The use of artificial intelligence and data analytics in life insurance
The emergence of data analytics and machine learning is providing insurers and reinsurers with new insights into how they drive and monitor their business.
The U.S. Department of Health and Human Services (HHS) Office of Inspector General (OIG) announced the final rule “Removal of Safe Harbor Protection for Rebates Involving Prescription Pharmaceuticals and Creation of New Safe Harbor Protection for Certain Point-of-Sale Reductions in Price on Prescription Pharmaceuticals and Certain Pharmacy Benefit Manager Service Fees, 1 which finalized the proposed rule2 published February 6, 2019, with a few key changes.
The Centers for Medicare and Medicaid Services (CMS) separately issued an interim final rule with a comment period (IFC) that adopts a Most Favored Nation (MFN) Model3 for Medicare Part B drugs. Final policies are adopted in an IFC when a federal agency finds good cause to issue a final rule without first publishing a proposed rule.4 CMS may make changes to the final policies in response to public comments or may confirm the final policies through a brief final rule published after the comment period closes.
While the rules were released as final and interim final, respectively, the road ahead is uncertain. It appears likely the rules will face legal challenges prior to or after the effective dates of their provisions. It is difficult to predict whether the policies finalized in these rules will reduce the total cost of drugs for the government and Medicare beneficiaries as intended.
Under the current Medicare program structure, drug manufacturers offer price concessions, known as rebates, to plan sponsors—or pharmacy benefit managers (PBMs) on their behalf—in exchange for drug formulary placement. Plan sponsors primarily use these rebates to reduce beneficiary premiums and government costs. This is possible because rebates are applied after drugs are dispensed (i.e., post-point-of-sale) as allowed under the current safe harbor protection for discounts under the Anti-Kickback Statute (AKS).5
Under the final policies, drug manufacturer rebates in Medicare Part D are to be excluded from the discount safe harbor. A new safe harbor allows plan sponsors to use the rebates to reduce drug costs at the point of sale (POS), allowing for the potential restructuring of post-POS manufacturer rebates as POS price concessions. The rule also creates a new safe harbor to retain protection for certain service fees paid by manufacturers to PBMs, though this article focuses on the changes to manufacturer rebates.
The final rule states that administration of pharmacy direct and indirect remuneration (DIR), which is a price concession between pharmacies and plan sponsors typically paid post-POS, is outside the scope of the rulemaking. The rule does not alter safe harbor protections for pharmacy DIR. Therefore, pharmacy DIR may continue to be applied after the POS.
A key change between the proposed and final rules for revisions to the discount safe harbor is that rebates offered from drug manufacturers directly to Medicaid managed care organizations (MCOs) will still be protected by the existing safe harbor if all conditions of the safe harbor are met. The rule only applies to Medicare Part D and does not apply to other government or commercial programs.
Additionally, the effective date for revisions to the discount safe harbor was updated from January 1, 2020, to January 1, 2022.
The revisions to the discount safe harbor are effective January 1, 2022. Medicare plan sponsors submit their bids (i.e., projected member premium plus government subsidies) for calendar year 2022 plans to CMS on June 7, 2021. Thus drug manufacturers, PBMs, plan sponsors, and other stakeholders need to begin working on changes now to ensure formularies, contracting, benefit designs, premiums, operations, and other key items consider the new rebate structure.
The rule provides limited guidance on how potential POS rebates under the revised discount safe harbor will be administered. For example, if, as expected, drug manufacturers continue providing rebates, new arrangements will be needed to administer the rebates as POS chargebacks, which are payments made from the manufacturer (directly or indirectly) to the pharmacy, equal to the negotiated rebate amount. HHS is vague about which entities can serve as POS chargeback administrators, noting their intention to provide flexibility and allow various types of entities to administer chargebacks.
Requiring rebates to be applied at the POS will reduce out-of-pocket cost sharing for certain beneficiaries and, without stakeholder behavioral changes, could increase beneficiary premiums and overall federal government costs. The rule will have widespread effects across all Medicare Part D stakeholders, including drug manufacturers, wholesalers, PBMs, pharmacies, plan sponsors, and beneficiaries. This includes direct cost impacts from the change in financial rebate dynamics, as well as operational and administrative costs and considerations. Stakeholder behavior changes are likely but difficult to predict.
