SummaryDeferral of care during the COVID-19 pandemic is resulting in fewer claims and diagnoses among Medicare Advantage (MA) enrollees, which could lead to a 3%–7% reduction in 2021 risk scores and lower plan payments in 2021.
On June 1, MA plans submitted their bids for 2021, including their projected risk scores for the upcoming plan year. Risk scores reflect an individual’s health status and are a critical part of determining plan payment. The higher the risk score, the greater the expected healthcare costs for that individual, and thus the plan receives a higher payment; the lower the risk score, the lower the expected cost for that beneficiary, and the lower the payment to the plan. More specifically, the Centers for Medicare & Medicaid Services (CMS) uses the projected risk scores from a plan’s bid, which reflect that plan’s estimate of the average risk score of their enrollees, to determine 2 components of the plan’s payment:
- The standardized payment amount, calculated as the plan’s bid divided by the projected average risk score from the plan
- The rebate amount, calculated as a percentage of the difference between the plan’s bid and the county benchmark in which the plan operates
For the rebate amount, the specific percentage used is based on the plan’s star rating. Plans use these rebates to lower premiums or provide supplemental benefits for their enrollees.
The CMS uses diagnoses data based on inpatient, outpatient, and physician services to calculate an MA plan enrollee’s risk score. For 2021, risk scores will be based on member demographics and diagnoses from services that occur in 2020. During the COVID-19 pandemic, beneficiaries have delayed accessing services, either to free up provider capacity to respond to the pandemic or to adhere to social distancing guidelines. As a result, as recently reported, medical claims are lower than they would have been otherwise, indicating that changes in utilization have occurred because of COVID-19.1
Because diagnoses from the 2020 claims are used as an input to determine 2021 risk scores, fewer claims in 2020 could mean lower risk scores and, consequently, lower payments for plans in 2021. The CMS recently announced that MA plans will be permitted to use diagnoses recorded during telehealth visits under certain circumstances for risk adjustment, which could mitigate some of the impact of the decrease of in-person visits.
While the risk scores included in plans’ bids for 2021 are estimates, the degree of uncertainty surrounding healthcare utilization during the pandemic substantially increases the risks associated with this projection as compared to previous years.2 If actual risk scores are lower than what plans projected in their bids because of decreased utilization, then the payments to plans could be lower than the plans’ costs.
Changes in Claims and Diagnosis Counts in 2020
To assess the potential impact of the pandemic on service utilization, Avalere analyzed 2019 and 2020 claims data from a sample of MA plans to identify changes in the number of claims and diagnoses recorded during the pandemic.
As shown in Figure 2, the percentage of enrollees who had at least 1 claim is dramatically lower in April 2020 than in both prior months of 2020 as well as in April of the prior year. In January, February, and March, 28%, 27%, and 21% of enrollees had at least 1 claim, respectively. But in April, only 4% of enrollees had a claim. This type of decrease is unusual, given that the percentage of enrollees with at least 1 claim in March and April 2019 was 23%. A reduction in enrollees with at least 1 claim could potentially have a large impact on risk scores, particularly if that claim was the only encounter where a diagnosis would have been collected.
Source: Avalere analysis of 2019 and 2020 claims and enrollment data.
As shown in Figure 3, the number of diagnoses recorded per enrollee decreased substantially during the public health emergency. In January and February, 28% of enrollees in the selected plans had at least 10 diagnoses, while in March, 24% had at least 10 diagnoses. Yet in April 2020, only 14% of enrollees had 10 or more diagnoses. Among members with a claim, there were, on average, 12 diagnoses per member in January and 11 diagnoses per member in February. This figure had dropped to an average of 10 diagnoses per member in March, and 8 diagnoses per member in April 2020.
Source: Avalere analysis of 2020 claims data.
The analyses show that the number of claims per person and the number of diagnoses per person have both declined during the pandemic. The reduction is particularly notable with respect to claim volume. The trend toward fewer claims and diagnoses than anticipated will likely continue through the rest of 2020, though the magnitude of the impact could vary depending on location, timelines for reopening, and the trajectory of COVID-19 diagnoses, among other reasons.
