On Friday, March 31, 2023, the Centers for Medicare & Medicaid Services (CMS) released the Calendar Year (CY) 2024 Medicare Advantage (MA) Capitation Rates and Part C and Part D Payment Policies (Rate Announcement). This Rate Announcement follows CMS’s February 1 notice of planned changes to rates and the risk adjustment methodology, which provided an opportunity for the public to submit comments during a 30-day period (Advance Notice), as required by Section 1853(b)(2) of the Social Security Act (the Act). The Rate Announcement — providing for 60 days prior to the bid submission deadline of June 5, 2023 — provides notice of the annual capitation for MA for CY2024 related to the benchmark, risk adjustment, and other factors to be used in adjusting rates and responds to all substantive comments received from the Advance Notice.
We summarize the key factors and adjustments to the overall expected average rate increase of 3.32% (which is about $13.8 billion more than CY2023, and an increase from the 1.03% in the Advance Notice), and comments from CMS. We also highlight other key developments affecting MA rates, notably relating to the MA risk adjustment methodology (the 2024 Risk Model). Major changes identified in the Rate Announcement include updates to the risk adjustment model that uses International Classification of Diseases (ICD)-10 codes instead of the ICD-9 system, using data from 2018 diagnoses and 2019 expenditures, and the removal or reclassification of codes disproportionately coded in MA compared to Medicare Fee-For-Service (FFS) that CMS does not consider to accurately reflect increased costs to care for beneficiaries.
General Overview of MA Rates
To help understand the impact of the Rate Announcement, we explain how rates are established pursuant to 1853 of the Act (42 U.S.C. §1395w-23):
- Medicare Advantage Organizations (MAOs) are paid a base rate by CMS. The base rate is the benchmark (bidding target) established by CMS, or if the MAO plan bid is lower than the benchmark, then the MAO is paid the bid amount. Bid amounts include medical costs plus plan administrative costs and profit.
- Base rates are risk adjusted to account for certain factors related to a beneficiary (e.g., health condition, age, gender, dual eligibility status, and demographics information) based off a risk score from CMS’s hierarchical condition category (HCC). The HCC uses diagnoses billed for inpatient, outpatient, and physician visits to account for the full range of healthcare spending for Medicare FFS beneficiaries. All ICD codes are grouped into diagnostic groups with similar medical conditions and then condition categories based on clinical profiles and expected costs (e.g., diabetes, depression, heart disease). Hierarchies are imposed on the conditions depending on if there are less and more severe manifestations of the disease. HCCs are then divided by codes that better predict costs and, thus, included in the risk adjustment model (referred to as “payment HCCs”), and those that do not anticipate costs because of the rare nature of a condition or the condition does not have well-specified diagnostic coding (referred to as non-payment HCCs). See Rate Announcement at p. 78. HCCs and changes to the HCCs are included as part of the annual rate notice (e.g., the Rate Announcement).
- Plans that bid above the benchmark receive a premium from beneficiaries that is equal to the difference between the bid and the benchmark. Plans that bid below the benchmark receive a rebate equal to a percentage of the difference between the benchmark and the bid. Any rebate received must be used for the benefit of beneficiaries in the form of supplemental benefits or lower premiums.
The benchmark is determined by calculating the average per capita FFS Medicare spending for each county. This is determined by calculating the national estimated FFS Medicare per capita costs for the following year ($1,105.10 for 2024), which is multiplied by a county-specific geographic index for the 5-year rolling average for FFS Medicare spending in the county, weighted for enrollment and average risk score. We note that there are variations of this formula for local and regional plans, which are outside the scope of this article. We also note that the benchmark is adjusted, subject to a cap, for MAOs that meet Quality Star Ratings (4 or higher). See Section 1853(o) of the Act; Medicare Managed Care.Ch.7.Risk Adjustment.
Summary of the Major Changes in the Rate Announcement
The overall expected average rate change is 3.32% of MA revenue, year over year, based on the following adjustments:
- Effective Growth Rate (EGR) of 2.28%
- Impact from Star Rating changes of -1.24%
- Risk model revision and normalization of -2.16%
- Risk score trend of 4.44%
EGR: The EGR reflects inflation and the most up-to-date estimate of Medicare FFS per capita costs adjusted to account for MA plan payments. The benchmark originates from Medicare FFS data, which does not account for practice patterns of providers contracted in managed care. The 2024 EGR also accounts for effects from COVID-19 and other changes, such as lower morbidity from excess COVID-19-related deaths, lower total spending due to shifts in care from inpatient to outpatient settings (e.g., hip and knee replacements), and to reflect an increase in dual-eligible beneficiaries enrolling in MA. We note that the greatest increase in enrollment in MA for 2024 was in dual-eligible special need plans. See Rate Announcement at pp. 3 and 101. EGR is also adjusted for a technicality related to removing indirect and direct medical education costs from historical and projected expenses associated with MA beneficiaries.