Beneficiaries most likely to experience reduced out-of-pocket costs are those who pay coinsurance on rebated drugs and do not receive income-based government cost-sharing subsidies. Beneficiaries paying fixed copays or taking drugs not subject to rebates may experience premium increases without a material change in their out-of-pocket cost sharing. One study estimated the beneficiaries most likely to realize reduced overall costs with the rule could be the approximately 10% of beneficiaries who do not receive income-based subsidies and end the year above the initial coverage limit Part D benefit phase.6 The rule also notes other beneficiaries may realize reduced costs due to changes in copayments needed to maintain actuarial equivalence requirements or due to a projected reduction in the standard Part D deductible. In aggregate, the estimates summarized in the rule project overall beneficiary savings, driven by reduced cost-sharing realized by certain beneficiaries.7
In President Trump’s executive order,8 a requirement for finalizing the rule was that the Secretary of HHS confirm “the action is not projected to increase Federal spending, Medicare beneficiary premiums, or patients’ total out-of-pocket costs.” HHS Secretary Azar released a statement, “My extensive experience in this field, coupled with the fifteen year history of the program, supports my projection that there will not be an increase in federal spending, patient out-of-pocket costs, or premiums for Part D beneficiaries under the Final Rule implementing the Executive Order.”9
Publicly available studies on the topic referenced in the final rule, including those commissioned by HHS and from CMS, estimate an increase in premiums with this change, under scenarios with and without possible stakeholder behavior changes.10,11 Behavior changes contemplated included, for example, higher or lower negotiated price concessions in response to the rule. While premiums were projected to increase in all cited scenarios, the rule notes the premium increases observed in two of HHS’s selected behavior change scenarios are considered de minimis—i.e., the monthly increase per beneficiary, on average over ten years, is less than the $2 de minimis threshold used in other Part D contexts.
In the final rule, HHS notes that it believes the rule will reduce drug list prices. However, given the rule's limited applicability to only the Medicare Part D program and not to other markets, it is difficult to predict how list prices will change. It is possible the rule could affect stakeholder strategies in other markets.
The MFN Model is a seven-year test through CMS’s innovation center, which changes the way Medicare pays for drugs by basing the price off the cost of drugs in other selected countries and implementing a flat rate add-on provider payment, as opposed to the current model, which is based on Average Sales Price (ASP) plus a 6% add-on payment. The stated intent of the MFN Model is to lower Medicare’s drug costs. The model applies to certain Part B (i.e., physician-administered) drugs. Specifically, the initial cohort will include 50 high-spend drugs. Many of these drugs are used to treat hematology/oncology conditions, though drugs that treat rheumatology and neurology are also well represented on the list. The full drug list for year 1 is included in the IFC.12 Collectively, these 50 drugs represent approximately 75% of total Part B drug spend in 2019. CMS has indicated the list of drugs will be updated in future years as spend patterns change. Some categories of drugs are excluded, such as certain vaccines, radiopharmaceuticals, oral drugs, compounded drugs, and intravenous immune globulin products. Drugs intended to treat COVID-19 are also excluded.13
Medicare Part D-covered drugs are not impacted by this rule.
Beneficiaries: Beneficiaries who are enrolled in Medicare fee-for-service (FFS) and not covered by any other group plan are impacted by the MFN Model. This includes beneficiaries who are also enrolled in Medicare Supplement (i.e., Medigap) plans. According to the rule, Part B premiums are expected to decrease. Additionally, beneficiaries who utilize any of the 50 drugs included in the MFN Model can expect additional cost-sharing savings. Medicare Advantage beneficiaries are excluded from the MFN Model, but would benefit from the Part B premium decrease.