The increase in the percentage of enrollees with at least 1 claim in January 2020 compared to January 2019 could reflect multiple factors but does not by itself indicate that risk scores would be higher in 2021 than in 2020. Risk scores are based on an entire year’s worth of data, rather than 1 or 2 months. That is, the key factor to consider for the 2021 risk scores is not an increase or decrease of claims in any 1 month. Rather, as discussed below, the most crucial element in determining 2021 risk scores is the claims volume from May through December 2020.
Risk Score Impact Scenarios
To evaluate the potential impact of the pandemic on 2021 risk scores, Avalere used 2020 claims data from January to April for the sample of plans analyzed and then modeled different risk score scenarios based on different estimated utilization patterns for the remainder of the year. First, Avalere modeled a baseline scenario in which claims from May through December of 2019 were used as a proxy for the claims for May through December 2020. Second, Avalere modeled 3 scenarios—low, medium, and high impact—and compared them to what risk scores likely would be compared to the baseline. While there could be an increase in volume in claims in 2020 due to pent-up demand, this analysis assumes that the increase would take place in 2021 due to a potential second wave of COVID-19 infections in the fall of 2020.
All 3 scenarios assume an 80% reduction in claims in May 2020, similar to the claim reduction in April 2020. The low impact scenario assumes a 30% reduction in claims in June, and a 10% reduction in each month from July through December 2020. The medium impact scenario assumes a 45% reduction in June and then a 25% reduction through the end of the year, while under the high impact scenario, claims would decrease by 60% in June and 40% each month thereafter.
|Baseline||Low Scenario||Medium Scenario||High Scenario|
|Estimated Risk Score||1.000||0.969||0.949||0.927|
|Percent Decrease in Risk Score Due to Pandemic||0%||3%||5%||7%|
Source: Avalere modeling based on 2019 and 2020 claims data. The 2020 baseline score has been normalized to 1.0.
Depending on the extent to which utilization resumes during the rest of 2020, MA plans could experience an average decrease in risk scores of between 3% and 7% below what they would have anticipated without the COVID-19 pandemic. Notably, if there is a second wave of cases, utilization could drop even lower than accounted for in the high impact scenario, more closely resembling the decreases in claims plans experienced in April 2020. In this case, risk scores could decrease by more than 7%.
Potential Implications for Plan Payment
Lower than anticipated 2021 risk scores would have an impact on payments for MA plans. Due in part to how the CMS calculates the payments to plans, lower risk scores could mean that plans would not receive payments equivalent to their expected costs in the coming year.
For example, if a plan’s bid is $900 per member/per month (PM/PM), the plan expects that the incurred costs per member will be $900 PM/PM in 2021. If the plan’s expected risk score is 1.0, as in the baseline scenario, the plan’s standardized payment amount would be $900, and the plan payments would be calculated by taking the product of $900 and the risk score. Therefore, a plan would receive a payment of $900 for a member with a risk score of 1.0 ($900 × 1.0) and $931.50 for a member with a risk score of 1.035 ($900 × 1.035). If, however, the actual risk score in 2021 is 0.927 or 7% lower than projected, as in the high impact scenario, then the payment would be $834.30 ($900 × 0.927), and the plan would lose nearly $66 PM/PM. Underpayments could range from $28 to $66 PM/PM based on the scenarios analyzed here.
Furthermore, deferred care from 2020 could lead to higher healthcare costs for plans in 2021. For example, Avalere analysis of pre-adjudicated fee-for-service Medicare claims finds a substantial decrease in the number of patients with chronic obstructive pulmonary disease and heart failure with a claim in April 2020 compared to April 2019—a reduction of 48% and 45%. This indicates that Medicare beneficiaries with chronic conditions may be having difficulty managing these conditions due to the pandemic. If individuals with chronic illnesses have not seen their physicians, their disease states could worsen and lead to higher costs for plans. That is, plan losses could be even greater if costs are higher than what plans predict in their 2021 bids.
Considerations for MA Plans
The COVID-19 pandemic will have an impact on utilization patterns through 2020 and potentially beyond. Plans must continue to adjust their strategies to account for the longer-term implications of deferred care. In particular, plans should consider the actions they can take now to ensure that beneficiaries are accessing appropriate care and that plans are accurately projecting risk scores, including:
- Increasing Use of Telehealth: Plans must ensure they are prepared to use the new flexibilities permitted by the CMS to engage with beneficiaries via telehealth and collect and submit data from these encounters. In particular, plans should work with primary care providers (PCPs) conducting telehealth visits to ensure that the providers are equipped (e.g., have the necessary training and tools) to identify, address, and document any gaps in in quality and risk reporting. Additionally, plans can also help their members become more comfortable using this technology to interact with their providers.