One of the biggest changes from the Advance Notice is that the technical update and, as discussed below, the risk model revisions, will be phased in over 3 years with a blend of 67% of the adjustment under the current 2020 model and 33% of the adjustment under the 2024 model (Phase In).
Impact from Star Rating: The overall Quality Star Rating, which can increase an MAOs benchmark rate, is anticipated to decrease for the 2023 Star Ratings (which is what is used for 2024 Quality payments).
Risk Model Revision and Normalization: Risk model revision and normalization reflects the changes expected due to the updated risk model. These changes are the most significant of all as reflected in the Advance Notice. More specifically, the model will now reflect use of ICD-10 coding, changes, and reclassifications to the HCC codes and updated underlying FFS data. Based on public comments from the Advance Notice, CMS will implement these changes subject to the Phase In and anticipates that by 2026 100% of the risk scores will be calculated using the updated model in the Rate Announcement.
MA Risk Score Trend: The MA Risk Score Trend is the CMS estimate of the average increase in risk scores, not accounting for normalization and MA coding adjustments. It is determined, using the upcoming risk adjustment model (in this case, CY 2024 Risk Model), by calculating the MA risk score over three prior years and then calculating the average annual change in risk scores across those three years based. See Rate Announcement at p. 87. We note that the MA Risk Score is also subject to the Phase In.
2024 Risk Model
It is important to keep in mind that risk adjustment was implemented to discourage MAOs from cherry picking healthy Medicare-eligible beneficiaries over those beneficiaries that are sicker and, thus, more costly. See UnitedHealthcare Ins. CO. v. Becerra, 16 F.4th 867, 873-74 (D.C. Cir. Aug. 13, 2021, reissued Nov. 1, 2021), cert. denied, 142 S. Ct. 2851 (U.S. June 21, 2022) (No. 21-1140) (“Risk Adjustment is intended to reduce or eliminate the incentives to enroll only the healthiest.”). The risk-adjusted premium protects MAOs by accounting for increased costs for care provided to beneficiaries with prior health conditions. In response to public comments, CMS explains:
“Since MA plans bear full financial risk for the population and services they cover, a main objective of risk adjustment is to minimize incentives for MA plans to compete for the healthiest beneficiaries. The risk adjustment model accomplishes this by accurately predicting relative risk across subgroups of beneficiaries and reimbursing plans more for populations that are expected to be sicker and have more complex health needs. This is achieved through the segmentation of the model that assigns unique risk scores for each condition based on demographic factors. A key focus of the MA risk adjustment model is to accurately predict costs that are attributable to characteristics that are present over time (e.g., chronic conditions that persist or affect longer term costs, demographics, etc.). In this sense, the model is an insurance-like model that seeks to balance the over and under prediction errors so that the average actual expenditures for a sufficiently large group of beneficiaries equals the average predicted expenditures.”
See Rate Announcement at p. 70. To predict foreseeable costs more accurately for subgroups of beneficiaries (those based on risk scores), CMS undertook a comprehensive evaluation of coding diagnoses under ICD-10 as compared to ICD-9 and conditions (i.e., HCCs) for inclusion in the 2024 Risk Model. CMS follows 10 principles to guide its categorization and determination of diagnostic codes to an HCC in its prediction of Medicare expenditures and how the diagnostic groupings should interact for risk adjustment purposes. See Report to Congress: Risk Adjustment in Medicare Advantage (Dec. 2021).
The 2024 Risk Model uses diagnoses from 2018 and costs reflected from 2019 as opposed to the 2020 Risk Model that used 2014 diagnoses and 2015 costs. In addition, the 2024 Risk Model uses ICD-10 codes to reflect how providers code for services, instead of the previous ICD-9 version. These changes impacted the HCC model, causing many codes to no longer map to an HCC. In addition, CMS made changes to payment HCCs. The HCC reconfigurations and conditions most impacted by the 2024 Risk Model include changes to the following disease categories or conditions: (1) vascular, (2) metabolic, (3) heart, (4) blood, (5) amputation, (6) neurological, (7) diabetes, (8) kidney, (9) psychiatric and (10) musculoskeletal. Here are examples to illustrate the impacts of the 2024 Risk Model:
Diabetes: Under the 2020 Risk Model, there were 3 payments HCCs. The 2024 Risk Model has 4 payment HCCs; however, codes for diabetes with unspecified complications or complications related to blood sugar were moved to the lowest payment rung, and drug-induced diabetes codes were categorized to a non-payment HCC.
Psychiatric: Several changes were made to the classification model. Notably, diagnoses for mild, unspecified remission, subsequent encounter, and sequela codes are now a non-payment HCC due to the diagnoses not being clinically meaningful or a reliable predicter of future costs.
Congestive Heart Failure: The 2024 Risk Model takes 1 HCC and breaks it into 5 HCCs to better capture the ranges of heart failure severity.