Providers: Providers who treat Medicare FFS patients will be mandatorily included in the MFN Model, with a list of exceptions. These excepted providers include children’s hospitals, Prospective Payment System (PPS)-exempt cancer hospitals, critical access hospitals, Indian Health Service facilities, Federally Qualified Health Centers, Rural Health Clinics, hospitals that are not “subsection (d)” and are paid on the basis of reasonable costs subject to a ceiling, extended neoplastic disease care hospitals, and certain acute care hospitals for the first two quarters of year 1. Note that 340B-covered entities are included in the MFN Model.14 Eligible providers can expect to see changes in reimbursement for the eligible MFN Model drugs. Providers will be impacted differently depending on which MFN Model drugs they administer and the proportion of FFS patients treated, which is likely to vary across different specialties.
Drug manufacturers: The MFN Model affects the price Medicare pays to providers. Manufacturers of drugs included in the MFN Model are not required to make any changes to the prices charged to providers. However, providers will likely suffer a financial loss if they are not able to acquire a MFN Model drug for a lower cost corresponding with their lower reimbursement, which could force manufacturers to reduce prices. This could lead to lower revenue for drug manufacturers for MFN Model drug utilization.
The basis of MFN pricing is that the United States will not pay more than what other comparable countries pay for eligible Part B drugs. Countries are considered comparable if they are members of the Organisation for Economic Co-operation and Development (OECD) and have a gross domestic product (GDP) per capita of at least 60% of the U.S. GDP. These criteria will be reevaluated quarterly. For the first quarter of year 1 of the MFN Model, the countries that meet the criteria for comparable pricing information are Australia, Austria, Belgium, Canada, Denmark, Finland, France, Germany, Iceland, Ireland, Israel, Italy, Japan, Republic of Korea, Luxembourg, Netherlands, New Zealand, Norway, Spain, Sweden, Switzerland, and the United Kingdom.15
The MFN price will be based upon the lowest GDP-adjusted country-level price for each drug. Pricing will include discounts and rebates when possible.16 CMS has indicated that the MFN discount is expected to be approximately 65% on average when fully phased-in, though discounts will vary by drug.17
The sources used to determine international prices typically do not include rebate arrangements between manufacturers and governments. CMS states that this will lead to the international prices used for the MFN to be potentially overstated from actual international prices.
The MFN Model will affect both provider payments and member cost sharing. First, the lower negotiated cost will reduce member cost sharing for FFS beneficiaries, which is 20% for Part B drugs. Second, changing the provider add-on payment from a 6% add-on to a flat rate add-on as defined by the rule ($148.73 in the first quarter of year 1), impacts the final provider reimbursement. The provider add-on payment may increase or decrease compared to the current level, depending on the ASP of each drug. According to CMS, the intention of switching from a percentage to a flat amount is to “remove the tie between drug cost and the add-on amount.”18 The flat add-on amount will be indexed quarterly for inflation.
Each drug and provider faces different implications, but Figure 2 shows an illustrative example of how the full implementation of the MFN pricing model (based on 2019 pricing levels) affects a provider and a member. We used the drug Eylea (aflibercept) as an example, because it is the highest-spend drug (based on 2019 Part B Rx drug spend) on the initial MFN list. We did not consider the impact to Part B member premiums in this example. The rule contains examples of gross cost for all drugs included in the initial quarter of the MFN Model.
|Provider Add-On Payment**||$54.19||$148.73||$94.54|
|Member Cost Sharing†||$191.47||$79.87||-$111.60|
* IFC Rule, Table 6, Q1 2019 price assuming MFN pricing fully phased in
** Current: 6% of gross cost. MFN Model: flat $148.73 per IFC rule
† 20% coinsurance (current: based on gross cost + provider add-on payment; MFN Model: based on gross cost)
The MFN Model is expected to be effective as of January 1, 2021, and will run through the end of 2027. MFN pricing will be phased in over the first three years of the program, using a blend of MFN price and the current ASP methodology.19
It is unknown how the industry will react to the MFN interim final rule, but some stakeholders will likely have strong opinions and may consider options to push back against the rule.20, 21 Some providers may seek to increase their charges to make up for revenue lost due to lower reimbursement, particularly in the commercial health insurance and Medicare Advantage markets.
