enrollees would be ranked by risk score, then divided into three equal groups of 1,000 enrollees each (to represent high, medium, and low strata). An equal number of enrollees will be randomly selected from each group. The weight for each sampled enrollee will equal 14.925 (i.e., 1,000/67).
CMS states that it will allow Medicare Advantage plan sponsors with audited contracts to submit multiple medical records for each CMS-HCC being validated. However, “all diagnoses will be abstracted from the first medical record that validates the CMS-HCC under review” and the “one best medical record policy will continue to apply to RADV audit dispute and appeal processes.” It remains to be seen how these policies can maintain consistency considering that the utility of submitting multiple medical records may be undercut by having the one best medical record policy continue to apply to the audit review and appeal processes. CMS offers that more guidance will be provided to Medicare Advantage plan sponsors addressing the RADV audit procedures at the time contracts are audited.
CMS will use the results of the RADV audit to calculate, at a 99 percent confidence interval, a contract’s payment error. If the confidence interval of the estimated payment error for the Medicare Advantage contract is zero or below zero, the recovery amount will be “constrained to zero.” If the confidence interval of the estimated payment error for the Medicare Advantage contract is greater than zero, then CMS will apply a “Fee-for-Service Adjuster” (“FFS Adjuster”) amount. Details surrounding the FFS Adjuster, such as the extent of and how often adjustments will be made, are unclear. CMS states:
The FFS Adjuster accounts for the fact that the documentation standard used in RADV audits to determine a contract’s payment error (medical records) is different from the documentation standard used to develop the Part C risk-adjustment model (FFS claims). The actual amount of the adjuster will be calculated by CMS based on a RADV-like review of records submitted to support FFS claims data.
Although the Methodology provides much needed guidance with respect to how the agency will conduct these audits, there are still a number of unanswered questions.