What have major Medicare Advantage settlements revealed about corporate coding and risk‑adjustment practices by state?
Executive summary
Major Medicare Advantage (MA) settlements have exposed recurring corporate practices—retrospective chart reviews, third‑party coding vendors paid on contingency, post‑visit addenda, and systematic submission of additional diagnosis codes to boost risk scores—that federal prosecutors say inflated payments to plans, with enforcement actions and Corporate Integrity Agreements arising in multiple states where large MA plans operate [1] [2] [3]. Empirical research and CMS adjustments also show that coding intensity varies across plans and likely across geographies, and that the federal “coding intensity” adjustment only partially offsets those differences [4] MedPACReportTo_Congress_SEC.pdf" target="blank" rel="noopener noreferrer">[5] [6].
1. Settlements expose a repeatable playbook: vendors, addenda, and retrospective coding
Recent high‑profile resolutions allege a common sequence: MA plans or their vendors review prior medical records, solicit additional diagnoses (sometimes via addenda months later), and submit those codes to CMS to raise enrollee risk scores—practices central to the Independent Health/DxID case out of New York and mirrored in other DOJ actions [2] [1] [7]. DOJ’s public materials and legal reporting show vendors like DxID were compensated on a contingency basis, creating an explicit financial incentive to identify and submit extra diagnoses [1] [3].
2. State footprints reflect where big plans and vendors operate, not a single‑state problem
Settlements have clustered around large MA operators and their vendors headquartered or active in specific states—Independent Health’s case involved a New York plan and vendor activity there, and DOJ’s later, larger settlements implicate other major operators such as Kaiser with broad regional footprints—indicating the risk‑adjustment enforcement pattern follows market concentration rather than unique state law environments [2] [8] [3]. Reporting does not support a conclusion that particular states are uniquely responsible; instead, actions track where big MA plans and coding vendors conduct chart‑review programs [8] [3].
3. Data studies confirm variation in coding intensity that maps onto plan and geographic differences
Academic and policy analyses find MA risk scores are systematically higher than comparable fee‑for‑service (FFS) scores—MedPAC and peer‑reviewed studies estimate coding intensity inflates MA risk scores by roughly 7–12 percent before CMS’s uniform adjustment—evidence that coding practices vary across plans and regions and can produce billions in excess payments nationally [5] [4] [9]. Those findings imply that state‑level or regional patterns of higher MA payments likely reflect both plan behavior and differences in how record‑keeping and coding are executed locally [5] [4].
4. CMS’s universal coding‑intensity adjustment is blunt and incomplete
CMS applies a uniform coding‑intensity adjustment (about a 5.9 percent deflation in recent years) to counterbalance MA plans’ stronger coding incentives, but researchers and MedPAC conclude that this adjustment only partially offsets estimated overcoding—leaving residual variation across plans and geographies that enforcement actions now target [4] [6] [5]. The settlements underscore government views that individual plans’ practices—not just a systemic payment rule—matter for liability and repayments [3] [10].
5. Enforcement patterns, corporate remedies, and whistleblowers shape compliance incentives
DOJ settlements routinely pair monetary payments with five‑year Corporate Integrity Agreements requiring independent reviews, auditors, or annual record audits; whistleblowers have driven many suits and received large relator shares, signaling heightened detection and private enforcement across states where MA plans operate [3] [1] [7]. Defendants often settle without admitting wrongdoing, reflecting litigation risk management while leaving open continued regulatory scrutiny [7].
6. Limits of the record and competing framings
Available reporting documents national trends, specific corporate practices, and plan‑level allegations, but does not provide a comprehensive, state‑by‑state map quantifying which states have the worst coding practices or the precise dollar impact by state; empirical studies illustrate geographic variation but cannot attribute all differences to fraudulent intent versus legitimate documentation intensity, and industry advocates argue that robust coding simply reflects better capture of patient complexity [4] [5] [6].