Additionally, manufacturers may increase prices in non-Medicare FFS markets to offset the cost impact of the MFN rule. CMS included language, including excluding MFN utilization from ASP pricing, to help avoid manufacturers increasing cost in other non-Medicare FFS markets. Manufacturers may also consider not entering or leaving countries included in the MFN group where the price is significantly lower than the current ASP.
Beneficiaries who receive MFN Model drugs from providers who participate in the program should see lower cost sharing. CMS is also expecting the rule to lead to a decrease in the Medicare Part B premiums, which would lead to savings for all Medicare beneficiaries.
1HHS (November 20, 2020). Final Rule to Bring Drug Discounts Directly to Seniors at the Pharmacy Counter - RIN 0936-AA08. Retrieved November 20, 2020, from https://www.hhs.gov/sites/default/files/rebate-rule-discount-and-pbm-service-fee-final-rule.pdf.
2HHS (February 6, 2019). Proposed rule is published in the Federal Register. Retrieved November 20, 2020, from https://www.federalregister.gov/documents/2019/02/06/2019-01026/fraud-and-abuse-removal-of-safe-harbor-protection-for-rebates-involving-prescription-pharmaceuticals.
3CMS (November 20, 2020). CMS-5528-IFC MFN. Retrieved November 20, 2020, from https://innovation.cms.gov/media/document/mfn-ifc-rule.
4See the rule making process at https://www.federalregister.gov/uploads/2011/01/the_rulemaking_process.pdf.
5See additional details about AKS as it relates to pharmacy rebates: https://www.milliman.com/en/insight/changing-the-rebate-game-a-primer-on-the-hhs-proposed-rule-to-shift-drug-rebates-to-pos
6Margiott, T. & D’Anna, S. (October 26, 2020). Impact of Medicare Part D Drug Manufacturer Rebates at the Point of Sale. Milliman Report. Retrieved November 23, 2020, from https://www.pcmanet.org/wp-content/uploads/2020/11/Part-D-POS-Rebates_FINAL.pdf.
8White House (July 24, 2020). Executive Order on Lowering Prices for Patients by Eliminating Kickbacks to Middlemen. Retrieved November 20, 2020, from https://www.whitehouse.gov/presidential-actions/executive-order-lowering-prices-patients-eliminating-kickbacks-middlemen/
9HHS (November 20, 2020). Secretary Azar Confirmation In Response to Executive Order on Lowering Prices for Patients by Eliminating Kickbacks to Middlemen. Retrieved November 20, 2020, from https://www.hhs.gov/about/news/2020/11/20/secretary-azar-confirmation-in-response-to-executive-order-on-lowering-prices-for-patients.html.
10CMS Office of the Actuary (August 30, 2018). Proposed Safe Harbor Regulation. Retrieved November 20, 2020, from https://www.cms.gov/Research-Statistics-Data-and-Systems/Research/ActuarialStudies/Downloads/ProposedSafeHarborRegulationImpact.pdf.
11Klaisner, J., Holcomb, K., & Filipek, T. (January 31, 2019). Impact of Potential Changes to the Treatment of Manufacturer Rebates. Milliman Client Report. Retrieved November 20, 2020, from https://aspe.hhs.gov/system/files/pdf/260591/MillimanReportImpactPartDRebateReform.pdf.
18CMS (November 20, 2020). FACT SHEET: Most Favored Nation Model for Medicare Part B Drugs and Biologicals Interim Final Rule With Comment Period. Retrieved November 23, 2020, from https://www.cms.gov/newsroom/fact-sheets/fact-sheet-most-favored-nation-model-medicare-part-b-drugs-and-biologicals-interim-final-rule.
20PhRMA (November 20, 2020). PhRMA Statement on Most Favored Nation Model and Finalized Rebate Rule. Retrieved November 23, 2020, from https://www.phrma.org/Press-Release/PhRMA-Statement-on-Most-Favored-Nation-Model-and-Finalized-Rebate-Rule.
21Community Oncology Alliance (November 20, 2020). COA Statement on Trump Administration Most Favored Nation Drug Pricing Announcement. Retrieved November 23, 2020, from https://communityoncology.org/coa-statement-on-trump-administration-most-favored-nation-drug-pricing-announcement/.