Removal of Certain HCCs: The 2024 Risk Model removes more than 2,000 codes from the model; however, 97% of this change is due to a code no longer mapping to an HCC. The net result of the HCC reclassification is 266 HCCs (as compared to 204 under the 2020 Risk Model) and 115 payment HCCs—representing 43.2%, as compared to 86 (42.2%) under the 2020 Risk Model. ICD-10 has a larger code set and includes more specificity. Bottom line is that 10.5% of ICD-10 codes now map to a payment HCC, as opposed to 13.3% under the 2020 Risk Model. Protein-Calorie Malnutrition is a key HCC removed because CMS indicated stated that empirical data showed that coding differences between FFS and MA were too variable and, thus, not a good predictor of costs for this diagnosis.
Risk Model Related to Dual-Eligibles
Per the 2024 Risk Model, risk scores for special needs plans (SNPs), e.g., those dually eligible for Medicare and Medicaid, are 50% higher than traditional MA plans. In addition, the risk score trend for dual-eligibles is 4.67% higher than for non-dual eligible beneficiaries. This is significant given the increased enrollment of Medicare beneficiaries that are dual-eligibles, as noted above.
The 2024 Risk Model is being implemented because CMS believes that the new model will better direct resources to beneficiaries with higher health care costs. CMS states that the new model improves or maintains predictive accuracy based on its examination of predictive ratios across the subgroups of beneficiaries, including dual-eligibles. Further, CMS notes that “[t]he risk adjustment model used for MA payment is not designed to drive clinical behavior to look for specific conditions or to be the sole purpose for MA organizations or health care providers to identify and treat conditions that are potential precursors to adverse medical events or complicating factors in the identification and treatment of other conditions.” See Rate Announcement at p. 76.
Other Activities to the MA Program that May Impact Risk Adjustment
In addition to the Rate Announcement, CMS recently finalized new regulations for conducting MA Risk Adjustment Data Validation (RADV) audits, effective April 3, 2023, which will be codified at 42 C.F.R. 422. See 88 Fed. Reg. 6643 (Feb. 1, 2023). In the Advance Notice, CMS predicted that the changes to the risk scores and HCC updates will help prevent overpayments by improving the accuracy of payments made to MAOs. CMS further stated:
“Separate from the proposals in the 2024 Advance Notice, CMS also recently finalized policies for the [RADV] program. The RADV program is the primary audit and oversight tool to address improper payments in the MA program. … CMS will extrapolate RADV audit findings beginning with Payment Year 2018 as part of our statutory and fiduciary responsibilities to reduce and recover improper expenditures of taxpayer dollars.”
See Advance Notice FAQs.
Also of note is the Medicare Payment Advisory Commission (MedPAC) March 2023 report, which includes recommendations to address disparities in coding intensity and other coding differences between Medicare FFS and MA that have led to increasingly inflated risk-adjusted payments. See March 2023 Report to the Congress: Medicare Payment Policy, MedPAC at 332 (Mar. 15, 2023) (MedPAC Report). The MedPAC Report also probes the efficacy of MA’s quality bonus program in rewarding or improving quality of care, and calls on Congress to implement a value-based incentive program instead.
Lastly, on December 27, 2022, CMS released a Proposed Rule that has the potential to impact MAO and provider value-based payment arrangements that look to incentive quality measures based on MA Star Ratings. In addition to making changes to quality measures and utilization management standards, the Proposed Rule looks to align the identification of overpayments with the False Claims Act (FCA) so that an overpayment is “identified” when the provider or supplier has actual knowledge of the existence of the overpayment or acts in reckless disregard or deliberate ignorance of the overpayment. This would effectively make improper coding (due to unsubstantiated medical records or other reasons) an overpayment subject to the FCA.
With the growth of the MA program, which is close to 50% of all eligible Medicare beneficiaries opting to enroll in a MAO over Traditional Medicare, there is increased scrutiny over how rates are calculated and adjusted. This all aligns with other federal agency actions in recent years related to scrutiny of MA and provider risk adjustment and coding practices, including numerous high-profile FCA lawsuits, investigations, and settlements premised on overpayments to MAOs due to alleged improper and fraudulent coding and documentation with an intent to receive high payments for beneficiaries.
Several commenters pointed out that value-based care models would suffer from the Advance Notice. CMS responded that MAOs are allowed to use a portion of savings to furnish additional benefits and to develop cost-efficiencies, and that reducing morbidity and mortality through early-stage disease or prevention is an “inherent expectation of a capitated managed care system.” CMS also noted that the 2024 Risk Model changes do not limit or change an MAO’s obligation to use a network of providers and to ensure that covered benefits are available and accessible to beneficiaries. One thing is clear, though: These recent developments will impact MA and any value-based payment arrangements entered by providers, risk-bearing entities, or intermediaries and an MAO. Plans and providers should be prepared to renegotiate and revise their payment arrangements considering the 2024 Risk Model to ensure that the financial expectations underlying the terms of any such contract remain intact.