- Facilitating Provider/Vendor Coordination: As plans collaborate with vendors to provide services and assess beneficiary health (e.g., document diagnoses), plans should ensure that information is being shared regularly with the enrollee’s PCP. In addition, plans should collaborate with vendors who interact directly with enrollees to encourage enrollees to visit their PCPs and enroll in available support services (e.g., disease management programs).
- Communicating with Beneficiaries: Beneficiaries may be uncertain about if and when they are able to schedule routine appointments during the pandemic. Plans should increase efforts to communicate with members to discuss scheduling or rescheduling appointments with them. Plans should also emphasize safety precautions taken by their providers in light of COVID-19.
- Addressing Social Risk Factors: In addition to addressing beneficiaries’ clinical needs, plans should use telehealth and other enrollee outreach strategies to identify beneficiaries who are experiencing social risk factors, such as social isolation, food or housing insecurity, and therapeutic non-adherence, which could contribute to worse health outcomes. Addressing social determinants of health is a critical component of care management; if plans do not appropriately coordinate care during this time, there could be a significant increase in utilization and costs in 2021.
- Improving Quality: While quality reporting to the CMS for 2020 has been delayed during the pandemic, plans should continue to undertake activities to promote high quality care and address gaps in quality. Plans will be expected to submit quality data in 2021 to the CMS; the 2021 data will include information on activities that will take place in 2020. Further, the 2021 quality data will be used to determine 2022 ratings, which in turn will affect 2023 payment. As such, if plans delay investments in quality activities in 2020, they could see lower Star Ratings and lower payments in future years. As a result, plans should continue to undertake quality initiatives.
- Analytics to Determine Drivers of 2021 Risk Scores: Plans should consider projecting 2021 risk scores to determine their potential underpayment. In addition, they should focus member interventions on those enrollees who have not been utilizing medical care appropriately during the pandemic by reviewing electronic health records or conducting traditional paper-based medical record retrieval.
- Evaluating Impact on Medical Loss Ratios (MLRs): If the lower claim volume from April 2020 continues, the trend would indicate that plan MLRs could be lower than the 85% threshold required by law. If a plan’s MLR is below 85%, then the CMS will recoup the difference between the plan’s MLR and the 85% threshold. Plans should forecast the impact of the pandemic on their 2020 MLRs in advance of the reporting deadline in December 2021 and consider investments in quality activities that might help them avoid reductions in their MLR and improve patient care.
Inovalon and Avalere are actively engaged with plans to help address these and other issues due to COVID-19. To learn more about our capabilities, connect with us.
Check out our COVID-19 Intel Center.
Avalere used claims data from a group of health plans for 2019 and through the first 4 months of 2020. Avalere identified enrollees as of April 2020 in these health plans and calculated risk scores for 2021 based on claims projections for 2020 for these enrollees. Risk scores were assigned based on the risk factor code contained in the CMS’s May 2020 Monthly Membership Report.
First, Avalere modeled a baseline scenario in which claims from May through December 2019 were used as a proxy for the claims for May through December 2020. Second, Avalere modeled 3 scenarios—low, medium, and high impact—and compared them to what risk scores likely would be compared to the baseline. Avalere calculated an average risk score for the baseline scenario. Then, Avalere divided all risk scores by this average risk score, so that the average risk score would equal 1.0. Avalere divided the risk scores for each of the 3 scenarios by the average risk score for the baseline scenario. Individuals on dialysis and those who have received a kidney transplant were excluded from the analysis.
- HCA. “Healthcare Reports First Quarter 2020 Results,” April 21, 2020; United Health Group. “First Quarter 2020 Earnings Release,” April 15, 2020 Medical Group Management Association. “COVID-19 Financial Impact on Medical Practices,” April 8, 2020.
- “COVID-19 Posed Challenges for 2021 Medicare Advantage Bids”, Health Plan Weekly 30.24, June 15, 2020.
produces measurable results. Let's work together.