Options to Make Medicare More Affordable For Beneficiaries Amid the COVID-19 Pandemic and Beyond
Data and Methods for Adding an Out-of-Pocket Limit to Traditional Medicare
To analyze the effects of an out-of-pocket limit, KFF collaborated with Actuarial Research Corporation (ARC) to develop a model to assess the spending effects for Medicare, beneficiaries, and other payers of options to add an annual out-of-pocket limit to Medicare, assuming full implementation in 2020. The model is primarily based on individual-level data from the Centers for Medicare & Medicaid Services (CMS) Medicare Current Beneficiary Survey (MCBS), which are calibrated to match aggregate Congressional Budget Office (CBO) Medicare spending and enrollment estimates and projections.
We first developed a current-law baseline for 2020 by identifying Medicare reimbursements for each individual in traditional Medicare (excluding beneficiaries enrolled in Medicare Advantage plans), inferring the individual’s cost-sharing obligations under current law, and dividing those obligations between the individual and their supplemental insurer as appropriate. We calculated Medicare and supplemental plan premiums and added these amounts to beneficiaries’ out-of-pocket costs. Next, we simulated the effects of adding an out-of-pocket limit by modifying cost-sharing obligations based on various limits. We assumed that beneficiaries would use more services with an annual spending limit and that some beneficiaries would switch into or out of traditional Medicare, Medigap, or Medicare Advantage in response to this change.
Although MCBS includes Medicare beneficiaries who are enrolled in Medicare Advantage, we excluded this group when evaluating the individual-level spending effects of adding an out-of-pocket limit because the option modifies traditional Medicare. The model does incorporate indirect effects on aggregate Medicare Advantage spending and enrollment, based on the assumptions that changes in traditional Medicare reimbursement would be reflected in Medicare Advantage payments, and that aggregate Medicare Advantage payments will change to the extent that some beneficiaries switch between traditional Medicare and Medicare Advantage.
Data and Methods for Options to Improve Financial Protections for Low-Income Beneficiaries
For these options, our analysis uses data from the following sources: the CMS Medicare Current Beneficiary Survey 2017 Survey file; a 20% sample of Medicare beneficiaries from the CMS Chronic Conditions Data Warehouse (CCW), 2017; and the Urban Institute’s DYNASIM4 microsimulation model, using data for 2017.
Some people dually eligible for Medicare and Medicaid have incomes and/or assets higher than the MSP and LIS eligibility limits because some states do not have asset tests and/or have higher income limits for their Medicare Savings Programs. States may also allow people with higher incomes and/or assets to qualify for Medicaid through specific pathways, such as the nursing home or medically needy pathways. All Medicare beneficiaries who are dually eligible for Medicaid automatically qualify for Part D LIS benefits, irrespective of their incomes and assets. The total number of people receiving MSP and Part D LIS benefits in 2017 came from the CCW.
We estimated the number of additional people who could be eligible for Medicare Savings Program benefits and Part D LIS benefits if eligibility criteria were changed to allow all people with incomes below 150% of the poverty level (FPL) to be eligible, with no limits on assets. The MCBS provided information about the distribution of income among people receiving Medicare Savings Program and Part D LIS benefits. The data from the DYNASIM model was used to estimate the total number of people with incomes below 150% of the FPL. The estimated number of MSP beneficiaries with incomes below 150% of the FPL was subtracted from the estimated total number of people with incomes below 150% of the FPL to estimate the number of additional people who would be eligible for the MSPs if the income threshold was 150% of the FPL and no asset test was imposed. The same process was used to estimate the number of Part D LIS beneficiaries who would be eligible if the income threshold was 150% of the FPL and no asset test was imposed.
We then estimated the number of additional people who would be eligible for MSPs and LIS if the income threshold was raised from 150% to 200% of the FPL, with no restrictions on assets. The MCBS was used to estimate the number of MSP beneficiaries who have incomes between 150% and 200% of the FPL. That estimate was then subtracted from the total number of people with incomes between 150% and 200% of the FPL, with the latter estimate coming from the DYNASIM model, to produce an estimate of the number of additional people who would be eligible for MSP and LIS if the income threshold was raised from 150% to 200% of the FPL, with no asset test.
The resulting estimates overstate the number of people who would be newly eligible for the MSPs and LIS under the expanded eligibility criteria because the estimates do not account for the Medicare beneficiaries who are eligible for these programs under existing income and asset limits but are not enrolled.
To estimate average savings that beneficiaries might achieve through expansions in eligibility for the Medicare Savings Programs, we estimated average annual Medicare Part A and B liability for individuals not receiving cost-sharing assistance through the MSPs using the 2017 CCW. Liability was based on beneficiaries in traditional fee-for-service who do not have Medicaid and represents the amount of cost sharing that beneficiaries would incur if they do not have any form of supplemental coverage. We inflated the 2017 average amount to a 2020 value using the average of the values for the average annual rate of growth in the Part A deductible between 2017 and 2020 (2.3%) and the average annual rate of growth in the Part B premium between 2017 to 2020 (2.6%); averaging these two values gave us a growth rate of 2.4%. This method assumes no change in utilization between 2017 and 2020. Savings on Part B premium is calculated based on the 12 months of the standard Part B premium for 2020.
A similar process was used to calculate how much individuals would save if they received cost-sharing assistance through the LIS program. We calculated average annual cost sharing on prescription drugs for non-LIS, partial LIS, and full LIS enrollees using Part D prescription drug event claims from the 2017 CCW. We inflated average spending on prescription drug cost sharing for these three groups of enrollees using the average annual growth rate in the Part D deductible from 2017 to 2020 (2.8%). We calculated average savings on cost sharing by taking the difference in average spending between various groups (such as the difference between average spending by full LIS enrollees and partial LIS). The calculation of premium savings is based on the Part D base beneficiary premium for 2020 ($32.74 per month); premium savings range depending on the level of premium subsidy received (full LIS enrollees receive a full premium subsidy; partial LIS enrollees receive premiums subsidies ranging from 100% to 25% of the monthly premium; non-LIS enrollees receive no premium subsidy). Estimated out-of-pocket savings for specific drugs is based on 2020 data from the Medicare Plan Finder. Using the plan finder for zip code 20902 in Maryland, we entered specific drugs and retrieved annual cost-sharing information for non-LIS, partial LIS, and full LIS enrollees, using amounts for the lowest-cost plan in the zip code based on prices at Costco, CVS, and Giant pharmacies located within this zip code. As with the overall average cost sharing savings calculation, we calculated savings on cost sharing for specialty drugs by taking the difference in spending on specific drugs across the three enrollee groups (non-LIS, partial LIS, and full LIS enrollees).